Concrete in The Service of Mankind - Appropriate Concrete Technology
Concrete in The Service of Mankind - Appropriate Concrete Technology
Hydraulic Structures
P.Novak, A.Moffat, C.Nalluri and R.Narayanan
Structural Grouts
Edited by P.L.J.Domone and S.A.Jefferis
For details, contact the Promotions Department, E & FN Spon, 26 Boundary Row,
London SE1 8HN, UK. Tel: 0171865 0066, Fax: 01715229623.
Appropriate Concrete Technology
Proceedings of the International Conference held at the University of
Dundee, Scotland, UK on 2426 June 1996
Edited by
Ravindra K.Dhir
Director, Concrete Technology Unit
University of Dundee
and
Michael J.McCarthy
Lecturer, Concrete Technology Unit
University of Dundee
E & FN SPON
An Imprint of Chapman & Hall
London Weinheim New York Tokyo Melbourne Madras
Published by E & FN Spon, an imprint of Chapman & Hall, 26 Boundary Row, London
SE1 8HN, UK
This edition published in the Taylor & Francis e-Library, 2006.
To purchase your own copy of this or any of Taylor & Francis or Routledges collection of
thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.
Chapman & Hall, 26 Boundary Row, London SE1 8HN, UK
Chapman & Hall, GmbH, Pappelallee 3, 69469 Weinheim, Germany
Chapman & Hall USA, 115 Fifth Avenue, New York, NY 10003, USA
Chapman & Hall Japan, ITP-Japan, Kyowa Building, 3F, 221 Hirakawacho, Chiyoda-ku, Tokyo
102, Japan
Chapman & Hall Australia, 102 Dodds Street, South Melbourne, Victoria 3205, Australia
Chapman & Hall India, R.Seshadri, 32 Second Main Road, CIT East, Madras 600 035
First edition 1996
1996 E & FN Spon
The cover illustration shows the Glendevon Reservoir spillway, constructed using GGBS concrete.
Photograph courtesy Castle Cement Ltd.
PREFACE
The specification of concrete has long been a major source of conflict between the
engineer, contractor and supplier. Much of this conflict stems from the increasing
litigious world in which modern construction takes place. Equally, it may arguably to be
due to the misunderstandings arising from what is the appropriate performance that will
be required from the materials used themselves and the structure manufactured.
To overcome this conflicting interest, there will have to be an evolutionary approach
away from the traditional method of specification and towards performance criteria. This
will also facilitate the formulation of integrated, harmonised international standards.
This transition will not be easy and it will almost certainly transfer responsibility for
some concrete material decisions to the supplier. It can only be hoped that total quality
management, which has done so much for the manufacturing sector, can also provide the
same degree of confidence to clients in concrete construction. Since most international
specifications bodies are, in principle, committed to this approach to specification by
performance, it is important that the implications of this are debated at a major forum
such as this international conference.
The versatility of concrete continues to be underlined through the development of new
binder technologies, alternative materials for reinforcement and novel design and
construction techniques, which continues apace.
The Proceeding of this Conference Appropriate Concrete Technology deal with all
of these subject areas under six clearly identified themes: (i) Criteria for Appropriateness,
(ii) Implications of Harmonisation, (iii) Versatility of Concrete, (iv) Binder Technology,
(v) Non-Ferrous Reinforcement and (vi) Design and Construction. Each theme started
with a Leader Paper presented by the foremost exponents in their respective fields. In
total, 65 paper were presented during the 3 day International Conference and compiled
into these Proceedings.
Ravindra K Dhir
Michael J McCarthy
Dundee
February 1996
ORGANISING COMMITTEE
Concrete Technology Unit
Dr H Z Al-Abideen
Assistant Deputy Minister
Ministry of Public Works and Housing, Saudi Arabia
Professor M S Akman
Professor of Civil Engineering
Istanbul Technical University, Turkey
Professor M G Alexanader
Professor of Civil Engineering
University of Cape Town, South Africa
Professor Carmen Andrad
Research Professor
Instituto Eduardo Tonoja of Construction Sciences, Spain
Mr J A Bickley
Implementation Manager
Concrete Canada, Canada
Professor J M J M Bijen
General Manager
Institute for Materials and Environmental Research, Netherlands
Professor A M Brandt
Head of Section
Institute of Fundamental Technological Research, Poland
Mr J C Caballero
General Director
Instituto Del Cemento Portland, Argentina
Professor A Ceccotti
Fornitek Laboratory, Canada
Professor O E Gjrv
Professor, Division of Building Materials
Norwegian Institute of Technology, Norway
Professor T C Hansen
Professor of Building Materials
Technical University of Denmark
Dr G C Hoff
Senior Associate Engineer
Mobil Research & Development Corporation, USA
Professor S Y Huang
Director
Shanghai Institute of Building Materials, China
Professor C Jaegermann
Professor Emeritus
National Building Research Institute, Israel
Professor F K Kong
Head of Division of Structures and Construction
Nanyang Technological University, Singapore
Professor I E Korish
Professor of Reinforced Concrete Structures
University of Alexandria, Egypt
Professor S Nagataki
Professor, Department of Civil Engineering
Tokyo Institute of Technology, Japan
Dr A M Paillere
Technical Director
Laboratoire Central des Ponts et Chausses, France
Mr W G Ryan
Chief Executive Officer
Cement and Concrete Association, Sydney, Australia
Professor A E Sarja
Research Professor
Technical Resarch Centre of Finland (VTT)
Professor P Schiessl
Director
Technical University of Aachen, Germany
Professor S P Shah
Director
ACBM, Northwestern University, USA
Professor T P Tassios
Director of Laboratory Concrete Structures
National Technical University of Athens, Greece
Professor D N Trikha
Director
Structural Engineering Research Centre, India
Professor K Tuutti
Skanska Teknik AB, Sweden
Professor F H Wittmann
Director
Institute for Building Materials, Switzerland
Professor A V ZaBegayev
Head of Department RC Structures
State University of Civil Engineering, Russia
TECHNICAL COMMITTEE
(from United Kingdom)
Professor A W Beeby
Professor of Structural Design, University of Leeds
Dr R D Browne
Consultant, Roger Browne Consultancy
Professor J H Bungey
Professor of Civil Engineering, University of Liverpool
Dr P Chana
Divisional Director, CRIC, Imperial College
Mr P M Deason
Managing Director, Trafalgar House Technology Ltd
Professor R K Dhir (Chairman)
Director, Concrete Technology Unit, University of Dundee
Mr C R Ecob
Divisional Director, Mott MacDonald Ltd
Professor F P Glasser
University of Aberdeen
Dr T A Harrison
Technical Director, British Ready Mixed Concrete Association
Professor P C Hewlett
Director, British Board of Agrment
Mr J Innes
Chief Bridge Engineer, Scottish Office
Mr K A L Johnson
Director, AMEC Civil Engineering Ltd
Professor A E Long
Director, School of the Built Environment, Queens University of Belfast
Mr G G T Masterton
Director, Babtie Group
Professor G C Mays
Head, Civil Engineering Group, Cranfield University
Mr L H McCurrich
Technical Director, Fosroc International Ltd
Dr J B Menzies
Engineering Consultant, John B Menzies
Professor R S Narayanan
Senior Partner, S B Tietz & Partners
Dr P J Nixon
Head, Inorganic Materials Division, Building Research Establishment
Dr D J Pollock
Director, Sir William Halcrow & Partners Ltd
Professor D C Spooner
Director, Materials and Standards Division, British Cement Association
Dr H P J Taylor
Director, Costain Building Products Ltd
Professor P Waldron
Director, Centre for Cement & Concrete, University of Sheffield
Mr KR Wilson
Technical Director, G Maunsell & Partners
SUPPORTING BODIES
Advantage Precast
Allied Bar Coaters
AMEC Civil Engineering
Ash Resources Ltd
Babtie Group
Blue Circle Cement
Boral Pozzolan Ltd
British Board of Agrment
British Cement Association
Building Research Establishment
Capco Test Equipment
Castle Cement Ltd
Cementitious Slag Makers Association
Cem-FIL International Ltd
CFPI
City of Dundee District Council
Colebrand Ltd
Du Pont de Nemours Internatioanl SA
E & FN Spon
ECC International (Europe) Ltd
Elkem Materials
Fibraflex
Fosroc International Ltd
Germann Instruments A/S
GOMACO International Ltd
Grace-Cormix
Harris Speciality Chemicals
I W Farmer & Partners Ltd
John Fyfe Ltd
Kerner Greenwood (UK) Ltd
Lafarge Aluminates
Minelco Ltd
Mott MacDonald Special Services
National Ash, National Power plc
PowerGen plc
Protovale (Oxford) Ltd
Ready Mixed Concrete (United Kingdom) Ltd
Rugby Cement
Scottish Enterprise Tayside
Scottish Power plcAsh Sales
Tarmac Topmix Ltd
TBV Stanger
Thomas Telford Publishing
Transport Research Laboratory
VNC Association of the Netherlands Cement Industry
Wexham Developments
CONTENTS
Preface v
Introduction vii
Organising Committee viii
International Advisory Committee ix
Technical Committee (from United Kingdom) xi
Supporting Bodies xiii
Sponsoring Organisations xiv
Exhibitors xv
Contents xviii
Leader Paper
Performance criteria for structural concrete 2
Professor G Somerville, British Cement Association, United Kingdom
Who likes concrete? 13
S MacCraith
Perfomance criteria: The pains and strains of change 22
C Oltean-Dumbrava
Application of a statistical methodology for the evaluation of the potential 28
market of new concrete products on the construction industry
C C Videla, H de T Solminihac, J de D S Ortzar and J P Retamal
Appropriate concrete design 36
J Wang, D Knight and L Swann
Assuring performance of concrete structures through a durability audit 47
B K Marsh and P J Nixon
Realistic prediction of time-dependent response of concrete structures 57
J Lopatic, and F Saje
Leader Paper
Implication of harmonisation 67
Mr R S Narayanan, SB Teitz Partners, Consulting Engineers, United Kingdom
Application of European concrete standards 72
D Stoelhorst and G P L den Boer
Implications of structural Eurocodes 81
S B Desai
Fracture mechanics parameters of concrete: Test methods development and 90
harmonization of standards
S N Leonovich
Effects of concrete creep on the ultimate compressive strength of concrete 99
columns
A F L Wong, C Arnaouti and N K Raji
Health and safety in construction in Nepal 108
R P Adhikari
Foundations for low rise buildings in Romania and UK: Comparisons and 115
effects
C Oltean-Dumbrava and W G Carter
Site workersthe challenge 121
D P Barnard
Computer-based concrete specification 132
A Osborne, A D Pullen and J B Newman
Leader Paper
The versatility of concrete 142
Professor F Wittmann Institute for Building Materials and Dr J Gebauer and Dr
Professor F Wittmann, Institute for Building Materials and Dr J Gebauer and Dr
R Torrent, Holderbank Management & Consulting Limited Switzerland
Concrete masonry: A different dimension for housing in developing countries. 154
The 40 years experience in Colombia
G G Madrid and L G Pelaez
Material matters in affordable housing 169
J Morris and R A Kruger
Cement stabilisation of soil for the production of building blocks 177
J Morris and G F Blight
An innovative labour-intensive method for construction of arch bridges using 186
uncut stone and mortar
R G D Rankine and G J Krige
Manpower motivation improves quality in construction 194
M S Chishty and M A Choudhry
Appreciating low-strength concrete 202
M A Abdul-Salam
Behaviour in extreme climates of concrete made with different types of cement 209
C C Videla, P T Covarrubias and J M D Pascual
Problems about optimization of porosity and properties of aerated concrete 223
R Cabrillac and Z Malou
The use of foamcrete for affordable development in third world countries 232
E P Kearsley
Some aspects of the design and production of foamed concrete 243
D E Wimpenny
Building underground: The concrete contribution 253
M A Clarke
Leader Paper
Hydraulic cementsNew types and raw materials and radically new 262
manufacturing methods
Professor A Samarin, University of Wollongong, Australia
Hydraulic behaviour of non-expansive sulfoaluminate cement 274
Ch Malami, Th Philippou, T Tsakiridis and V K Rigopulou
Phase equilibria in the CaO-Al2O3-SiO2-H2O system in relation to blended HAC 281
cements
K C Quillin
High performance concrete containing modified rice husk ash 291
B Chatveera and P Nimityongskul
Slag alkaline polymer cement concretes 298
P V Krivenko, V A Raksha and L V Raksha
RHA-cement as a replacement for Portland cement in rural Tanzanian villages 307
P Stroeven and E I Sabuni
Effect of the binary admixtures of fly ash and shale ash on the fluidity and 321
compressive strength of cement concrete
N-Q Feng, Q-F Zhuang and D-H Wang
A study of carbonation in binding system with cement and hydraulic admixtures 328
I Robu
Application of the cement hydration equation to concrete 335
K K Sideris, M S Konsta and C G Karayannis
Properties of high temperature-humidity cured OPC-GGBS fibre concrete 344
roofing tiles
H C Uzoegbo
Micropore structure variances during hydration of cement pastes 355
Ch Ftikos, A Georgiades and J Marinos
Improving the properties of adobe by pozzolanic materials 363
B Baradan
Leader Paper
Tailoring the properties of concrete structures with appropriate non-ferrous 369
reinforcements
Professor A E Sarja, Technical Research Centre of Finland, Finland
Development of cover meters for stainless reinforcementTwo successful 385
approaches
J C Alldred
Structural design using epoxy coated reinforcement 395
J Cairns
Supercover concrete: A new method of preventing reinforcement corrosion in 408
concrete structures using GFRP rebars
C Arya and G Pirathapan
Cracking of carbon fibre reinforced mortars 420
L Kucharska and A M Brandt
Low cost fibre reinforced cement products by using inexpensive additives 430
I Papayianni, N Economou, D Leventis and G Xanthakos
Biaxial fracture behaviour of fibre reinforced concrete 441
E K Tschegg, M Elser and S E Tschegg-Stanzl
Designing structures using non-ferrous reinforcement 453
J L Clarke and P ORegan
J L Clarke and P ORegan
A preliminary investigation into the comparative performance of FRP and steel 462
reinforcement in medium and high strength concrete
C Ellis and E M Ulas
Structural behaviour of bamboo reinforced concrete beams and slabs 472
K Ghavami, A J S Mina, H C L Junior and N P Barbosa
Leader Paper
Design and construction 483
Professor T P Tassios, National Technical University of Athens, Greece
The design of framed structuresA proposal for change 495
A W Beeby and F Fathibitaraf
Development of moment continuity connection for simply supported precast 506
concrete bridges
G Davies, K S Elliott and A A H Arshad
Evaluation of honeycomb concrete 518
G W Seegebrecht, S H Gebler and B G Stejskal
Flexible reinforced concrete beams on elastic rubber foundation model 525
A Hassani and D W Cox
Portal A-frame vs portal H-frame structures: A choice based on total costs 536
D R Lemelin
Preliminary design rules for structural steel shearheads in concrete slabs 544
P S Chana
Shear resistance of reinforced concrete members at high temperatures 551
S B Desai, N K Subedi and K S Virdi
Repeated and cyclic loading tests on precast concrete beam-to-column 563
connections
Y C Loo, B Z Yao and S Takheklambam
SISMO-building technology: Plain concrete in high rise buildings 574
E van Rensbergen, R Bellers and D van Gemert
Factory & site productionA comparison 588
H P J Taylor
Development of self-cure concrete 596
R K Dhir, P Hewlett and T D Dyer
Vibrators for compacting concreteA radical look 609
P G F Banfill
The planning and design of concrete mixes for transporting, placing and 616
finishing
G G T Masterton and R A Wilson
Roller compacted concrete: Development and use in Queensland, Australia 629
H Reid
CLOSING ADDRESS
Chairman Dr T A Harrison British Ready Mixed Concrete Association, United
Kingdom Presented by Professor P C Hewlett, Director British Board of
Agrment, United Kingdom
Index of Authors 638
Subject Index 642
OPENING ADDRESS
Leader Paper
INTRODUCTION
As a student in the 1950s, the author was taught the rudiments of structural design in
terms of using numerical methods to provide adequate strength, stiffness, stability and
serviceabilitywith some guidance on how to do that at least first cost. The performance
criteria, margins, and factors of safety were given in Codes, and accepted without
question. These were assumed to prevail as soon as the structure was built, as well as
during its entire useful life (although this was never specified). This was not checked in
any numerical sense, but assumed to be so, if material and workmanship specifications
were satisfied during the construction phase. In short, the provision of durability i.e. the
maintenance of the required margins and factors, was on a purely prescriptive basis.
In the 1990s, that basic scenario has not really changed. To be sure, the prescriptions,
specifications and factors of safety have altered, but the overall design approach is
fundamentally the same.
Performance criteria for structural concrete 3
This is no different from establishing minimum requirements for safety and (say) crack
width, in structural design; without such metre-sticks, it is not possible to compare
alternative options in whole life cost/benefit terms.
Performance criteria for structural concrete 5
In performance criteria terms, durability is not the only issue. At present in the UK,
the construction industry is itself undergoing change, in the procurement process. For
some time now, concrete technology has had to cope with large continuous pours, with
pumping, and with round-the-clock production in a precast factor, and with delivery in a
ready-mix truck. However, demand for increased productivity and quality, linked to
speed of construction, is imposing increased pressure on technology for the construction
phase. Design is becoming more closely related to the construction process, and therefore
we have to strike a balance between buildability and satisfactory performance in service.
This means that performance criteria must be derived to permit that to happen, i.e. to
balance the needs both of buildability and durability; the merits of alternative design
options can then be judged on a common basis.
In brief, the time has come when durability should be considered as an integral part of
structural designideally based on the same principle of establishing margins or factors
of safety, in providing resistance against known loads, while satisfying established
performance criteria.
The purpose of this paper is to propose a framework for such a design method, before
going on to suggest some limiting performance criteria.
Data of the type provided by Paterson [3] and Wallbank [4] clearly demonstrate that the
level of deterioration in individual cases is dictated by a combination of factors, in which
design and construction issues are significant. The provision of durable concrete is
important, but it is possible to have failure even with the best materials, if design and
construction standards are poor.
Moreover, the spectre of obsolescence also enters the arena. Many buildings and
bridges have had to be upgraded or replaced, because their functional needs have
changedquite apart from any decrease in technical performance. In addition, different
components in individual artifacts (eg. cladding in buildings; expansion joints in bridges)
have been shown to have useful lives much less that those for the basic structures.
Growing recognition of this situation has led to the development of a performance
profile approach to design, eg. by White [5], for buildings, and by BSI [6] in a formalised
general approach. It is not always possible to precisely define the useful life of a structure
at the design stage or to second-guess the need for future upgrading. However, in many
cases an informed decision can be taken; for example, in the UK, much of the motorway
network is being upgraded within 3040 years of construction.
This leads to a basic proportion as follows:- durability should be an integral part of
design: the essential requirement is to ensure fitness for purpose, while taking account, as
far as possible, of future functional and financial needs, in whole life costing terms.
Table 2 shows the five essential elements which make up the detailed process of
structural design. To gain acceptability, durability design has to be seen alongside these,
Appropriate concrete technology 6
and parallel proposals are also made in Table 2. Currently we concentrate on item 5,
while trying to improve, via research mainly associated with items 1 and 3. Missing is
any serious consideration of items 2 and 4, and yet both are fundamental to a quantitative
approach; we need to know what we are trying to achieve, in what will never be an exact
science, and we need to have some confidence in the predictions we make.
To understand better what is intended here, Figure 1 is proposed as a possible
relationship between the elements in Table 2. Each aspect of Figure 1 will now be
considered in more detail.
practice indicate that ponding and run down of water is probably more severe than spray,
but that the process of wetting and drying can override that.
Table 3 also stresses the need not only to predict deterioration itself but also to assess
the possible effects on structural resistance. Some structural forms are more sensitive to
the effects of deterioration than others, and the risk factor requires evaluation.
A proposal for this is shown in Figure 2. The front face of the cube is used to make a
statement about the importance of the structure or element, in terms of life and
criticality. While there has been discussion in the literature regarding specifying design
lives for different types of structure in years, this may not be necessary. The proposal
here is that zones be defined (A-G in the Figure), which set down performance
requirements in terms of life and criticality. Criticality is intended to recognise both the
importance of structural elements in load-carrying terms, and also the difficulty of repair
or replacement, and the disruption that this would cause. A possible grading for criticality
might be:
High : failure would cause cessation of function and/or major disruption during remedial
work.
Medium efficiency of operation would be reduced, but replacement/remedial work could be
: done out of normal working hours.
Low : not critical. Any necessary maintenance or remedial work could be done without
inconvenience.
If necessary, life could be defined in years, but the zonal approach forces designers to
take early decisions on maintenance/replacement strategy, related to perceived future
functional requirements. It also helps define limiting performance criteria for durability.
This brings us to the third axis of the cubethe performance profile rating.
Rigorous design for durability effectively requires the definition of appropriate
performance criteria for each aggressive actionequivalent to maximum deflections or
crack widths in structural design. The intention can be illustrated by considering
corrosion; possible limiting criteria are:
1. The mean maximum level of a carbonation or chloride front should not penetrate more
than (a)% of the nominal cover.
2. Corrosion has just started, i.e. the critical front has reached the reinforcement, both
oxygen and water are available, and, in the case of chlorides, the critical chloride
threshold level has been reached.
3. Corrosion has caused cracking parallel to the reinforcement and the crack width is
equal to (b)mm.
4. Corrosion has removed (c)% of the cross-section from (d)% of the reinforcement.
If these are then related to Figure 2, then examples of limiting criteria which might then
emerge are:-
Zone A : mean maximum level of a chloride profile should not reach the rebar for [15
years]; loss of rebar section (<10%) after [30 years].
Zone G: mean maximum level of the chloride profile should not reach the rebar for [100
years].
These examples require very different design solutions. The concern at this point is with
performance criteria for structural elements. The objective is to create flexibility,
whereby the designer can consciously make a choice commensurate with his structure
and its maintenance/replacement strategy. In effect, this approach is an extension of
methods already in use for assessment work, which involve the definition of a range of
damage levels.
Having established limiting criteria, the next question is how can a designer ensure
that these are satisfied. One obvious direct approach is via Zone B in Table 3. For one-off
special structures, this may be the best way forward. However, for general use, there is an
alternative, which is outlined below.
This is illustrated in Figure 3. The actions foreseen are in the diagram itself, and the
Notes indicate how the system would work.
The whole emphasis in Figure 3 is still on durability design. However, this would have
to be considered alongside the needs of conventional structural design (Table 2) and of
buildability. As an example, consider a bridge using precast standard bridge beams. For
load bearing and prestressing, probably a Grade 60 concrete would be required. For rapid
production, a high early strength would also be required. The structure could be in Zones
F or G in Figure 2, with a consequential requirement that the chloride profile should not
reach the tendons for (say) 100 years. These basic needs then establish a foothold in
Table 4, and, assuming that reliable performance profiles are available, a choice can be
made from various resistance options, to achieve the required performance.
Table 4 Shopping list of resistance options for
corrosion
A Materials Concrete quality - mix proportions
- mix ingredients
Cover - minimum
- tolerances
Permeable formwork Concrete protection - sealers
- coatings
- penetrants
- layers
Rebar protection - epoxy coating
Special steels
Non-corrodible reinforcement Cathodic protection
B Design Design concept Structural detailing Cladding, services, fittings, finishes
Performance criteria for structural concrete 11
CONCLUDING REMARKS
A proposal is made for a framework for durability design. It is not a ready-made system,
but is put forward at this time to stimulate discussion, and to give a sense of direction to
all the current R & D work on the different facets of durability.
The emphasis is on design, while taking account of material and construction matters.
Design is not synonymous with analysis, and will never be an exact science. However, by
properly defining performance criteria (both in breadth and in depth), then the strength of
predictive modelling can be harnessed, to produce structures which better meet our needs,
in life cycle terms.
Notes to Figure 3
1. In the overall framework, at least 3 possible levels of design are foreseen:
(a) A recipe approach, similar to current practice.
(b) A more rigorous numerate approach, as briefly outlined in the text to Table 4.
(c) An analytical approach, as briefly mentioned in zone B of Table 3, while meeting
the performance requirements of Figure 2, and supported by the shopping list in
Table 4.
Appropriate concrete technology 12
2. It is important that all possible aggressive actions are identified and treated. Some,
such as sulfate attack or abrasion can be dealt with purely by a recipe approach.
However, others, such as corrosion and freeze-thaw, may produce synergetic effects.
3. Conventional structural design is on a semi-probabilistic basis. This suggests that
durability design should be on the same basis; however, support should also be
provided by Risk Analysis.
REFERENCES
1. SOMERVILLE G. The design life of concrete structures. The Structural Engineer Vol. 64A, N
2, February 1986, pp 6071, The Institution of Structural Engineers, London.
2. CONCRETE BRIDGE DEVELOPMENT GROUP. Proceedings of a 1-day seminar on Whole
Life CostingConcrete Bridges, 25 April 1995, the Concrete Bridge Development Group,
Crowthorne, Berks.
3. PATERSON A.C. The structural engineer in context. The Structural Engineer Vol. 62A, N11,
November 1984, pp 335342, The Institution of Structural Engineers, London.
4. WILLBANK E.J. Performance of concrete in bridges: a survey of 200 highway bridges. A report
prepared for the Department of Transport by G.Maunsell & Partners, HMSO, London, April
1989.
5. WHITE K.H. Building performance and cost-in-use. The Structural Engineer Vol. 69, N7, pp
148151, April 1991. The Institution of Structural Engineers, London.
6. BRITISH STANDARDS INSTITUTION. Guide to: Durability of buildings and building
elements, products and components BS7543:1992, BSI, London.
7. RAISS M.E. Design details for durable grouted post-tensioned concrete bridges. Proceedings of
a 1-day seminar held by the Concrete Society and the Concrete Bridge Development Group, 18
May 1994, The Concrete Society, Slough.
WHO LIKES CONCRETE?
S MacCraith
University College Dublin
Eire
ABSTRACT. People live with concrete. The paper deals with the public
perception of concrete, particularly as revealed by the misuse of the word
itself. A survey of peoples attitudes to concrete is described briefly. The
problems in achieving satisfactory concrete surfaces are outlined. The
author describes a new serviceability limit state of appearance which
might be an answer to the vexed question of what is an appropriate
surface in a given situation.
INTRODUCTION
families moving to the country, wrote city folk embarked on the intrepid transition to
fresh air and country living, far from the concrete and grime. (5). There is the subtle
association of concrete with grime, to add emphasis to the point. The experts plan to
place a huge ring of cement underground around the foundations (6). The confusion of
cement with concrete is quite common.
Other examples can be noted. This awful structure must surely be the embodiment in
concrete of the attitude which spawned. If the planners persist in putting concrete
and cars before people, the city will die. Dublin is depressing enough in the grey
weather, with its lines of concrete-rendered housing in bleak and treeless streets. .....a
campus landmark to atone for the concrete cock-ups of the sixties and seventies. .....an
orgy of concrete........ .....a huge steel-and-contrete butterfly..... A jungle of concrete,
glass and indifference,.......tall inanimate staring columns of concrete blocks......
The point at issue here is not the validity of what is being stated but rather the use of
one of our best known building materials to create a distasteful picture by association.
Why is this so? Why do people misuse language like this? Lack of education? A
wrong type of education? A distrust of what engineers, architects, planners, etc. do,
especially if it is out of taxpayers money, in order to create a civilised living for all?
Society needs houses, clean water, comfort, buildings of every type, good roads and
railways and so on. Whether tastefully done or not, such monuments to societys needs
must not, in the eyes of the beholders, be a blur on the environment. The deliberate
choice of the word concrete, noun or adjective, to describe any of them, whether the
concrete is visible or not in the finished product, immediately condemns them to the
realm of awfulness.
The concrete industry has adverted to this phenomenon on occasions in the past. In
1991, that marvellous magazine Concrete Quarterly, now sadly no longer with us
(although at the time of writing, hopes of its revival exist), Martin Clarke quoted Max
Hutchinson as saying that The words concrete and architect occupy a special place in the
publics affections. It is a secure place rather like a prison wing reserved for special
offenders (7). However, he also refers to another, and more positive, side to concrete.
Political commentaries and speeches are liberally, and increasingly, dotted with concrete
proposals and relationships being cemented with concrete solutions and plans for
concrete action. There is demonstrated here a certain dichotomy, a trace of
ambivalence about the term concrete. While being, on the one hand, regarded as
synonymous with ugliness and lowliness, it is also, on the other hand, used to convey an
impression of strength, solidity, reliability and durability. But not beauty.
A similar contradiction can be seen to exist in peoples attitudes to animals. Pets are
generally loved for lots of good reasons, their beauty, companionship, attractiveness,
cuddliness etc. Yet, at the same time, the English language is full of the most derogatory
references to animals. For example, sly people are compared to foxes, horrible-looking
foods are called dogs dinners, a person may be as blind as a bat or as dirty as a rat. There
are many others. This of course may be of little comfort to lovers or promoters of
concrete. But it is nonetheless an interesting aside to the thinking of unthinking people.
Appropriate concrete technology 16
Parallel to all this activity, but eventually converging with the concreting, the required
reinforcement will have been determined, detailed, scheduled, ordered, cut, bent,
bundled, labelled, transported, assembled and so positioned in the formwork that it will
simultaneously fulfil its structural duty and also be adequately protected from corrosion
for the rest of its life. Even here, the potential exists for something somewhere along the
line to go wrong in a way that could affect the long-term appearance of the concrete. The
concrete, containing its cargo of reinforcement, must next be adequately but not too
enthusiastically compacted by vibration. The next few days will be crucialassurance of
adequate curing is vital. Freezing weather, hot days, windy weather, variable humidity,
all impact in their own way on the curing period. There are other matters that add to the
drama. The materials and construction of the formwork play a large role in how the
finished concrete will look. The size and extent of the pour may cause headaches
human and concreteif either is very large.
And then, when the formwork is stripped, the conception period is over and the
concrete is bornlife now begins in earnest for the concrete. The rain will lash, the wind
will howl, the sun will scorch, the frost will bite, dust will blow and settle and unsettle
again, oxygen and carbon dioxide will both permeate through the concrete, each with its
own hidden agenda over time. The pollution of a billion exhaust/kilometres will seek
refuge in the pores of the concrete. The spores of algae and fungi and mosses will
conspire to see if a living can be made off the concrete surface. Life with the neighbours,
the bricks, stones, windows, polymers, metals, etc., may be a little strained. On top of all
these there are the effects of the stresses and deflections and cracking due to the variety
and extent of dead and applied loads. In brief, weathering has begun!
It may seem, after all this, that to achieve a concrete finish that will be perfect for all
time is impossible. It isnt. But it is so nearly impossible, even with present day
technology, that it would be unwise to try always to get it. Perfection can only be bought
at a price and the price is not always acceptable or available.
This is a simple fact of life. But it does not help the perception of concrete by the
general public. Whatever about perfect concrete, will people accept imperfect concrete?
There is a challenge here for the whole concrete industry, namely, to educate the public
to understand and appreciate concrete at both a conscious level and more importantly if
less easily, at a subconscious level. The public may still not like it, but it could now
rationalise its dislike.
Limit States
Modern structural design codes allow for two principle limit states. The first and more
important one is the limit state of collapse, or more commonly, the ultimate limit state
(ULS). The primary concern of the ultimate limit state is with safety of people and is not
an issue as far as this article is concerned. Secondly and less importantly, there is the
serviceability limit state (SLS). According to IS326 (10) and BS8110 (11), structures
should not become unfit for their purpose by collapse, overturning, buckling (ultimate
limit state), deformation, cracking, vibration, etc. (serviceability Limit states) and that the
Who likes concrete? 19
structure should not deteriorate unduly under the action of the environment over the
design life, i.e. will be durable. (Clause 2.2.1).
What, however, is not included, under the serviceability limit state is a state that would
reflect an observers acceptance of, or satisfaction with, the appearance of a building,
whether it be the apparent structural safety of the building or the purely visual aspects.
Part of what would impinge on the casual passer-by, even if only subconsciously, would
be the visible surfaces. And if concrete is a part of the building and is exposed to view,
then, taking account of its particular relation with adjoining surfaces, it must be
satisfactory to the casual observer.
The various surface conditions, each with its own scale of five, would cover at least the
following:
i. crackingfrom invisible at close quarters(or none) to visible at a distance;
ii. tactilefrom very smooth to very rough (terms such as off the shutter, fairface, etc.,
are too vague). This could possibly be subdivided to deal with exposed aggregate
finish, bush hammered finish, board marked finish, etc., as well as applied
cementitious finishes, each with an associated scale of five;
iii. colour variationfrom none to, say, noticeable at a distance of x metres at y months;
iv. jointsfrom specially treated for invisibility or acceptable appearance to no particular
control;
v. honeycombingbest described quantitatively, e.g. 1% to 10% or more of a surface
area. The randomness of its occurrence should deserve attention also;
vi. leachingthis could be broadened to include efflorescence, algal bloom, fungal
growth, moss, lime leaching, and a quantitative scale would possibly be appropriate.
There are difficulties with this since none of these phenomena are obvious on striking
of the formwork and may take years to become evident;
vii. shutter markssimilar approach to joints.
Thus, a Grade III surface finish would consist of a grade 3 in each sub-element in the
subset of finish descriptions. Therefore, rather than having a proliferation of possible
combinations of finishes, there would be only five grades, each grade consisting of one
subset from the list of seven finish requirements.
To operate the system in practice, a designer would need to describe the details of the
seven (or more) finish descriptions and agree them with, or present them to, a client.
Then, a judgement would be made on the importance of the various concrete surfaces (A
to E) and a judgement made on the distances of the surfaces from some agreed
observation point (1 to 5). These two, when plotted on the perception chart of Figure 1,
would give the surface quality grade required. For example, a C-rating for importance
with a 1-rating for distance would require a Finish Type III. Similarly, a building of top
rate importance, i.e. with an A-rating on the importance scale, but only visible in the far
distance, would have grade IV surface finishes, externally anyway.
CONCLUSION
The question Who likes concrete has not really been answered. But if careful attention
is paid to concrete finishes from the point of view of the casual observer as well as for
durability purposes, perhaps along the lines indicated, then maybe people, over time, will
come to accept that concrete is not really so bad after all. Indeed they might even agree
Who likes concrete? 21
than concrete can be beautiful. And some other word will then have to be used to conjure
up the nastiness and distaste that seems to be the present purpose in the public eye of the
word concrete.
REFERENCES
One of the most pressing needs has been to modify the legislation under which the design
and construction of buildings are built. In addition to this has been the crucial process of
ensuring that the performance concept and the need for producing sensible criteria has
been developed and effectively implemented. In particular, the identification and precise
determination of client needs by the designer/constructor, in both qualitative and
quantitative terms, has been a demanding part of the process.
The paper will address the means by which these new concepts have been adapted
and adopted in Romania. It will highlight the problems that have been experienced and
expound upon how they are being resolved in an emerging free market economy.
Keywords: Construction, performance concept, performance criteria, legal liabilities,
cultural differences.
Crina Oltean-Dumbrava is a Romanian citizen currently appointed as a lecturer at the
University of Abertay Dundee, UK. She specialises in decision making on the basis of
multi-criteria analysis.
INTRODUCTION
The design and construction process has as its most important goal to identify the needs
of the client and to reconcile them with the design, technological, economic and
architectural aspects of the project. These aspects have to answer, in a suitable way, to the
ecological and social requirements. This is a difficult problem to solve if the financial
resources of the client and society generally are constrained, which is often the case for
those countries in transition to a market economy.
Perfomance criteria: The pains and strains of change 23
In the past, Romanian civil, structural, building and services engineers developed a
high standard of expertise. In structural design, especially of reinforced, prestressed or
partially prestressed concrete structures and of steel structures, this was particularly the
case. For these design engineers, all their efforts were concentrated in obtaining strength,
stability and durability of the structure by complying with the main building material
(cement, steel, timber) consumption indices fixed by arbitrary regulations ordered by the
communist party. The partys wish was to constantly decrease the indices in order to
increase efficiency in the building process. On occasions, because of the very strict
liabilities, some of the structures, especially in areas affected by earthquakes, were over-
designed in order to protect the designers and builders who were responsible for the
strength and stability of the building for its entire lifetime.
Little or no concern was given to the finishes or to the comfort of the buildings and
although the architects tried to influence the aesthetics of buildings (mainly in blocks of
flats), it was very difficult to achieve this. This was because of the high level of
standardisation of these buildings since the majority were constructed of large
prefabricated reinforced concrete elements.
Specialists in the construction field increasingly became dissatisfied with the way these
blocks of flats looked and particularly with the level of comfort offered to their tenants.
Therefore, these specialists brought to the attention of the authorities the need to improve
the quality of housing, which would also improve the health of the occupants.
It was with the help of the performance concept, which was introduced in the different
regulations concerning the quality of the buildings, that these specialists tried to change
the entire philosophy of the house building process. Before the introduction of the
performance concept, the focus was on the quantity of buildings produced in order to
satisfy the social demands prescribed by the socialist and communist economy. The
effect of this approach was that quality was neglected. Through the introduction of the
performance concept, it became possible to address quality issues and standards were
subsequently able to be improved.
There was and still is a difficult problem to solve in Romania because of the small
range and inadequate supply of building materials available for finishes and because of
the necessity to create a culture for quality. Previously, people typically accepted the flat
offered by the state at a low rent without complaining about its quality or comfort. The
designer and the builder were concerned about the strength and stability aspects of
quality and did not consider the quality of finishes as important. However, this was in
great contrast to the high specification finishes of some residential buildings built for
some communist party members compared to the quality offered to the public generally.
The House of the Republique in Bucharest with its exceptionally high quality is only one
example of the high level of craftsmanship and skill of Romanian builders.
The performance concept is a methodology which enables the establishment of the
necessary quality characteristics of a building to be met. In this way, the user
requirements can be identified without necessarily taking into account the means by
Appropriate concrete technology 24
which they will be achieved. It will, therefore, be appreciated that the introduction of the
performance concept in Romanian building regulations was so important. In addition to
the criteria of strength and stability, fire risks, impermeability and durability, it is
necessary to add criteria which focus on physical and psychological comfort, including
hydrothermal, atmospheric environment, acoustics, artificial lighting, tactile, hygiene,
anthropometries, radiation and sun brightness, electric field and ionic density, security,
privacy and servicing continuity. Each of these help to ensure that building occupants are
provided with good living accommodation. It should be appreciated, too, that some of
these requirements existed either explicitly or implicitly in the former regulations in
Romania, but they could only be requested rather than demanded. Many others have been
identified by the author of this paper as a result of a deep study of human needs and
requirements being undertaken as part of doctoral research. It is not the intention to insist
on the functional architectural requirements which are increasing the quality of buildings.
However, the freedon to create without being forced to comply to the strict rules of
standardisation brought a significant change in project originality with respect to site
integration, general space organisation, flexibility and aesthetics. Further, to create a
culture for quality means much more than a quality flat or house. It has to take into
account the urban integration, the quality of the services offered and the built
environment as a whole. People, too, have to be taught to be more demanding when
considering the environment in which they live and should ask for the provision of
schools, leisure facilities and shopping areas which will improve the quality of their
living environment.
The distance from an individuals house to their place of work is also an issue viewed
differently by people from different countries. The Anglo-Saxon approach generally
considers that to travel up to 25 miles from home to their place of work is reasonable and
therefore presents few problems. In the Latin culture and the culture of some Slavonic
countries in the East or Central Europe, it is essential to have a house as near as possible
to the place of employment. This need is not only a matter of culture, but also a matter of
the economic level of development. In countries with poor infrastructure, poor local
travel systems and where access to and use of a car is considered a luxury, it is difficult to
consider a house if it is a long way from the place of work.
It is clear that the level of these criteria and not simply their quantity is dependent
upon the wealth of the country. For example, the minimum temperature in wintertime
which should be provided in a house in France, as with most European developed
countries where a temperate climate exists, is 24 Celsius. In Romania before 1990, the
figure was 18 Celsius (but this was never respected). Now, this has been raised to 21
Celsius. In 1995, this new figure is still not being achieved. In reality, such requirements
can only be recommendations because on several days per year, these figures cannot
always be achieved even in Western European countries. In Romania, to even achieve
figures of 21 Celsius for the majority of the time is still a major goal.
Another perspective on building is what is considered to be their lifespan. Should the
philosophy be that buildings are considered as a consumption product with a lifespan of
30 years maximum for houses as in America, or should the lifespan be 100 years as was
the case with the blocks of flats in Romania. In the 30 year lifespan example, timber
frames and lightweight materials are used, whereas in the 100 year lifespan example,
reinforced concrete elements provide the framing and cladding system. Further, the 30
Perfomance criteria: The pains and strains of change 25
year lifespan philosophy does not provide a means of transferring the culture and
tradition onto the next generation. One therefore builds for one generation mainly in the
fashion of the day using the latest materials and services. In the 100 year lifespan, a
means of transferring to the next generation a heritage is provided. All the processes from
inception to completion are different and the differences are not only in technology and
philosophy, but also in culture. Comparisons between Northern and Southern Europe are
difficult, not least because of climatic differences. Likewise, it is not sensible to compare
houses built in Japan with those built in Africa. This is not only because of climate but
also, and importantly, in the concept of family life which is radically different. Finally,
the potential for natural phenomena occurring, such as earthquakes, will influence the
design, technology and materials used in buildings.
The improvement of quality by Romanian specialists and the crucial role played by the
performance concept and the performance criteria are major inputs. These provide an
open feedback system which is the quality assurance system (see Figure 1). During the
contract execution, contractors have major responsibilities from the point of view of
quality. Romanian Law No. 10/1995 Law Concerning the Quality in the Construction
Industry [1] sets out the requirements of the building works, the quality system and the
obligations and responsibilities of all the actors taking part in the erection of a building.
In order to achieve a high quality building, the law demands the following requirements,
which shall be met during the lifetime of the building. These are: strength and stability,
safety in operation, fire safety, hygiene, minimum health standards, environmental
recovery and protection, thermal and weather protection, energy saving and noise
protection.
In accordance with this law, the contractor, in summary, has the following primary
responsibilities:
a) to inform the investors on the lack of conformity or agreement in the project with the
aim of solving such problems;
Appropriate concrete technology 26
b) to begin works only with approval in accordance with the present legal requirements
and with projects checked by chartered project experts;
c) to ensure the proper quality level through his/her own quality system, devised and
achieved by his/her own staff, with the execution and technical responsibilities being
given to chartered personnel;
d) to identify the factors involved in the checking of the works in each building phase and
to ensure the requirements are achieved;
e) to give solutions for lacks of conformity, defects and dysfunctions during erection only
on the basis of the designers solutions and with the agreement of the investor;
f) to use only products and procedures provided in the project. These products and
procedures have to be checked to ensure compliance with the quality standards. The
replacement of the products and procedures in the project by others with similar
features is only possible if the designers solutions and investors agreement are
obtained;
g) to carry out all details in order to meet the quality requirements;
h) to inform the State Control Office for Buildings, Public Works, Urban and Land
Planning within 24 hours if technical disturbances appear during erection;
i) to submit for inspection only high quality works meeting the requirements and which
have the necessary technical documents needed to fill in the estate documents;
j) to carry out all measures established by control or checking bodies or by work
reception documents;
k) to repair at the contractors own expense any quality failure either caused during
erection or during the guarantee period;
l) to bring back the temporarily engaged land to the original state when works are
completed;
m) to decide the exact responsibilities of all participants taking part in the process of
production, ie responsible parties, co-workers, suppliers and subcontractors. These
responsibilities will be assigned according to the performers own quality system and
legal provisions.
The legal penalties for non-compliance with the law are very severe. For a loss of
strength and stability of a building which directly leads to loss of life or serious injury or
partial or total destruction of the building or other severe consequence, could lead to
imprisonment of between 5 and 20 years. Where other non-compliance occurs, fines
ranging from 10,000 to 30,000,000 lei (4300 lei=one sterling in December 1995) are
applied.
The penalties described above might be regarded as harsh, but they are necessarily so
to ensure that total quality management systems are applied to the entire building process.
The quality assurance system in Romania is not yet fully in place, but it is essential for
the building to be designed to high standards irrespective of this.
The law also prescribes the obligations of the owner or the tenants of the building.
This is a very sensitive issue because although the owners and tenants are forced by law
to maintain their buildings to a high standard, they often cannot comply with these
requirements because of the lack of financial resources. This is also the case with
stateowned buildings. It is, therefore, necessary to create, as an essential step in the
insurance system, a facility to enable owners to receive low-interest rate loans to enable
Perfomance criteria: The pains and strains of change 27
maintenance work to be undertaken. This will also ensure that the previous investment in
the built environment will be protected.
CONCLUSIONS
This paper describes the difficulties that a country such as Romania in transition to a free
market economy faces when it endeavours to achieve high quality buildings in an effort
to satisfy the needs of users. The change is not only difficult because of economic,
technological and technical constraints, but also because of the necessity to implement a
new legal framework. The creation of social and public services to produce the necessary
infrastructure is essential too. But what is arguably even more difficult is to create in
people a quality culture and for them to realise that the only reason for such systems is
to help satisfy their requirements. Implicit in all the issues on quality is the need to realise
the impact that construction has on the environment and be aware that the concept of
environmental protection and sustainability is of paramount importance. Only when all
these issues are fully understood and implemented will the pains and strains of change
diminish.
REFERENCES
1. LAW No. 10/1995. Law Concerning the Quality in the Construction Industry. Romania, 1995.
APPLICATION OF A STATISTICAL
METHODOLOGY FOR THE
EVALUATION OF THE POTENTIAL
MARKET OF NEW CONCRETE
PRODUCTS ON THE CONSTRUCTION
INDUSTRY
C C Videla
H de T Solminihac
J de D S Ortzar
J P Retamal
Pontificia Universidad Catolica de Chile
Chile
ABSTRACT. This paper presents the results of a case study carried out
for the evaluation of the impact of new concrete products or technologies
on the construction market. The applied methodology consisted on the
planning and implementation of a survey of the Chilean construction
industry and its later statistical analysis. Stated Preference data and the
discrete choice Logit models were used to determine the potential demand
function for the product. The model allows to know the probability of
choosing a new product compared with a traditional one as a function of
certain product attributes. It is worth recalling that these tools are
frequently used in market research and transportation engineering.
INTRODUCTION
The present work forms part of the study Assessment of the Potential Market for
Minimum Cracking Concrete, MCC [1], performed by the authors for a Chilean cement
company. This paper examines a methodology to evaluate the impact of new concrete
products on the construction market and allows to know the probability of choosing a
new product compared with a traditional one as a function of the level of service of
certain product attributes. The procedure was used to quantify the potential users market
of fibre reinforced concrete (FC). The characteristics and the average willingness to pay
of the Chilean concrete market for the new product were studied. The authors express
their acknowledgement for the technical and economical support given by the company
POLPAICO S.A. and to Andrs Iacobelli for his unselfish technical assistance.
METHODOLOGY OF ANALYSIS
increase of impact strength (%). To model the users behaviour the random utility theory
was used [2]; thus assumes that individuals select an option from a finite set of
alternatives based on the relative attractiveness of the option, i.e. they always select that
option which maximises their net personal utility. However, a wide range of variables or
attributes can define the utility that a particular product can offer. Also, not all individuals
consider the same variables for their choices. Therefore, the net utility of a individual is
modelled as the sum of a representative utility (or utility function from now on, which is
a function of attributes that are assumed to be considered by all individuals) and a
random component (which consider that the modeller does not possess complete
information about all elements considered by individuals making a choice, together with
any observational errors). The form of the utility function is linear, e.g. for each
alternative of this particular case it is as follows [2]:
(1)
(2)
where, Utility function of individuals belonging to segment q for concrete with (FC)
and without (TC) fibres, respectively.
Parameters for the ith variable in the utility functions of segment q, for the
FC and TC alternatives. They represent the relative influence of each
attribute in terms of contribution to the overall satisfaction produced by the
alternative (marginal utility).
i: Variable representing and attribute. The attributes i were the following:
Pr=Price (UF/m3), UF is a commercial interchange unit used in Chile and its
value is approximately US$ 32; Cr=Reduction of cracking (%); Ar=Abrasion
resistance (%); IS=Impact strength (%).
c: Calibration or modal constant.
Because the net utility of an individual is a random variable it is only possible to work
with probabilities, i.e. to predict the probability of choosing a certain alternative,
according to the level of service offered by each of the available options. From the utility
functions of each group q and defining certain levels of service for each option, it is
possible to determine the probability of choosing each alternative. Because this is a
binary choice the sum of the probabilities is equal to:
(3)
DISCUSSION OF RESULTS
Market Characteristics
Concerning the market characteristics only the results related to the main problems
encountered in concrete construction and their causes will be deal with, because lack of
space. More information can be found in reference [1]. Figure 1 shows the most frequent
problems encountered in Chilean concrete construction works. An interesting aspect
highlighted on this figure, is that concrete cracking is the most significant problem from
the point of view of both designers and constructors, representing around 40% and 33%
of the answers respectively; thermal cracking appears as the most frequent problem for
constructors (25%) and designers (19%). This result also indicates that the most severe
problems are encountered with cracks occurring during the first month of the concrete
life. Also, results showed that construction problems appears to be by far the major cause
of failure (41%).
The subjective marginal value or willingness to pay of a group for the improved
characteristics offered by FC concrete can be estimated for the three technical variables
considered in the model. However, in the analysis that follows, only the value of those
Application of a statistical methodology for the evaluation of the potential market 33
(4)
The WP is the extra amount of money in US$/m3, that a individual should be willing to
pay for each 1% improvement of a specific characteristic offered by a new product [2].
The most significant variables for the group of designers are clearly price and
percentage of crack reduction [1]. The willingness to pay of designers for FC concrete
offering a 1% reduction of cracking (VMCr) is 0.25 US$/m3/1% crack reduction.
Therefore, if FC were capable of reducing by 50% the amount of crack formed, the
designers would be willing to spend up to US$ 12.5 per cubic meter in FC more than the
amount they would spend if they bought a traditional product [1].
Similarly, the statistically significant variables for the group of constructors are price
and percentage of crack reduction. For this group of professionals WPCr is 0.21
US$/m3/1% crack reduction. It is interesting to note that the designers willingness to pay
for reduction of cracking is 20% higher than that revealed by the constructors [1].
In the case of hydraulic works not only price and percentage of crack reduction are
statistically significant; also abrasion resistance is important. The willingness to pay of
abrasion (WPAr) is 0.45 US$/m3/1% abrasion resistance increase; in this model WPCr is
0.22 US$/m3/1% crack reduction. The results show clearly that individuals working on
hydraulic structures values much more the abrasion resistance than the reduction of
cracking (WPAr is more than 100% higher than WPCr) [1].
For the group of pavement construction, the abrasion resistance and impact strength
variables were not included because they had significance levels far below 1.96. For this
reason the model calibration was performed only for price and percentage of reduction of
cracking. The calculated WPCr value was 0.23 US$/m3/1% crack reduction [1].
(5)
For example, if it is desired to know the proportion of the hydraulic structures market
selecting FC the expressions for V in model 3 must be replaced in the above formula.
Certain values must be assumed for the price of each option as well as for the percentage
of technical improvements of FC concrete compared with TC concrete. The following
limits were used: 30% and 90% crack reduction; 0% and 20% abrasion resistance
increase; 10% and 60% impact strength increase. Table 2 shows an example of the results
obtained when market segmentation procedure was applied [1].
Appropriate concrete technology 34
Table 2 shows that it can be expected that more than 66% of the hydraulic structures
market should use FC if it is defined as having 30% crack reduction, 10% impact strength
increase and equal price and abrasion resistance than a traditional concrete. For the same
FC (equal technical characteristics) but with a price 9.6 US$/m3 higher than a TC
concrete, then the percentage of the market interested in FC will diminished to almost
32%. The remainder part of the table can be analysed in the same way [1].
Table 2 Example of the market segmentation of
hydraulic structures for FC product
Cracks Abrasion Impact Price TC Price FC FC market
reduction resistance strength (UF/m3) (UF/m3) participation (%)
30% 0% 10% 3.0 3.0 66.26
3.3 31.71
3.6 9.89
3.9 2.25
60% 10% 35% 3.0 3.0 93.69
3.3 77.85
3.6 45.38
3.9 16.42
CONCLUSIONS
This paper has shown the potentials of a new methodology for the evaluation of the
impact of new concrete products on the construction market. One of the conclusions
obtained is that Chilean concrete construction market considers as a recurrent problem
the cracking of concrete at early-ages (first month); therefore, a product like fibre
reinforced concrete should a priori occupy an important place in that market. With
respect to the group division it can be concluded that designers have a much greater
willingness to pay for crack reduction of concrete than constructors; this seems logical
because the latter group perceives the effects of higher costs more closely than the
former. An important difference can be appreciated between professionals working on
pavement construction and hydraulic projects; in the first case they mainly value crack
reduction, while the others give priority to improved abrasion resistance. Also, it can be
said that the reduction of cracking is a characteristic valued for all individuals. Finally, it
must be pointed out the very low value of the calibration constant (with no statistical
significance); because this constant captures all those variables not considered in the
modelling, it may be concluded that the chosen variables practically explain all the
selection process.
Application of a statistical methodology for the evaluation of the potential market 35
REFERENCES
INTRODUCTION
There are two aspects to the design of concrete: materials and techniques. The purpose of
design is to ensure that, under given circumstances, the most appropriate materials are
selected and that compatible techniques of concreting are deployed so that the project
specification is achieved.
There are broadly two types of specifications, either based on materials or on
performance. With the former type of specification, materials requirements have already
been researched and are laid down by the specifier, leaving little scope for modifications
by the contractor. The design thus becomes the selection of the right techniques to go
with the chosen materials. With the latter type of specification, the performance of the
end product is specified and the contractor has the freedom in his design to suit the
particular circumstance. This type of specification encourages innovative and economical
design and is gaining popularity. Whichever type of specification is used, however,
someone, somewhere has to translate the project specifications into specific requirements
for concrete technology. Therefore, this paper takes a broad approach to concrete design
and aims to highlight some important, and yet often neglected, considerations in the
design process.
The design of concrete, unlike structural design, is not a well developed process. In the
past, it was often taken as a means of achieving an adequate strength. Over recent
decades, especially with the wide-spread use of reinforced concrete, it has gradually been
accepted that concrete needs durability as much as strength, if not more so. This
transition is justified by the fact that many structures have failed, not due to inadequate
strength of their members, but their inability to stand up to deterioration caused by the
environment. In todays world, concrete design takes on new meaning and puts emphasis
on the practicality, economy and environmental aspects of the design. As a volume
construction material, concrete can be found in every developed and developing country
and, with thoughtful design, concrete should and can meet the varying and sometimes
very stringent requirements.
This paper aims to categorise and discuss some important issues concerning concrete
design. It does not intend to become a design manual, but to provoke further thought by
people involved in concrete design. On the basis of the discussions, criteria are proposed
for the assessment of the appropriateness of a design.
It is considered that there are four major areas which concrete design has to take into
account and they are strength, durability, practicality and economy. There are also some
other factors, such as the environment, and they are discussed briefly under a separate
heading.
While it is possible to design concrete with a range of properties, flexibility in design
is often restricted by codes of practice and Building Regulations. It has been suggested
that codifying aspects such as durability is not possible(1). However, amendments in
codes and regulations are certainly required to incorporate ideas of modifying materials
and techniques in order to achieve the flexibility in concrete design.
Appropriate concrete technology 38
Strength
Concrete is usually classified by its strength, typically at an age of 28 days. Whilst this is
convenient and does provide a base for comparison, it may not be the most appropriate
under certain situations. For the design of strength one needs to know which of the
following criteria to satisfy:
strength at stripping shutters
strength at different times of construction
strength at application of full service load
The problem of accepting different strengths at different times is one of current code
requirements. While early strength gain is often checked by testing cubes or cylinders at
1 and 3 days to determine whether formwork can be stripped or removed (as in
slipforming), the full service load in many situations is only applied a substantial time
after construction commences, hence a design strength at 90 or 180 days or sometimes
even longer may be appropriate.
The significance of this change in specification is that it would increase the choice of
cementitious materials and allows a more effective use of blended cements. It is known
that, at the same water-cement ratio, the strength development of blended cement is
slower than its ordinary Portland cement (OPC) counterpart, but its ultimate strength may
be considerably higher(2,3), as illustrated schematically in Figure 1.
especially in the longer term, the ingress of chloride ions(6), which is often associated
with the premature deterioration of marine structures.
iii) By allowing the use of blended cements, it helps to produce economic concrete
mixes. For instance, if the 28 day strength requirement is replaced by that at 90 days, pfa
may be used in cast-in-situ concrete piles more effectively. The economic benefit is that,
for a typical pile of 5m in length and 0.30.4 m in diameter, the materials cost would be
reduced by about 1, or 5 to 7%,. For a medium size piling company with an annual pile
production of 300,000 linear meters, this would be translated into a saving of 60,000 per
year, not a small sum for such little effort.
Durability
Many structures have failed prematurely, not because of inadequate strength, but because
of inadequate durability. The cost of the consequent repair work takes up a significant
part of the construction expenditure and this is particularly so in developed countries. For
instance, it is estimated that in the UK alone, expenditure on repair/maintenance-related
problems was in 1993 in the region of 20 billion, or 42% of the total construction
output(7). So, what shall we do?
Reinforcement
Steel reinforcement is often taken for granted in our structural design. However, do we
really need steel in every situation? The fact that some Roman concrete structures have
lasted for thousands of years suggests that concrete structures would be better without
reinforcement as far as durability goes. The single most important cause of deterioration
of concrete structures is due to the corrosion of steel reinforcement(8). In some parts of the
Appropriate concrete technology 40
structures, e.g. mass concrete foundations, the use of steel reinforcement can be
completely avoided. Despite much successful experience, many contractors, clients and
statutory authorities feel uneasy about this option and further education may be required
to convince people that, where only compressive loads are applied, reinforcement is not
needed.
In most instances, the use of reinforcement is probably justified. However, if the
environment is aggressive, we can consider increasing the level of protection. For a start,
the depth of concrete cover to the steel can be increased. There are special types of
reinforcement which may be used, such as epoxy-coated rebars, non-ferrous fibres or
reinforcement. If the environment is really aggressive, consideration can even be given to
the installation on the structure of a cathodic prevention or cathodic protection system as
a further preventive measure.
As a first approximation, it may be assumed that the concrete cover depth follows a
Normal distribution, as indicated in Figure 2 (For lower nominal cover the distribution
curve may become positively skewed). So there is a probability, however remote, that a
result will be obtained which is less than the specified value. It is therefore sensible not to
specify a minimum cover depth but a characteristic cover depth, below which a
specified proportion of the test results, or the defect, may be expected to fall. The
characteristic cover depth may be defined to have any proportion of defects. BS 8110
adopts 5% defective level for strength. However, it provides a further safeguard against
failure by providing a relatively high partial material factor of 1.5. Given the
consequences of low cover and lack of other protective measures, a defect level of 2.5%
is considered as appropriate.
As a result of the variability of site practice it is necessary to design concrete elements
to have a mean cover depth greater than the specified characteristic cover depth by an
amount, i.e. the margin. Thus:
dm=dc+ks
Hence, for a specified characteristic cover depth of 30mm and 97.5% probability of
exceeding this value, the target mean cover depth can be worked out as below:
dm=30+1.96 (0.230)=42 mm
Clearly this concept needs further investigation, particularly with respect to the standard
deviation of cover. Consideration also needs to be given to the type and nature of
construction, the level of supervision and applied quality control.
Surface protection
Appropriateness in concrete design means the recognition in a positive manner of the
merits and limitations of the materials. Protection to concrete structures under some
circumstances is not only desirable but essential. The recently completed Shantou
Harbour Bridge in southern China is perhaps the longest suspension bridge using a
concrete deck, with a centre span of 450 metres. Figure 4 shows the bridge during
construction. For such a long bridge deck, self weight was a major problem and in order
to minimise it, a combination of high performance concrete and a small cross section had
to be employed. With the small cross section, the effective cover depth at the soffit of the
deck was only 18mm and, since the bridge is exposed to a semimarine environment, its
durability was a serious concern. To compensate for the low cover depth and to help
achieve the design life of 120 years, it was decided by the engineers involved in the
design and construction of the bridge to provide a surface coating to the structure. In this
particular case, a high performance, composite epoxy coating system was selected, which
provides excellent resistance to ingress of both carbon dioxide and chloride ions. It has a
design life in excess of 10 years, after which recoating may become necessary.
Appropriate concrete design 43
Practicality
Experienced engineers will know that it is no better to have a theoretically perfect, but,
impracticable, design than to have a bad one. It is therefore important that, in carrying out
the design, the practicality is carefully considered. This includes the availability of
materials, their quality and consistency, the availability of a trained workforce, quality
control and the construction conditions.
At present the use of high strength high performance concrete (HPC) is gaining
popularity in Hong Kong. Grade 100 concrete mix was used, for the first time, on a
commercial project last year(11). Along with its use, there were many problems which had
to be resolved. For instance, while most concrete constituents were obtained locally,
silica fume, which was required for the mix, had to be imported. This meant that, before
the use of HPC, appropriate and sufficient storage of silica fume at the site of the ready
mix supplier had to be resolved.
Quality procedures have to be tightened to cope with the smaller tolerance of the
concrete mix. The accuracy of batching and weighing machines has to be frequently
checked and maintained, as small changes in water or cement measurement will result in
a relatively large variation in water-cement ratio and, hence, the strength. For instance,
within the normal accuracy of a balance, a variation in weighing cement or water of
2kg may result in a change in strength of 36N/mm2 for a grade 100 mix compared with
only 13 N/mm for a grade 40 mix. As the total water content of the mix is less, any loss
in evaporation will have a detrimental effect on the properties of the concrete, especially
that near the surface. This requires that special care is taken in the curing of the concrete
and water loss from concrete is minimised.
Appropriate concrete technology 44
With the use of HPC, the concrete mix tends to have a higher cementitious content.
The amount of heat generated during cement hydration and, hence, the temperature rise
of the concrete can be very significant. For a typical Grade 100 mix and a concrete
section size of 2 meters, if no special measures are taken, concrete temperature can rise
by as much as 70C during the first 24 hours after casting. With such a steep rise in
temperature, thermal cracks can develop and the high temperature could also lead to the
development of a deleterious mechanism termed delayed ettringite formation, which in
the long term leads to cracking of the structure(12). Both these factors could have a
detrimental effect on the durability of the concrete structure and, therefore, require
careful consideration at an early stage for their avoidance.
Economy
In the search for a technical solution, an engineer has a duty to ensure that his design
meets the requirement of the client and is economic. He will have to consider the cost of
the materials, plant, labour and also the quality of the end product. One thing worth
bearing in mind is that an economic solution does not necessarily imply a cheaper initial
installation. The whole life costing of the structure also has to be considered. This enables
the construction and operation costs over the life of a structure to be expressed in present
values and different strategies to be compared on a common basis. Substantial savings
may be achieved by ensuring durability at the design and construction stages.
It has been suggested that there is a law of five(13) which states that one dollar spent
in Phase A(see Figure 5) equals five dollars spent in Phase B equals twenty-five dollars
spent in Phase C and equals a hundred and twenty-five dollars spent in Phase D. What
this means is that a little extra attention to durability during design and construction (say
one dollar) can lead to substantial gain with respect to renovation cost (of say 125
dollars).
Again taking the Shantou Harbour Bridge as an example. The decision to protect the
bridge deck with a high performance epoxy coating system itself is based on the
consideration of minimising the whole life costing of the bridge. Having selected the
main component of the coating as being epoxy, a further question is the selection of the
top coat. There were a number of choices, including chlorinated rubber, acrylic and
polyurethane. From weathering tests conducted on these specific products, it was found
that chlorinated rubber offered the least resistance to weathering and the polyurethane top
coat was some 50% more durable than the acrylic one. As the bridge is to be operated as
a toll bridge, it was obviously in the projects interest to maximise the maintenance
period, which could further reduce the associated costs in materials, labour, access
provision and, more importantly, lane possessions of the motorway. Therefore, although
the cost of polyurethane top coat was almost 100% more expensive than its acrylic
counterpart, it became the most favoured candidate for application on the bridge as the
top coat.
Appropriate concrete design 45
Other Considerations
Engineers have a responsibility to the society in which they live. In this respect, they
have to consider other aspects of the concrete technology, e.g. the law and regulations,
health and safety in the use of the materials, environmental protection etc. Of particular
relevance may be the protection of the environment. When using concrete technology,
engineers should consider how the use of waste materials from other industries will
benefit not only concrete technology but also the environment. The use of pfa and ground
granulated blastfurnace slag will reduce the need for costly disposal of these materials
and will also reduce the resources and energy needed to produce cementitious material.
These materials, when used judicially, have the added benefit of improving concrete
durability.
In order to maximise the benefits of employing appropriate concrete technology,
efforts should be made to reduce regulation and over-specification of concrete to permit
innovation.
Based on the above discussions, the authors would like to propose the following four
main criteria for assessing the appropriateness in concrete design.
First and foremost, a concrete design should be assessed on the basis of fitness for
purpose. This includes the engineering properties as well as the durability aspects of the
materials, taking into consideration what is desirable and what is actually achievable in
practice.
Appropriate concrete technology 46
ACKNOWLEDGEMENT
The authors would like to thank the Shantou Bay Bridge Company, Shantou, China, for
permission to publish some of the test results.
REFERENCES
INTRODUCTION
This paper describes the concepts of a Durability Audit currently under development at
BRE in order to enable the construction industry to achieve this aim. The paper identifies
the different stages of the construction process and how they affect durability, previews
how the audit might operate and predicts its likely impact on the durability of future
concrete construction.
Premature deterioration of concrete and reinforced concrete in structures has been a
matter of serious concern for several decades and remains so at present. Although some
of the recent changes in the use of materials may have a limited impact on the general
level of durability, little of fundamental significance would seem to have been applied
across the construction industry that is likely to cause a radical improvement in
durability. For example, a recent as yet unpublished study of the achievement of cover to
reinforcement in 25 structures, under construction in the UK during 1993/94, showed that
failure to achieve specified cover is still commonplace. Inadequate cover is, of course, a
major contributory factor in the deterioration of concrete through reinforcement
corrosion. It is thus likely that a significant proportion of concrete structures currently
being built will lack durability in the future, thus justifying the introduction of radical
measures aimed at tackling the whole problem.
It will be a formal system in which the decisions and actions affecting durability will be
examined at all stages of the construction process. The principle stages of the
construction process and how they affect durability are considered in the next section. If
adopted, the audit would most likely be required by the client for the structure; it would
comprise a scrutiny of appropriate evidence of how durability has been taken into
account. This evidence would be provided by the party responsible for the stage of
construction under consideration.
Assuring performance of concrete structures through a durability audit 49
To ensure the effectiveness of the Durability Audit, it will cover all the decisions and
activities that have an influence on the durability performance of the structure, at all
stages of the construction process. The main stages are shown in figure 1.
Some of the principle questions that will be addressed by the audit at each stage, are
given below:
Assuring performance of concrete structures through a durability audit 51
Briefing
Have the clients requirements for durability been established and clearly stated in terms
of required service life and acceptable maintenance levels? Has an adequate construction
period been allowed with regard both to duration and to likely weather conditions?
Pre-Execution Planning
Has a sufficient time been allowed to ensure all operations can be performed to the
standard required, thus ensuring the structure can be built as designed? Has consideration
been given to ensuring that forseeable weather conditions do not reduce the quality of the
construction?
Execution
Has the structure been built as it was designed, with a sufficient level of quality to ensure
it will be able to last for its design life? In particular, has the concrete been adequately
compacted, finished and cured and is the cover to reinforcement within specification?
Use
Are the planned inspections and maintenance being carried out at the required intervals
and to the required standard? Is the structure performing as designed?
Modernisation
Has the effect on the durability performance of the structure of any change of use been
assessed in relation to the required service life?
AUDITABLE ITEMS
The contents of the clients brief can be assessed against a list of essential requirements
based on BS7543 [5] and/or the draft checklist produced by the UK Construction
Industry Board Working Group 1 in conjunction with Reading University. The brief,
particularly the statement of required service life, will then form a principal standard
against which subsequent decisions and actions are audited.
The design can be audited against the clients brief and the principles of the draft
European Standard for the basis for design [6]. Auditing of the detailed durability design
will depend on the basis upon which the design was produced. The flexibility of the audit
system will ensure that no constraint is placed upon the designer to choose a particular
durability strategy or durability design method. For example, design could be based on a
particular deterioration model provided it can be shown that the particular model is
appropriate for the exposure conditions to be resisted and that the validity of the model
can be demonstrated. This, of course, may cause some problems in the early days of
operation of the system but, hopefully, no designer would choose to use a particular
model unless its validity can be readily demonstrated.
Delivery tickets for materials and products can be audited against design
specifications.
Execution and execution records (inspection records, delivery tickets for materials,
etc) can be audited against the designers drawings and specifications.
Records of maintenance can be audited against the designers maintenance plan.
DEMONSTRATION OF COMPLIANCE
The audit will require demonstration of compliance of decisions and actions with
appropriate and acceptable standards, some of which have been suggested in the
Assuring performance of concrete structures through a durability audit 53
British Standard BS5502 : Part 22:1993 [9] for the design of buildings and structures for
agriculture is a notable exception in that it gives values of concrete quality (minimum
cement content and maximum water/cement ratio) and cover to reinforcement for stated
durability lives of 10 and 50 years for concrete containing reinforcement, and 20 and 50
years for unreinforced concrete.
It is thus clear that improved guidance is required from National Standards and Codes
of Practice on expected service lives from different concretes in different environments.
Such change is, however, likely to be slow to develop and may require extensive review
of existing information on performance in service. In some cases it may be necessary to
commission new research.
Successful application of the Durability Audit will depend, for some parts of the
construction process, on scrutiny of Quality Assurance records so that the auditor does
not have to, say, perform a covermeter survey of the structure to check compliance with
the specification.
Successful demonstration of compliance in the audit will result in the issue of an
unqualified audit report. This report will be a statement that, in the opinion of the
auditors, the records of decisions and actions related to durability demonstrate an
acceptable probability of achievement of the (clients) requirements for the structure.
A qualified audit report would be issued where full compliance cannot be
demonstrated. A qualified report must command the status of serious concern to the client
for the audit. Nevertheless, only something important will cause auditors to quality their
report. The qualification might be due, for example, to a lack of proper records or
disagreement on acceptability of any standard against which a durability related
decision or action has been made.
Appropriate concrete technology 54
Further development of the system is being enabled by application of some of its main
principles to the construction of a new office and lecture theatre complex on BREs
Garston site. This building comprises a reinforced concrete frame and precast vaulted
concrete floors with an in-situ concrete topping. It also features an in-situ corridor slab of
100mm thickness and a 250mm thickness in-situ slab as part of the floors in seminar
areas. Foundations are a ground supported pad in class 2 sulfate soil. Much of the in-situ
concrete will contain recycled aggregates.
Much of this building is being constructed from materials other than concrete, such as
the stainless steel roof, the brick faade, timber window frames and some timber wall-
cladding, but the audit trial will initially be limited to the concrete components.
OTHER CONSIDERATIONS
It is important to minimise the extra workload resulting from the audit; although, even in
the short term, audits should reduce the overall workload simply by clarifying the
requirements at each stage of construction. As far as possible, therefore, it is desirable to
integrate the Durability Audit with other initiatives such as energy and environmental
assessment, CDM (Construction, Design and Management) regulations for safety, and
life-cycle costing to reduce the burden on designers and constructors, etc. For example,
there is a considerable amount of overlap between the information required by the
Durability Audit and that contained in the H&S file required by the CDM regulations.
Application of the Durability Audit will initially be on a voluntary basis, probably at
the request of the client. Nevertheless, for some types of structure difficulty might be
experienced in encouraging the use of the system, for example, where the client is a
developer whose financial interest will cease once the structure, or full-repairing lease,
has been sold. The party who, in that case, will bear the financial liability for inadequate
durability will have had no involvement in the briefing, design or execution stages, but
would clearly stand to benefit from ownership of a structure where the audit system has
been used.
Development of the system will include the necessary gathering of information on many
aspects of the construction process. This will include the procurement of data on the
normal variation of properties of concrete in structures. For example, cover to
reinforcement for in-situ concrete in the UK is commonly specified, in accordance with
BS8110 [7], as a nominal value with a negative tolerance of 5mm to allow for the
variability of fixing. Many studies have been made of cover achieved in practice [10] and
it has generally been shown that the maximum deviation from the specified nominal
cover is commonly significantly greater than the tolerance of 5mm. A durability design
that relies on achievement of a minimum cover of the nominal cover less a 5mm
Assuring performance of concrete structures through a durability audit 55
tolerance is therefore flawed; this problem will be identified within the Durability Audit.
If, however, it can be demonstrated that special measures will ensure that a specified
5mm tolerance will be achieved in practice, the audit will accept this.
There are few, if any, structures where concerns about durability can be limited to a
single material such as concrete. It is intended that the Durability Audit will, in due
course, be extended to cover all materials and components within a structure.
An audit administration body could be established, the functions of which would
include advising clients on the appointment of auditors and deciding on the acceptability
of audit standards such as guidance documents, design models and codes of practice,
and the eligibility of evidence such as Quality Assurance records. In many ways the
concept of the audit parallels the BREEAM methodology, run by BRE for the
environmental assessment of buildings, and BRE could provide the ownership function
whilst industry undertook the auditing.
The primary benefits of the Durability Audit will be the reduced incidence of repair,
reduced lifetime costs and reduced disruption of use due to unforseen maintenance or
repair. A successful outcome, i.e. an unqualified audit report, will provide clients with
greater assurance that the structures they have commissioned will perform as required,
and without significant unplanned expenditure, during their required service lives.
Acceptance of the audit system could provide a valuable guide for insurers in
assessment of their risk. A favourable audit report should allow the insurer to offer
significant discounts in reflection of the reduced risk.
The need for definitive guidance on design life and durability performance will
provide the incentive for more extensive continuous review of standards and codes of
practice than is currently the case.
Development of this Durability Audit will identity areas where essential work is
needed to provide guidance on durability design or specification for a given design life,
against which an audit can be performed.
CONCLUDING REMARKS
This paper has described a proposed Durability Audit and how it can play a major role in
assuring the durability performance of future concrete construction. Indeed, the principles
involved extend well beyond just concrete construction and, in due course, the audit
system can be extended to include all construction materials and construction types.
Introduction of the Durability Audit will encourage the client, designer and
constructor to take a much harder look at building durability into their structure than may
commonly be the case at present. It is acknowledged that within many prestige projects,
provision of durability is already receiving extensive consideration. This is to be
commended but, hopefully, the Durability Audit will encourage such detailed
consideration to be extended to normal structures. BRE is ready to play its part in
developing the scheme together with industry.
Appropriate concrete technology 56
ACKNOWLEDGEMENTS
The authors are indebted to the members of the Department of the Environments
Durability by Intent Strategy Steering Committee and colleagues within BRE for helpful
guidance, suggestions and encouragement. In particular the authors would like to express
their thanks to Kathryn Bourke of Construction Audit Limited for discussions on the
HAPM technical audit system.
REFERENCES
Assistant Professor Franc Saje is Head of the Structural and Traffic Laboratory at the
Department of Civil Engineering, University of Ljubljana. He lectures on concrete,
masonry and timber structures. His research activities mainly concern the design, safety
and reliability of concrete, masonry and timber structures, mechanical properties of
concrete at normal and elevated temperatures, nonlinear analysis of reinforced concrete
structures, as well as fire resistance of concrete and timber structures. He is Chairman of
the Slovenian Group of IABSE and national delegate to the permanent committee of
IABSE and to CEB.
INTRODUCTION
Increased demands for shortening the constructional time often have adverse effects on
the quality of structures. Strict demands of investors can frequently be satisfied only by
using special building technologies which enable faster finishing of construction and
certain rationalizations, but also demand special technical treatment. With certain
technologies or construction sequences (e.g. free cantilever construction of bridges) also
the assurance of adequate final geometry of structure can present a special problem.
Almost all structures, apart from the simplest ones, are constructed gradually. Besides
material and geometrical nonlinearities, usually included in a computational analysis,
their gradual construction is another important contribution to nonlinear behaviour of
structures. In that case a part of the structure is loaded and deformed even before the
completion of the whole structure whereas the structural system can be changed
basically. In reinforced or prestressed concrete structures additional handling is necessary
due to rheology of material. Owing to the fast progress of works there is often a desire to
prestress and de-shutter the structures soon after the concreting. During prestressing the
stress level can be higher than later, during the normal use of the structure. Furthermore,
the concrete is still young in this period, which increases the initial and time-dependent
strains due to rheology. Rheology causes in concrete structures time changes of
displacements, whereas in statically indeterminate structures it also causes redistribution
of internal forces. Stress-independent rheological phenomena are aging and shrinkage of
concrete, stress-dependent ones are creep of concrete and relaxation of steel. Within the
range of working stress creep of concrete depends linearly on the stress, whereas at
higher stress levels this dependence becomes explicitly non-linear.
There are relatively few theoretical papers which deal with the problem of rheology in
conjunction with the variable structural system, and there are even fewer presentations of
appropriate experimental researches. The problem is in most cases solved by generalized
method of forces where creep of concrete is only considered according to linear theory.
Such an approach naturally brings up to the system of Volterra integral equations for
unknown forces which can be solved only numerically in general. The complexity of the
system of equations depends on the type of structures which can be divided into
homogenous and heterogeneous ones [1], with respect to static determination, different
ages of individual parts, existence of reinforcement and possible elastic supports. For the
realistic prediction of time response each reinforced concrete structure must be treated as
a heterogeneous one. In this case, however, the systems of integral equations become so
Realistic prediction of time-dependent response of concrete structures 59
complicated that these methods are actually unsuitable for practical use in civil
engineering.
The method of finite elements (FEM) has been applied in the presented method of
analysis of time response of concrete structures considering together the rheology,
variable structural system, cracks and geometrical non-linearity. For modelling of
viscoelastic properties of aging concrete the generalized Kelvin model has been used
which can in addition to experimentally acquired functions of concrete creep approximate
also creep functions of concrete according to the CEB-FIP 1990 model code [2]. Non-
linearity of creep was included by functions which in dependence on the momentary
stress level increase the speed of creep [3,4]. For time integration, the method with
discretisation of inelastic strain, considering the linear interpolation of stress within the
time interval, as presented by Baant [5], has partly been modified. As a basis for
software development the NONFRAN computer program [6] has been used which is
intended for general static analysis of planar frames considering material and geometrical
non-linearity. The elaborated software allows consideration of gradual changing of
structural system through constructing phases. Graduallity of constructing can be
simulated on the level of whole elements or on the level of cross-sections consisting of
subsections.
(1)
(3)
Appropriate concrete technology 60
Indices r-1/2 describe the values of quantities in characteristic time tr1/2 of interval,
E(tr1/2, r-1, ) is stress dependent secant elastic module in the time tr1/2, i is parameter
of i-th Kelvin unit, n is the total number of Kelvin units, o is a stress-independent
increment of strain due to shrinkage of concrete and temperature influences, f() is
function which represents the influence of nonlinear creep, is the hidden variable of i-
th Kelvin unit (time derivation of deformation) being for the time tr defined by the
equation (2). i is the retardation time of i-th Kelvin unit.
As the function f(), representing the influence of nonlinear creep in dependence on
stress level the formulation (4) from the CEB-FIP MC 90 [2] can be
used, being applicable to the stress level or the suggestion (5) or (6) by
Baant, Prassanan and Kim [3,4], which can also be considered as a rough estimation in
the area of higher stress, fc(t) is the average compressive strength of concrete in time t.
(6)
When concrete is subjected to sustained high compressive stresses the strength decreases
with time of the load. In the computational analysis of structures reduced compressive
strength of concrete is also considered due to previous sustained loads. Additionally, a
special algorithm describes the behaviour of concrete during cyclic loading in time after
appearance of cracks, when the concrete is capable to take over only compressive stress.
In the analysis the strain approach is used, where the change of stress is determined
for defined increment of deformation . Equation 1 is written as a function of the
change of stress in the form g()=0 and solved iteratively by improved secant method.
If the formulation of creep from the CEB-FIP MC 90 [2] is taken as the basis, the
parameters of Kelvin units can be defined as follows. The function which describes the
development of creep c(t-t0, H) (eq. 2.170 [2]) is developped into Dirichlet series
. Retardation times i are chosen in advance [3,5], coefficients
Ai(H) are determined according to the least square method. By means of this method
polynomes, of relatively low degree, can be prepared in advance which allow to
approximate accurately enough the coefficients Ai(H) in the range of expected values H.
The next step is to define parameters of Kelvin units according to equation (7). Quantities
(fcm) and Eci are determined by equations (2.166), (2.167) and (2.116) in CEB-
FIP MC 90 [2].
(7)
The main advantage of the chosen incremental constitutive law of concrete is that only a
stress, a situation indicator (possible tension or compression failure) and hidden variables
of Kelvin units in the previous time have to be saved for each integrating point. This
enables an analysis of large structures at almost unlimited number of time steps also on a
PC.
Realistic prediction of time-dependent response of concrete structures 61
(8)
COMPUTATIONAL EXAMPLES
As computational examples two benchmark tests for evaluation of creep and shrinkage
analysis computer programs, proposed By Subcommittee 3 of RILEM TC 114 [8], are
analyzed. They concern relatively simple structural problems for which usual engineering
assumptions may be done. More comprehensive data relating to the analyzed
experimental tests can be found in [8]. The tensile strength, shrinkage and creep functions
of concrete, used in the analysis, are derived from its compressive strength according to
CEB-FIP MC 90 recommendations, without fitting in any way of the model data for the
optimization of calculated results.
Benchmark no. 1 consists of seven tests of simply supported reinforced concrete slabs
with a span of 310cm. The geometrical description, element model, cross-section and lay-
out of reinforcement in the typical cross-section are shown in Figure 1. At the age of
t=28 days the slabs were loaded by up to five different load levels between P=2.89kN
and P=15.73kN. Thereafter the deflection was measured for a time span of up to 510
days. During this time the relative humidity was 60% and the temperature 20C. Mean
compressive strength at 28 days was 30.6 MPa. In the analysis the compressive strengths
of individual samples were used. Figure 2 presents a comparison of measured and
calculated deflections of slabs.
Benchmark no. 2 comprises short- and long-time tests of three slender reinforced
concrete columns with rectangular cross-section, built in at one end and free at the other
with L=225cm. These tests provide an illustration of the creep buckling phenomenon.
The geometrical description, element model, cross-section and lay-out of reinforcement
in the typical cross-section are shown in Figure 3. An axial load P is applied at the free
end of the columns with an eccentricity e=1,5cm. A load P=280kN was applied to the
column II-2 at t=28 days and sustained for 197 days when failure occurred. Calculations
are made for different functions f() (Equations 4, 5 and 6), which represent the influence
of nonlinear creep in dependence on stress level. Figure 4 presents a comparison of
measured and calculated deflection values of column I-2.
CONCLUSION
REFERENCES
Leader Paper
Implications of Harmonisation
Mr R S Narayanan
SB Teitz & Partners, Consulting Engineers
United Kingdom
IMPLICATION OF HARMONISATION
R S Narayanan
SB Tietz & Partners, Consulting Engineers
UK
INTRODUCTION
Mention the word harmonisation. The reaction of many in the construction industry in
the UK will be anything but harmonious. It generally brings out the euro sceptics in us.
The common perception is that Brussels is imposing new methods of working, obliging
us all to use new codes and standards, new methods of production design and testing.
There is a natural reluctance to change but it is generally accepted that an expensive
changeover is upon us.
Therefore, for the purposes of this paper, I take the title to mean European
Harmonisation, which is slowly but surely creeping on us. It will affect our lives in many
spheres including the field of construction. In the area of Building and Civil Engineering
alone, there now exist 92 dual Euro/British Standards. The number of British Standards
which had been withdrawn as at summer 1995 stands at 24. According to BSI, 1996 is
likely to see a major acceleration in the birth of new Euro Standards.
In this sceptical climate, it well behoves us to examine some of the details of what is
being proposed and the positive aspects of the impending change.
Appropriate concrete technology 68
SINGLE MARKET
The Single European Act came into force in 1987. Under the provisions of the Act, a
huge market of some 370 million people was created on 1st January 1993 across Europe.
It has no internal borders. Goods, services, people and capital should move freely
throughout this internal market.
Some 300 individual measures had to be adopted by the Commission to facilitate this
arrangement. They are aimed at:
a. The removal of physical barriers to trade;
b. The removal of technical barriers to trade through the Communitys new approach to
technical harmonisation and standards policy and the mutual recognition of testing and
certification arrangements;
c. Liberalising public procurement;
d. Free movement of labour and professions;
e. Liberalising financial services;
f. Liberalising transport services;
g. Freedom of capital movement;
h. Company Law
i. Intellectual and industrial property; and
k. Removal of fiscal barriers.
Not all the measures proposed are of equal significance for the construction industry.
Those which most directly affect the industry include the Construction Products Directive
to remove barriers to trade in construction products.
C E MARKING
ATTESTATION OF CONFORMITY
Attestation of conformity is a legal statement that the product conforms to the relevant
harmonised specification, coupled with a declaration by the manufacturer that he
maintains an appropriate factory production control. The procedures, which are defined
in the Directive, range from a manufacturers declaration to third party testing and/or
certification. The Certificate of Conformity will be issued by a designated body in each
member state. Notification of a testing or certification body in one state creates defacto
recognition by all other member states, thus avoiding duplication and eliminating any
need for mutual recognition.
Appropriate concrete technology 70
Public procurement is a key element of the Single Market, both on account of the scale of
expenditure and because of the perceived susceptibility to political or other pressures to
favour a domestic supplier or contractor. Also the contract letting process has been made
more visible. Public sector specifiers are required to refer to European Standards or
European Technical Approvals where appropriate. Thus the product specification
developed to support the Construction Products Directive will effectively be mandatory
under the Public Procurement Procedures.
* Products are incorporated into structures. If products are designed to national standards,
then the whole structure needs to be designed to the same standards. This is not
harmonisation but preserving the status quo and entirely against the spirit of the single
market.
* In the UK, BSI and the industry, are unlikely to have sufficient resources to maintain
Eurocodes and British Codes. Such limited resources as are available, will be better
channelled into achieving satisfactory Eurocodes, which are made sufficiently flexible
that they are able to accommodate well established methods.
Another sector which will benefit from the harmonised Eurocodes, is the software and
publishing industries. This arises from the size of the potential market for their products
(eg) design aids. This will only be possible if all States work on the same design code.
Regulatory authorities could be confronted in the future, with unfamiliar or an unusual
designs and details, on which they will need to pronounce. This is because engineers and
manufacturers will come up with different solutions even when working to the same
Code. The solutions will be largely based on the technical culture in which they are
rooted.
Summarising, the question is no longer whether there should be harmonisation.
Politicians have decided that it should take place. The current debate is all about details.
It therefore, seems sensible to play an active part and influence the debate. The evolution
is unlikely to be smooth or cheap; but it seems inevitable.
REFERENCES
INTRODUCTION
Concrete construction has in all European countries a long tradition. After about fifty
years of harmonization on a voluntary basis individual national design rules lead to
results which differ more or less significantly. This shows the long breath we have to
have to finish the process of European harmonization, which is in itself not very amazing:
Application of European concrete standards 73
also when national Codes are to be replaced by more sophisticated rules, as has happened
in The Netherlands in 1993, the proces of adaption requires years. To meet the
implementation problems, CEN has introduced a gradual approach by the use of
prestandards, Principles, application rules and boxed values. The items which are now
most discussed with regards to Eurocode 2 part 1, are its clearness and usefriendlyness,
the reliability level and economical consequences in relation to national practice.
As the conversion from ENV to EN has been agreed lately, these problems should be
solved before 1998.
DESIGN AIDS
The use of design aids in the design proces relating to Codes are common use in different
European countries. Generally, the process is identified with Code Checking. Codes are
presented in such a way that the design process is to be passed through in an easy way.
Codes are not educational books: they are structured in a way to really design a structure,
which means finding the right shape, size and material specifications for a structural
solution. Although, during this preliminary design, anticipation on final code checking
takes place.
Same can be said with reference to educational books: the are not ideal to perform a
design job. The design process needs its own tools, and in each stage different tools are
used, as shown in the table.
Tools freq. used Stage of the design process
Reports (spec, sit.) Literature on 1. Searching for shape (establishment of prismatic/haunched
spec. subjects Graphs and tables beams, mushroom /flat slabs)
Literature on costs 2. Sizing of given shape
Product information Computer 3. Re-considering shape and sizes (considering possible mix
progr. for optimization design, fully/partially reinforced concrete, alternatives etc.)
Graphs and tables 4. Specifying reinforcement
Computer progr. 5. Code checking
Examples, tables 6. Detailing
This was shown by the inquiry in the design processes of different countries, performed
within the EU-SPRINT programme by the Concrete Societies of Netherlands, Germany
and UK. The use of present-day and future design aids was subject of a questionnaire. It
showed a clear preference in all countries for conventional design aids as graphs and
tables, with handbooks and calculation examples. The bureaus and engineers which were
asked to fill in the questionnaire is not a mathemathical random sample: to reduce costs
and work a selection of representative companies was made to achive representative
figures. The results of the questionnaires are summarized:
Appropriate concrete technology 74
1. Num. Examples
2. Charts and tables
3. Flow charts
4. Manual/Handbook
5. Computer programms
Some highlights from the questionnaire in The Netherlands (1994):
Most of the leading engineers (74%) are informed roughly on EC2, so is their staff
(58%). More than one half of the participants in one way or another has already
studied on EC2 items, but 74% has never made any calculations.
81% has no company-initiated information flow on EC2 in their organization structure.
The software used for different calculations is not prepared for EC2, according to 85%
Estimating the costs for implementation of EC2, most engineers are concerned and
think the costs overall will be reasonable/ to be expected (50%) or very high (47%).
Especially computer programs and CAD systems are mentioned.
Engineers are not unanimous in the judgement on possibilities concerning apllication of
EC2:61 % sees no advantage in the use of EC2 with reference to freedom of design,
against 32% who does see advantages.
Examples, courses and design charts and tables are important for the use of EC2
Germany:
Most of the leading engineers (74%) are informed roughly on EC2, so is their staff
(58%). More than one half of the participants in one way or another has already
studied on EC2 items, but 74% has never made any calculations.
81% has no company-initiated information flow on EC2 in their organization structure.
The software used for different calculations is not prepared for EC2, according to 85%
Estimating the costs for implementation of EC2, most engineers are concerned and
think the costs overall will be reasonable/ to be expected (50%) or very high (47%).
Especially computer programs and CAD systems are mentioned.
Engineers are not unanimous in the judgement on possibilities concerning apllication of
EC2:61 % sees no advantage in the use of EC2 with reference to freedom of design,
against 32% who does see advantages.
Examples, courses and design charts and tables are important for the use of EC2
United Kingdom:
75% of the participants have the opinion a Manual/handbook (with examples) is
essential, followed by design charts and tables (42%). No immediate value is credit to
Expert systems, CAD systems, textbooks and computer programms
A simplified version of the Code is also mentioned, probably because this suits in the
present design process in the UK
A question about a realistic price for desing tools, showed that especially flow charts
and a textbook have to be rather cheap.
However in the inquiry the use of IT tools are not envisaged as of main importance, IT
will influence the design process more and more. The actual advance of the computer in
the design process started in the seventies, although it has not -up untill now- pushed
aside the tables and graphs yet. EC2 may be used as a starting point and may stimulate
Application of European concrete standards 75
Untill now in The Netherlands not much experience has been gained on the use of
Eurocode 2, although a persistant group of about hundred engineers and designers are
working with or their work is related to the Eurocode. Last few years some companies are
really working with EC2 e.g. because it was prescribed. In a large Demonstration
project on Eurocode 2 the Concrete Society together with experts from the building
industry have designed a complete office building according to EC2, and have compared
the results to a design concept with the national Code.
When the NAD was composed, deliberately a limited amount of adaptions on EC2 has
foreseen, to anticipate on future harmonization.
Depending on the expectations one could draw specific conclusions with reference to
the results. As there are differences, there are also many similarities. Main differences did
occur on following items:
Stress-strain diagram
material properties
strength classes
concrete cover related to certain parts and crack width
shear
columns
fire resistance
bending
The required reinforcement for beams and plates according to EC2 is appr. 10% higher
compared to the national Code. Main factor is the increased Load-factor for permanent
loads. When the NAD is used this percentage is reduced because regular load factors
from the Dutch Code are used.
Appropriate concrete technology 76
For columns according to EC2 about 530% more reinforcement is needed. Again,
using the NAD this is reduced to 020%. The difference increases when the influence of
the normal force is getting bigger (with a materials factor).
With reference to deflection of slabs in EC2 some tables are provided. Basic ratios for
span/effective depth are given related to the reinforcement ratio. Looking at the Dutch
Code, using slabs with high permanent loads the basic ratio of span/effective depth a
smaller ratio is allowed. Slabs with limited loads are allowed to have a higher basic ratio.
Eurocode 2 leads to about 10% higher reinforcement for shear. Applying the NAD
about 5% difference remains. Sometimes this difference may be larger; e.g. when
detailing provisions are taken into account. In Eurocode 2 minimum shear reinforcement
is at least two times the value in the Dutch Code.
The size (thickness) of flat slabs is -related to punching- about 30% higher in
Eurocode 2. at a much earlier stage column heads or extra reinforcement for punching is
needed. As in The Netherlands this kind of slab is frequently used, this would not be
appreciated very much.
Furthermore, the aspect punching is subject to a major difference between EC2 and
the Dutch Code (value for 1). Another major difference is the way excentric punching is
handled; a rather simple method in used in EC2, but the question can be raised if this
method follows the real punching behaviour.
The approach for torsion in EC2 is based on the fact that all moments need
reinforcement, whereas the Dutch Code provides a threshold. These different approaches
may lead to different amounts of reinforcement, which is fostered by the higher minimum
value for reinforcement stirrups. Furthermore, detailing provisions in EC2 may lead in
general to more required reinforcement ratios.
Again, the NAD reduces these differences.
Some of the results are shown in figures 13.
Summarized, some calculations showed a more conservative approach (related to
reinforcement %) with EC2. On the contrary, e.g. shear calculation methods used in the
Dutch Code are far more concervative compared to EC2.
Use of EC2+NAD appeared not to be not complicated with reference to the National
Dutch Code. Structure of the Eurocode 2 is nevertheless more complicated (where is
what to be found). Sometimes it is hard to get a clear view on all relevant aspects and
articles related to a specific subject.
A representative designer was able to use EC2+NAD in a few days, and had the
opinion that design aids as graphs, tables and flow charts were quite necessary. As they
are to be supplied by the SPRINT project in March 1996, and the official NAD and ENV
199211 have been published in The Netherlands last year, nothing is keeping the Dutch
engineer from use of EC2+NAD, however the acceptance of calculations by the building
authorities should be clear. At the moment the Building Authorities are not all quite
convinced of the equality of the design concept of EC2+NAD and the Dutch Code, and
they have no experience in control of EC2 calculations. Administrative and
organizational initiatives are
Application of European concrete standards 77
taken at the moment to improve the situation. One can deliver EC2+NAD design
calculations at the desk of the Building Authority, but still has to prove the conformity
with the National Dutch Code before they are accepted. It is to be foreseen that in 1996 in
The Netherlands more and more EC2+NAD calculations will be made and authorized.
The Concrete Society of the Netherlands is stimulating this process by offering the
building practice a package for the use of EC2, containing a (draft) translation of ENV
199211, the Book on Design Aids, an informative brochure on use and implications
and the official issue of the NAD.
ENV 206
With reference to ENV 206 this document was also tested the process of execution of
concrete structures at the building site of two tunnels at Schiphol Airport. It has to be
emphasized that the application of ENV 206 was not initiated-to adapt the ENV 206 to
the relevant Dutch Code (VBT/VBU). Our opinion is that a widely accepted Code has to
contain performance requirements in order to enable all countries to use their national
methods as long as they satisfy these requirements. Although there are textual remarks
(more concrete), too many references and too many notes, the ENV 206 is generally
considered as applicable, apart from the folowing aspects:
FUTURE OUTLOOKS
EU with an approved Eurocode Programme means a market with more ideas and funds to
initiate, develop and implement new technologies. As EC2 provides new challenges for
automatic procedures because it is not related to any specific (conservative) design
practice, this is in favour of market possibilities for development of software. Especially
for smaal countries certain developments are just unfeasable when they are taken up on a
national basis only. EC2 is in fact more intricate in the procedures to follow compared to
procedures from the past that are less formal but easier to apply. IT can and will probably
play a major role in handling of partial safety coefficients and the increased amount of
checks on load combinations and limited states to be envisaged.
The use of the NAD is an inevitable step towards total implementation of the
Eurocodes. Preferably, NADs should be as short as possible, to facilitate introduction of
the Eurocodes as ENs. When the national NAD is well structured and concise,
principally it is not a problem to use it in practice, as trial calculations in The Netherlands
point out.
With the increasing activity on IT tools, the use of CAD/CAM is increasing also in
many countries. Especially this area needs a harmonized Code, for the different systems,
various national Codes, symbology, language and representation of reinforcement are still
very far from each other. E.g. an English reinforcement drawing may well be totally
unusable in Germany. For the execution of concrete structures and information exchange
in the European building process these aspects are of main importance, beside the design
Code.
New developments, leading to better hardware and software for less costs will
influence the future of design. As use of commonly used tools as graphs and tables, IT
tools should be incorporated in education. International cooperation to develop and
implement such software, based on EC2, should be made. When we look at the European
Appropriate concrete technology 80
steel construction initiatives in this field, we may conclude they are ahead of the concrete
IT-design and concrete IT education programmes (of course Standards for Steel
Construction are more suitable for IT application). The availability of flow-charts for
EC2 is a step forward in this matter. Flow charts are very useful for development of
software programmes.
Government and Industry should not hesitate to use advanced but realistic information
systems in construction. But together with implementation of these developments, quality
control and the engineers touch will be needed. The engineer is responsable for its
design and the calculations; he cannot blame the IT tools he has used.
Further to EC2, in 1998 a package of (a limited number) ENs for design of concrete
structures is foreseen. When we look at the publication of European technical books on
concrete construction, we may determine a small but clear increase of books on
Eurocodes.
With this in mind, and knowing that some aspects of Eurocode 2 need a sound
discussion of some design problems, the time-path towards 1998 is short, however
achievable. With proper design aids and increasing use of IT tools in the near future, the
treshold to use EC2 is lowered.
REFERENCES
1. Jong, ir. P.de e.o., Demonstratieproject Europese Regelgeving, Betonvereniging, 1994, appr. 120
pp.
2. Stasse, Ing. J, e.a., Vergelijkingsberekeningen VBC-ENV 199211, CUR, 1993, 110pp.
3. Proceedings International Congress European Concrete Standards in Practice, Betonvereniging,
December 1994
4. Boer, ir. G.P.L.den, Litzner, Dr.-Ing. H.-U., Walker, M., Final Report SPRINT project RA 358
phase 2, 1994, 30 pp.
IMPLICATIONS OF STRUCTURAL
EUROCODES
S B Desai
Department of the Environment
UK
INTRODUCTION
The European Community (EC) had the concept of creating a common market, which
was in a continuous state of development since the year 1950. In 1985, therefore, the
Member States asked the Commission of the European Communities (CEC), which is
one of the institutions of the EC, to put forward a plan to achieve a unified internal
market by the year 1992. It was recognised that the construction industry and
construction products represented a large proportion of the market. For a considerably
Appropriate concrete technology 82
long time, the EC had tried to introduce an agreed EC standard for health and safety, with
adjustments to the individual national regulations. The CEC recognised that this task
would be too difficult and it would not be practicable to achieve it within a finite time-
scale. The CEC adopted a new approach to technical harmonisation of standards of
health and safety for the EC. The EC policy on technical harmonisation is put in place
through its Directives. For the construction industry, the most important Directive is the
Construction Products Directive (CPD).
The CPD is aimed at providing freedom of sale and use of construction products,
provided that the products have such characteristics that the structures in which they are
incorporated meet the following essential requirements.
1. Mechanical resistance and stability
2. Safety in case of fire
3. Hygiene, health and environment
4. Safety in use
5. Protection against noise
6. Energy conservation and heat retention
Each requirement is associated with an Interpretative Document (ID), which interprets
the expectations of the requirement. In the main, the IDs are not intended for the use of
designers of structures. The IDs are often described as the codes for writers of the
Eurocodes. For example, the Limit State Concept is incorporated in the ID for the
requirement Mechanical resistance and stability and this concept is the basis of all
structural Eurocodes.
The structural Eurocodes are intended to serve as reference documents for two main
purposes. First, they should serve as a means of proving compliance of building and civil
engineering works, mainly with the essential requirement Mechanical resistance and
stability. The Eurocodes also concern the second requirement Safety in case of fire,
since fire is accounted for as an accidental action in the design of structures. Second, they
are expected to form a framework for drawing up harmonised technical specifications for
construction products.
The CEC had intended that the Eurocodes should initially serve as an alternative to the
different national codes in use in the various Member States. The CEC had expected that
the Eurocodes would ultimately replace these national codes. In 1990, after consulting the
Member States, the CEC transferred the work of further development, issue and updating
of the structural Eurocodes to CEN, the European Committee for Standardisation. CEN
technical committee CEN/TC250 has the responsibility for these structural Eurocodes.
The sub-committees of CEN TC 250 (SC1, SC2 etc) are responsible for Eurocodes 1, 2,
etc.
CURRENT POSITION
In the first instance, various parts of the Eurocodes are published as ENV (European pre-
standard), when the final drafts prepared by the project teams are approved by the
appropriate sub-committee of CEN/TC250. The Member States are expected to
Implications of structural Eurocodes 83
encourage the use of ENV stage codes accompanied by the National Application
Documents (NADs), which contain the following information:
a) The values of parameters acceptable to the National Body (for example, BSI in the
UK) in place of the Boxed values given in the ENV; (The boxed values are without
prejudice and are not necessarily those which will be agreed by the UK at the EN
stage.)
b) Reference to supporting standards and codes which are necessary for the use of the
ENV. (e.g., loading/actions, material specifications, fire etc.);
c) Additional information and guidance required for the use of the ENV.
In the UK, the ENV, NAD and the National Foreword are published by the British
Standards Institution (BSI) as single document DD ENV .. as a draft for
development. Preparation of NADs is overseen by the relevant sub-committee of the
BSI. The actual work is usually supported from the Department of Environment (DOE)
and it is led by the Building Research Establishment (BRE), on the basis of the following
examinations and studies:
i) Textual Examination
A detailed examination is carried out to provide comments on the following topics, using
the relevant national standard as a model:
a) General lay out, cross-referencing and scope of the subjects covered by the ENV;
b) Consistency of the draft rules;
c) Significant differences in formulae and procedures;
d) General clarity, readability and usability of the draft;
e) Implications for designers legal responsibilities;
f) Feasibility for outlining a step by step design procedure.
indicated as -x, -y, and -z will be given in chronological order of approval. The last
column indicates whether the European Pre-standard is in English (E), French (F) or
German (D), or whether it is approved and it is in the process of publication (Ap), or at
which target date (9Y-MM) the approval is foreseen. A letter for the language (E,F or D)
followed by a date YY-MM indicates the month in which the publication is foreseen in
that language.
Table 1: Eurocode EC1 (Basis of Design and
Actions on Structures)
ENV 19911:1994 Part 1: Basis of design E
ENV 19912 1:1995 Part 21: Actions on structures E
Densities, self- weight and imposed loads 9502
ENV 19912 2:1995 Part 22: Actions on structures E
Actions on structures exposed to fire 9502
ENV 19912 3:1995 Part 23: Actions on structures E
Snow loads 9502
ENV 199124 Part 24: Actions on structures E
Wind actions 9503
After the ENV stage of some 3 years, the ENVs should be transformed into EN
(European Standard). All the available ENVs can be purchased from the European
National Standards Institutes which are members of CEN. For countries which are not
members of CEN, the English version can be purchased from BSI in London, the German
version from DIN (Deutsches Institut fr Normung) in Berlin, and the French version
from AFNOR (Association Franchise de Normalisation) in Paris.
SUMMARY OF UK ACTIONS
The UK experts have had a large share of representation on the project Teams drafting
the Eurocodes. In addition, other activities dedicated to the Eurocode cause have been
progressing, mainly sponsored by the DOE, with participation by the Building Research
Establishment (BRE), BSI and a number of private sector experts. Some of these
activities are listed below.
Technical Coordinators
Technical Coordinators have been appointed for each Eurocode, through the
commissions funded by the DOE, to serve the relevant BSI committees. Each Technical
Coordinator represents the views of the BSI committee at the CEN subcommittee
meeting and reports to the chairman of the BSI sub-committee, and liaises with the DOE
on matters concerning implications for Building Regulations.
Order of production
The ENVs of structural material codes (EC2, EC3 for concrete, steel etc.) were issued in
advance of the ENV versions of EC1 (Basis of Design and Actions). The NADs of EC2
and EC3 were based on their comparisons with the corresponding national codes, using
the information given in the BS loading codes. As a result, the NAD for ENV Eurocode
EC1 may have to align the values of loading closely with the corresponding figures in the
BS, in order to maintain the validity of the comparison exercises used as the basis of
NADs of EC2, EC3, etc.
relevant to the Eurocode Parts 1.2 (Structural Fire Design), which refer to the standards
on certain protection measures and tests.
At their recent meeting in Lisbon, the committee stated key target dates, relating to those
parts of the Eurocodes necessary for the design of buildings. CEN/TC250 is developing
an action plan to achieve the following:
i) All the EN Eurocodes necessary for the design of structures for buildings comprising
Basis of Design, Loading and General Design Rules including those for Fire and
Seismic Resistance, to be published by 1 January 2000.
ii) The first group of documents consists of Part 1 and Parts 21, 23 and 24 of
Eurocode 1 and Parts 11 of Eurocodes 2, 3 and 4 to be published before the end of
1998.
The ENVs for bridge parts are to be approved by the end of 1996 to allow their
conversion to ENs before the end of 2001.
The development of the Eurocodes is still a step-by-step process. The plans for
action and the dates, therefore, can only be conditional or subject to completion of certain
key activities. CEN rules require that, when a full EN standard is issued, the conflicting
national standards have to be withdrawn within 6 months. However, in the case of
Eurocodes, the CEC mandate for CEN specifically states that this will not happen. After
the Eurocodes reach EN status, CEN, CEC and the national governments will consider
the arrangements for superseding national standards and they may agree a period of
coexistence of the national standards and the Eurocodes.
The regulatory systems of other countries differ from the UK system. In some
countries, the adoption of structural Eurocodes may produce legal and administrative
problems that may be more difficult to overcome, quite apart from any commercial or
technical issues. It will be some time before one can say anything definite about the
decision for replacing the national standards by Eurocodes or the length of the transition
period. The DOE is actively pursuing these issues with the CEC.
The best indication so far on transition, is that several countries are already revising
their national standards for steelwork to bring them voluntarily closer to the Eurocode 3
(for example The Netherlands and Switzerland). In France the ENV version of EC3
already has full legal equivalence with other standards. The smaller countries (Ireland,
Implications of structural Eurocodes 89
Luxembourg etc) are amongst the keenest on harmonisation. Even CEN associates (for
example Rumania) are issuing versions of EC3 as national standards.
The CEN asked the 18 members, if they should proceed to convert the present ENV
version of EC2, EC3 and EC4 into ENs and, if so, whether with or without technical
changes. This enquiry has received a positive response. The CEC/CEN agreement
provides that, in any case, CEN cannot proceed with the conversion without the consent
of the European Union and the European Free Trading Area. It is expected that CEC will
base their decision on views from national governments through the Standing Committee
for the Construction Products Directive.
In accordance with CEN rules, all standards are subject to a 5 year review and the
Structural Eurocodes will be no exception. The plans for final versions in 1998 will,
therefore, require a rapid conversion. (There is little more than 12 months for technical
discussion, the rest is needed for translation, editing and such other procedures). The
timetable can only be met if technical changes are restricted. It is envisaged that
delegations will agree to this, only if they are promised an early start on a more radical
review in some 5 years time. The CEN rules also allow for a maximum 3 amendments
between revisions. Whether this will be found useful (or even practicable) remains to be
seen.
The Eurocodes provide a relevant method for demonstrating suitability of products.
Although the Eurocodes are not formally linked with the CPD, they have a role in the
practical implementation of the CPD. In addition, the Eurocodes have a role concerning
responsibilities on designers, amongst others, which are stated in the new Construction
(Design and Management) Regulations 1995.
The Eurocodes have an important role in relation to Public Procurement (works,
utilities). Public Procurement Directives require tenders to use European standards, where
they exist. A number of CEN product standards (typically for fabricated products, rather
than basic products) have a need for design rules. The policy is that they should refer to
Eurocode rules, as far as practicable. This is an important reason why the EN versions of
Eurocodes are needed promptly.
FRACTURE MECHANICS
PARAMETERS OF CONCRETE: TEST
METHODS DEVELOPMENT AND
HARMONIZATION OF STANDARDS
S N Leonovich
Byelorussian State Polytechnical Academy
Republic of Belarus
INTRODUCTION
Situations where tensile strength and tensile toughness are of particular importance are,
for example, anchorage of deformed bars, shear forces in slabs and beams, splitting under
concentrated forces, and unreinforced pipes.
With authors participating State Standard 2916791 Concretes. Methods for
determination of fracture toughness characteristics has been developed.
This paper examines how the crack resistance of concrete varies with the volume
concentration of different aggregates .
By this time the large volume of experimental data has been accumulated on the main
parameters of fracture mechanics of concrete [17]. Nevertheless,, the application of
gained values of force and energetic parameters fracture parameters in practical
calculations is connected with many difficulties. It is explained by the lack of normalized
fracture parameters gained by completely balanced diagrams of deformation.
The research has been started in the CEB in the direction of development of concrete
classification approach. The concrete classes statements defined by the parameter of
fracture energy Gf, applied for the fracture were included in the cods model. Concrete are
classified on dependence on the size of coarse aggregate [8] :
The stable loaddeflection curves from three-point bend tests for different types of
concrete are presented in Fig. 2, 3. As can be seen in Fig. 4, 5 and Table 1, the strength
(R), deformativity (Eb), crack resistance of concrete are strongly affected by the quality
of the aggregate. In comparison with cementsand mortar (j=0), the compressive strength
of granitoid concrete with volume concentration of crushed-stone j=0.353 was increased
Test methods development and harmonization of standards 93
by 6.5% and with j= 0.621by 55% respectively. The compressive strength of gravel
concrete decreases when the coarse aggregate /cement-sand mortar ratio increases: by
1.4% (j=0.375) and by 25% =0.643).
In Fig. 5 the experimental energetic (Gf, Ji) and force (Kic) parameters of concrete against
volume concentration of granite or gravel are presented. It can be seen that both the
resistance to crack initiation (in the ascending branch) and resistance to crack propagation
(in the descending branch) increase with the increasing volume concentration of crushed
stone approximately by 2 times (j=0.353), by 2.3 times (j=0.621) in comparison with the
cement-sand mortar. Increase in resistance to growth of crack for concrete with the
concentration of crushed stone (j=0.353) is essential (Gf by 79 %).
Coarse aggregate is more important for greater crack resistance than for strength of heavy
weight concrete.
There were some differences between the results obtained with crushed stone concrete
and those with gravel concrete.
For the gravel concrete with the moderate concentration of the coarse aggregate j=0.
375 there was a considerable increase in crack resistance parameters at the post-peak
stage (Gf by 81%), and their constant values at the pre-peak stage of deformation (Figure
5).
For the gravel concrete with the maximum concentration in coarse aggregate j=0.643
there was the considerable decrease of the crack resistance parameters at the both stages
of deformation.
The results of this study show conclusively that the use of optimum volume concentration
of coarse aggregate in concrete contributes significantly to its crack resistance.
The optimum volume concentration is less than maximum concentration. We
suggested that the method of concrete classification by crack resistance should be built in
this sequence:
a) normalized values of crack resistance parameters depending on the concrete classes by
strength;
Appropriate concrete technology 96
In the tables 3, 4 are presented crack resistance indexes Gf, Kic of concrete based on
crushed stone and gravel in dependence on fine concrete strength and volume
concentration of coarse aggregate.
Table 3 Fracture energy Gf(N/m) of Crushed-Stone
Concrete
VOLUME CONCENTRATI ON OF STRENGTH OF CEMENT-PASTE Rcp
COARSE AGGREGATE (%) (N/mm2)
15 25 37 50 60 70
10 28.9 36.5 45.7 55.6 63.2 70.9
20 29.2 36.8 46.0 55.9 63.5 71.2
30 29.5 37.2 46.3 56.2 63.8 71.5
Test methods development and harmonization of standards 97
By this time the large volume of experimental data has been accumulated on the main
parameters of fracture mechanics of fine concrete [17]:
Kic=m1Rbt,
Kic=m2R,
Kic=nKic,
CONCLUSIONS
1. The classifications of concrete crack resistance on the base of force and energetic of
fracture mechanics criteria is proposed.
2. The formulas for determination of force and energetic parameters in dependence on
fine concrete strength and coarse aggregate concentration are obtained on the base of
experimental data.
3. Correlation between crack resistance of cement stone and crushed-stone concrete and
gravel concrete is determined.
REFERENCES
ABSTRACT. The allowances for the long term creep effects in the
British Standard and in the Eurocode 2 for the design of reinforced
concrete columns have shown some discrepancies. If the British Standard
for the design of a reinforced concrete column were followed, a smaller
column section would be required than if the Eurocode 2 were adhered to.
As a result of this, the British design approach will be more economical
and thus more competitive. In order to understand fully the long term
behaviour of concrete columns subjected to high stresses and to improve
the level of confidence in our design, this research project has been
established.
INTRODUCTION
The new Eurocode 2: Design of concrete structures (EC 2) [1] has put more emphasis on
the time-dependent effects on the design than that recommended by BS 8110: Use of
structural concrete [2]. The design approach is similar between the BS 8110 and the new
EC 2 in large except in the treatment of time-depending effects. In BS 8110, a strain of
0.0035 is treated as the ultimate strain of a concrete element, whereas in EC 2, various
Appropriate concrete technology 100
situations are considered and limits are set for certain situations. Sections which are
subjected to compression throughout are allowed a limiting total strain of 0.002 and those
that have their neutral axes on the least stressed side of the section are allowed a limiting
strain of 0.0035. For sections intermediate between the two situations, interpolation
between the two limits is allowed.
Most of the studies on concrete creep behaviour have concentrated on members
subjected to medium or low stresses. The limited published data in studies of creep
behaviours in highly stressed members have been found to be contradictory. In this study,
the model columns were cast and were stressed to above 70% of their ultimate strengths.
These prisms were allowed to undergo further deformation under the applied load.
Creep under very high stress accompanied by very high lateral creep will lead to
internal cracking when Poissons ratio is in excess of 0.5 [3]. This can in turn affect the
ultimate compressive strength of the member.
Work by Reid et. al. [4] on eccentrically loaded columns found that creep strains about
the minor principal axis due to bending are greater than those about the major principal
axis and axial creep strain is less than creep strain in bending. This again supported the
EC 2 approach for the design concrete member in compression.
In BS 8110, the limiting strain of 0.0035 is allowed for the total strain which includes the
elastic strain due to the applied load and other effects such as shrinkage and creep. A
simplified approach for estimating ultimate creep is recommended in BS 8110. The
approach follows an earlier method based on Comite Europeen du Beton (CEB) 1970.
The ultimate creep function is expressed as:
where is the ultimate creep coefficient which is obtained from Figure 7.1 of BS 8110:
Part 2. Thus from knowing the ambient humidity, age at loading and effective section
thickness, the ultimate coefficient can be estimated.
Where there is no moisture exchange, i.e. in sealed or mass concrete, basic creep is
assumed to be equivalent to that with a volume/surface ratio greater than 200 mm at
100% relative humidity.
Although there is no provision for estimating creep coefficient as a continuous
function of time, approximately 80, 50 and 30 percent of the ultimate creep will develop
at a rate corresponding to an effective thickness of greater than 400 mm.
Effects of concrete creep on the ultimate compressive strength of concrete columns 101
EC 2 treated shrinkage and creep effect as second order effects. Section 3.1 and appendix
1 of EC 2 gives guidelines on the estimation of time-dependent effects. The creep
function is given by:
Neville [5] suggested that failure of concrete would occur at a limiting strain whether that
strain is reached by rapidly applied high stress or a lower sustained stress. Neville [6]
showed that for concrete loaded at 7 days and 28 days old to a stress/strength ratio of
higher than 0.85, failure by creep could occur. Continho [7], found that for concrete
failure induced by creep to occur in older concrete, the stress/strength ratio had to be even
higher. An experiment carried out on six months old concrete demonstrated that a
stress/strength ratio of 0.96 could cause the concrete to fail by creep.
Hughes et. al. [8] loaded concrete up to 25% of its 28 day strength and maintained the
load for over ten years before assessing its ultimate strength The strength of the
specimens which had been subjected to long term loading were found to be similar to
those which were not loaded during that period.
Roll et. at. [9] conducted tests similar to Hughes et. al.. A range of stress/strength
ratios were studied and it was found that the ultimate strength of the specimens which
were subjected to long term loading was increased.
Sturman [10] proposed a relationship for eccentrically and concentrically loaded
concrete elements to express the fact that the maximum stress and strain could be higher
in a member subjected to flexural compression than that subjected to axial compression.
EXPERIMENTS
Unreinforced concrete prisms, 5050300 mm were cast for this study. After curing for
28 days in the curing tank, some of the prisms were put under an axial load in a test rig
and the others were left unstressed under the same environment. The prisms that were
stressed were taken to close to their ultimate strength at that stage. The load in the prisms
was maintained in the rig for a period of testing.
Applied Load
The testing procedure for the long- and short-term creep effects varied slightly. For the
prisms which were stressed to about 90% of their predicted strength at that age, were
tested by applying a constant axial load through the hydraulic ram of an Amsler testing
machine. The change of strain in the prisms was measured throughout the test. Prisms
which were stressed to a lower stress/strength ratio were expected to creep without
causing failure for a longer time. They were placed in a testing rig under which the
compression was maintained by a spring. During the testing period, these prisms were
kept in the curing room where the ambient temperature and the moisture were kept
constant.
Duration Of Test
The prisms, which were subjected to spring loads, were monitored for their deformation
for four months. In general, the spring loads were then removed from these prisms and
they were allowed to recover for three weeks before they were load tested to failure.
Monitoring
The creep and the shrinkage in the prisms were monitored for the testing period. Prisms
from the same batch which were not stressed were placed next to the axially loaded
prisms to account for other effects such as shrinkage and drying creep. The difference
between the stressed and the unstressed prisms gave a measurement of the creep in the
prism.
Both strain gauge and Demac gauge were used to measure the creep strain and
shrinkage in the prisms. Strain readings at mid-height of the prisms were presented.
Samples
Three sets of prisms were cast to investigate the effect of creep on the ultimate strength of
the prisms. Four more sets of prisms were cast and were subjected to compression in a
testing rig for a duration of three months. During this period of time, the specimens were
kept in the curing room. In each batch of tests, three prisms were subjected to
compression and another three prisms were left unloaded in the curing room to monitor
any environmental effects such as shrinkage of the prisms. The prisms in the testing rig
were then unloaded and were kept in the curing room for a further three weeks to allow
them to recover from both elastic and creep strain.
Effects of concrete creep on the ultimate compressive strength of concrete columns 103
Figure 1 shows the stress/strain characteristics of a prism loaded to 80% of its predicted
strength and allowed to creep. The creep effect slowed down quickly after the load
increment ceased. The prism did not fail in this case and the ultimate strength was as
predicted.
Figure 2 shows the stress/strain plot of a prism which was initially loaded to 85% of its
predicted compressive strength and allowed to creep. The compression was increased
when the increment in the strain became insignificant with time; about two hours after the
load was maintained. The load was increased to 90% of its predicted capacity. The load
was maintained and the prism failed about twenty minutes later. The failure of the prisms
was similar to that due to direct compression.
Appropriate concrete technology 104
Figure 3 shows a stress/strain plot of a set of prisms subjected to a stress of 70% of their
predicted strength and allowed to creep for four months. In this situation, concrete creep
has not caused failure in the prisms. The prisms recovered some of their elastic strain and
creep strain when the compressive loads were removed. The load test of these prisms
showed that the ultimate strength of these prisms was not different from those prisms
which had not been subjected to axial compression.
ULTIMATE STRENGTH
From the investigations of the long- and short-term loading of the prisms, it appeared that
creep strain could lead to failure of a compression member when its total strain exceeded
a limiting strain. Where the creep strain in a member did not cause the total strain to
exceed the limiting strain, its effect on the ultimate strength was insignificant.
Effects of concrete creep on the ultimate compressive strength of concrete columns 105
The failure stresses of the prisms from three sets of tests are shown in Table 1. The
results suggest that the ultimate strength of a prism is not significantly affected by the
axial load applied previously.
Table 1 Comparison of the ultimate strength of the
concrete prisms
Failure stress of prisms which were Failure stress of prisms which were not
subjected to compression before subjected to compression before
(kN/mm2) (kN/mm2)
Batch 21.8 24.6 24.6 23.9 22.6
1
Batch 24.1 24.4 26.0 24.8 23.8 27.1
2
Batch 24.0 20.6 28.6 26.4 26.4 26.5
3
Appropriate concrete technology 106
The creep coefficients obtained from BS 8110: Part 2 for the concrete prisms were found
to be larger than those obtained from EC 2 [11]. That is the predicted ratios of specific
creep to the specific elastic strain on the application of load were higher when applying
the procedures recommended in BS 8110: Part 2.
Brooks [12] compared various methods for the prediction of concrete deformations.
For mass concrete members, it was shown that the creep function predicted by BS 8110
was higher than that found from EC 2.
CONCLUSIONS
The short term loading on the prisms has given valuable information in relation to the
strength of concrete columns. It has been shown that the effect of creep in a compression
member could lead to failure of a compression member when the total strain in the
member exceeded the ultimate strain. The allowance of an ultimate strain of 0.002 in EC
2 for members subjected to axial compression appeared to be more relevant to the real
behaviour of the member.
The results from the two stage loading of the prisms have shown that the small amount
of creep strain in comparison with the elastic strain is not usually the major cause of
failure of the member.
The results from the prisms which were subjected to long-term loads have shown that,
where the total strain in a member does not exceed the ultimate strain during its life, the
ultimate compressive strength of the member is not significantly different from those
which were not loaded. The members recovered from both elastic and creep strain when
unloaded.
The predicted creep coefficient and the total strain obtained from BS 8110: Part 2
were found to be higher than those obtained from other prediction methods.
Further work is being undertaken to investigate the ultimate strength of concrete
prisms which are not allowed to recover from elastic and creep strain.
FURTHER WORK
Further work is being undertaken to investigate the creep effect on the ultimate strength
of concrete prisms which have been subjected to long-term creep effects and were load
tested to failure without recovering from elastic strain, creep strain or shrinkage strain.
An assessment of the actual total strain in concrete columns designed following BS
8110 is being carried out to verify that they will not fail during their design life as a result
of the total strain in the columns exceeding the limiting strain.
Effects of concrete creep on the ultimate compressive strength of concrete columns 107
REFERENCES
INTRODUCTION
dangerous and risky, construction is not possible without the workers involvement.
Despite of mechanisation of construction in developed countries, manual operation are
still essential.
Shortly after 1800, Portland Cement was discovered. Since then, it has become the
most used cementing material in the construction industry. [2] Because of its versatile
nature, concrete can be formed into different shapes as required and thus it has become an
important and widely used material in modem construction.
Construction has direct relationship with development because, it fulfills peoples
desire for socio-economic development. Development has always been for the service of
mankind. This paper tries to focus the condition of such workers who work with
concrete/cement for the service of mankind and need special consideration.
Construction workers in Nepal, basically, are peasants and are thus, seasonal. Most of
them are not permanent in the trade. Annual requirement of construction workers is about
200,000. Influx of construction workers, having more skill and experience are increasing
day by day from the neighbouring country, India. Construction workers both skilled and
unskilled do not need formal training or license to practise. They are most unorganised
and they do not have any unions. These all have made them very weak in protecting their
interest and enhance the trade.
Daily wages of skilled and unskilled workers fixed by the Government are Rs. 130 and
Rs. 80 (1 US$ is equal to Rs. 50.00) respectively. The prevailing wage in residential
construction is Rs. 140 to Rs. 200 for skilled and Rs. 70 to Rs. 100 for unskilled.
Normally, if the job is for longer duration and if the workers are working with a
contractor, lower rate are paid and in the smaller projects directly handled by the owner,
higher rate are paid. However, the wage obtained by the construction workers is
encouraging in comparision with the other workers in other trades. Apparently the basic
reasons are lack of job security, appropriate working condition and facilities, safety
measures etc. Paid leave is not available to them. Except a snack break, they do not even
get one day paid leave in a week. As a consequence of this, against the ILO Standard of
at least 24 consecutive hours of rest per week (Convention No. 14), workers are
compelled to work without taking any rest. To make more money, if job is available, they
work up to 14 hours a day.
Temporary labour huts with poor ventilation and light are provided for the workers
who live at construction site. Huts are built with stacking bricks or by curtaining with
CGI sheets. In some huts, cement is also stored.
By virtue of socio-economic condition of developing countries, involvement of
women workers in construction labour has become a common scenario and their
participation in heavy manual labour has a considerable effect on their health. As
revealed by the survey carried out by the International Labour Organisation (ILO) among
women construction workers in Bombay- miscarriages and menstrual disorders are just
two of the many health hazards these workers have to battle against. Many such females
suffer from prolapse of the uterus and backaches. [3] The situation of women workers in
Nepal is not so severe as the survey revealed. There still exists the general concept that
Appropriate concrete technology 110
heavy load carrying and overtime working are generally responsibility of male
construction workers. Comparatively lighter activities like marble chips rubbing, sand
screening, mortar preparation, moving concrete pan from hand to hand for concreting are
the major jobs women construction workers perform. The heavy type of job they mostly
perform is crushing stones manually for producing aggregates. These jobs are considered
suitable for women workers.
The revised version of Safety Provision (Building) Convention 1937 (No 62) is the
Safety and Health in Construction Convention 1988 (No 167). This Convention applies to
all construction activities (building, civil engineering, erection and dismantling on a site),
from the preparation of the site to the completion of the project. [6] The main aim of this
Convention is to ensure safety and health in construction. As per this Convention all
appropriate precautions shall be taken to ensure that all workplaces are safe and without
Health and safety in construction in Nepal 111
risk of injury to health. However, such important Conventions are still to be ratified by
the Government.
After the collapse of the Panchayat (partyless) system of government in May 1990, the
new Constitution was promulgated in November 1990. The Labour Act 2048 came into
force on 15 May 1992 which has some special provisions related with construction
industry. Before this, there was no special law regarding labour in construction industry.
The law regarding Factory and Factory Workers 2019 which was formulated under the
Factory and Factory Workers Act 2016 was the reference at that time. Labour Act
2048 has following special provisions applicable to the construction industry:[7]
a. sufficient tools and equipment are to be provided by the management at construction
site.
b. it is managements responsibility to manage shelter, fooding, drinking water etc., if
more than 50 workers are involved in a construction.
c. it is managements duty to have workers insurance against accident.
d. it is managements duty to keep construction site safe.
e. it is managements duty to provide necessary personal protective equipment (PPE) for
the workers involved in construction.
In general, Labour Act 2048 has various provisions and are also supposed to be
applicable for the construction workers also. Some of them are:
a. Working hours should not be more than eight hours a day or forty eight hours a week
and there should also be a holiday in a week.
b. Government shall fix the minimum wage for the workers and no workers shall be paid
less wage than the minimum.
c. Various clauses regarding health and safety.
d. Clauses regarding welfare activities such as welfare fund, compensation, rest room etc.
The provisions of the Act have an effective impact on the factory workers, but for the
construction workers, they have very little practical implication. Thus, they are not found
practically strong to favor the construction workers in Nepal.
Stone, brick, timber, mud and lime mortar are the traditional building materials widely
used in Nepal. Concrete construction in Nepal started about half a century ago. Palace of
then Prime Minister Mohan Sumsher (April 1948-November 1951) is considered as a
first concrete structure in Nepal. The technology used in this structure was influenced by
the British Indian technology prevalent in the then India. At the beginning concrete
constructions were a status symbol. Government buildings were built in accordance with
it. The early application of the concrete were limited to the slab-like structures.
Gradually, people became aware of the inherent properties of the concrete-strength,
durability and easy to cast into desired shape. Tradesmen got opportunity to be trained in
concrete structures. This attracted ordinary people towards concrete structures.
Deforestation limited the availability of timber. Cost of the materials were also
increased. On the other hand, Nepal being a mountainous country, aggregates are
Appropriate concrete technology 112
Most of the large/major constructions in Nepal follow the FIDIC Conditions of Contract
or WB guideline for procurement. There is not uniformity in Conditions of Contract
applicable to constructions. Sometimes, only simple agreement works between the Owner
and the Contractor. It is also found through observation that even the Department of
Building has not prepared detailed conditions of contract and specifications. In such a
situation, effective implementation of health and safety standards are hardly possible. No
construction site is inspected by the officials from the labour office and in general, no
Safety Officer is appointed in the construction site.
The form work and false work are not properly designed. In March 1994, form work
and false work of a boiler house of a five star hotel in the capital city collapsed during
concreting and claimed the life of a person and eleven others injured. [8] Similarly in
September 1994, while removing formwork from the newly casted roof of a school
building, whole of the slab collapsed. [9] There are many such unreported cases.
However, before concreting, the formworks are inspected by the engineer. There is
also no practice of hammering the nails exposed on the planks immediately. Piercing by
rusted nails are common and such happenings are not reported in any form.
Concrete mixer is used in mixing concrete. Pumping system is yet to be introduced.
Thus, concrete is kept in an iron pan and moved hand to hand to the place of casting.
Personal protective equipment such as boots, gloves, protective clothes are not
compulsory to all the workers. Some workers are seen with rubber boots but not all.
Some roll polythene sheet on their foot during concreting. Patches of cement slurry are
seen on the body of the workers.
The workers are not only untrained they are not orientated with the common hazards
and probable safety measures to be adopted. They are unaware of this. Working with
concrete does not exclude them handling cement. Government owned cement distributing
company operates from a godown with very poor ventilation and lighting. The porters
involved in loading and unloading of cement are found fully covered with cement dust.
They are not provided protective equipment like clothes, boots, gloves and masks. The
floor is found covered with a thick layer of cement dust.
Health and safety in construction in Nepal 113
ACKNOWLEDGMENT
During the preparation of this paper I am obliged to the following persons and wish to
herewith forward my sincere acknowledgment and thanks:
1. Professor Prem Nath Maskey, Department of Civil Engineering, Pulchowk Campus,
Institute of Engineering, Lalitpur, Nepal.
2. Mr. Deepak Bhattarai, Principal, Nepal Engineering College, Kupondole, Lalitpur,
Nepal.
REFERENCES
11. MAMORIA, C.B., Personnel Management, Himalaya Publishing House, Bombay, 1990. pp.
905
12. MAMORIA, C.B., Personnel Management, Himalaya Publishing House, Bombay, 1990. pp.
906
13. NATIONAL INSTITUTE OF CONSTRUCTION MANAGEMENT AND RESEARCH
(NICMAR), BOMBAY, Common Safety Practices in Construction Sites, pp 30
FOUNDATIONS FOR LOW RISE
BUILDINGS IN ROMANIA AND UK:
COMPARISONS AND EFFECTS
C Oltean-Dumbrava
University of Abertay, Dundee
W G Carter
Nottingham Trent University
UK
The paper will report on aspects of this research, with reference to the technical, quality
and other administrative procedures to show the differences in approach in the selected
countries.
Keywords: Residential development, foundation failure, legal requirements and quality
procedures in UK and Romania.
Crina Oltean-Dumbrava is a Romanian citizen currently appointed as a lecturer at the
University of Abertay Dundee, UK. She specialises in decision making on the basis of
multi-criteria analysis, total quality management in construction and other related
subjects in construction management and construction economics.
Geoffrey Carter is currently the Research Project Director exploring the nature of
residential foundation failure in several countries. He is a principal lecturer in building
technology at The Nottingham Trent University, UK.
Appropriate concrete technology 116
INTRODUCTION
Changing political climates in countries can impact on society and across the world in
many different ways. The effects of change can take time to percolate through a nation or
in some instances such change can be swift. The transition from one political system to
another brings demands that are sometimes painful, especially when a political system
has been in place for decades and urgent changes are demanded. Romania is one such
country. For over forty years, Romania functioned under a command economy political
system, then in 1989, a revolution occurred. Since 1990, Romania has sought to adopt a
western style of market economy. The process of change has not been easy for the
country as a whole or for its construction industry. Commercial pressures, of the kind
previously unknown, have brought with it many different stresses.
Despite frequently held opinion to the contrary, Romania is a technologically
sophisticated country. Romanias construction industry, through political dictat, was
forced to provide mass housing for a rapidly growing population and the migration of
people into cities and towns, because of the socialist industrialisation. Low and medium
rise blocks of flats were, of necessity, constructed to satisfy this demand. As there were
suitable, locally available raw materials and an abundance of labour, it was inevitable that
concrete would become one of Romanias primary building materials.
The form of residential development in Romania in the last five years has changed.
The construction of blocks of flats has given way to two and three-storey housing units.
However, the expertise of Romanian engineers is still reflected in the design and
construction of these units and the whole is the subject of an international research
project based in the UK.
This paper will look at the nature of housebuilding in Romania and make some
comparisons with the housebuilding process in the UK. It will identity aspects of building
control and associated legislative issues when considering the construction of foundations
in concrete.
There are some striking similarities between aspects of housing construction in the UK
and Romania. In the UK, there was a phase during the 1960s1970s where the demand
for social housing was high and the authorities deemed that a speedy and efficient method
of providing these was by the construction of blocks of flats. This strategy was the same
in Romania, but the timeframe was extended to include most of the 1980s too. A
difference in approach is conspicuous in that in Romania during this period, the vast
majority of dwellings were constructed of blocks of flats, whereas in the UK, there was a
private housing sector developing to satisfy the demands of individuals wishing to buy
their own property. In Romania before 1945 and the onset of a communist regime, there
was a housebuilding industry constructing dwellings typically of single and two-storey
traditional brick construction. Further, since the revolution in 1989, Romania has ceased
to build new blocks of flats in preference to contemporary, single and two-storey housing
in a detached, semi-detached and terrace form. In this way, therefore, a comparison
between UK and Romanian residential development is both timely and relevant.
Foundations for low rise buildings in Romania and UK: Comparisons and effects 117
In the UK, the subject of foundations for low-rise residential development in shrinkable
clay soils has been the subject of much research and debate for many years. The debate
has often centred on the depth at which foundations should be placed in shrinkable clay
soils. Foundation depth is important because of the effects that seasonal changes can
make to expansive clays. This is exacerbated in the UK when there are periods of severe
drought as have occurred in 1976, 1984, 1989, 1991 and 1995. Trees too can significantly
affect shrinkable clays. In the UK, the failure to construct house foundations at a suitable
depth has meant that there have been numerous foundations which do not support the
superstructure properly because they have been disrupted, cracked, or been subjected to
undue stress for which they were not designed. In essence, therefore, these foundations
have failed and underpinning is necessary to rectify this damage. A simple statistic is that
in 1991, the value of insurance claims for such remedial works was 540 million. It is
unnecessary to expound upon the typical details of UK foundations in shrinkable clay
soils, apart from plain (unreinforced) concrete is often used.
In contrast to the way UK foundations are constructed, the construction adopted in
Romania for two-storey houses on shrinkable clay is rather different. Firstly, in all cases,
foundations for every building are purpose designed. The minimum foundation depths
used varies depending upon the water level. The minimum depth of foundation is 1.5
metres, although this is in some instances increased to 2.0 metres. In addition, all the
foundations are constructed of reinforced concrete. Whilst such specifications are not
particularly complex, it is important that this difference in construction is noted. The
subject of legal and quality controls, too, is important.
In both the UK and Romania, legal requirements and quality procedures exist to ensure
concrete foundations meet the agreed standards.
In the UK, housebuilders normally construct houses which, upon satisfactory
completion, a ten year warranty is provided to housebuyers against major structural
defects. To satisfy the requirement for the granting of a warranty, housebuilders agree to
build in accordance with the Building Regulation [1] and, for example, the Standards [2]
of the National Housebuilding Council (NHBC). Together, these requirements specify
steps to be taken for constructing foundations in clay soils, or other difficult ground In
some instances, it will be necessary to undertake extensive ground investigations and
submit for checking proposals for foundations in such ground. During the construction
stage, inspections will be undertaken by field inspectors employed by the NHBC and/or
Appropriate concrete technology 118
by the local authority. Once the work is approved, the housebuilder may proceed with
subsequent construction work. If the work is not satisfactory, the housebuilder will be
required to rectify the work before proceeding. Although materials could be tested by the
inspectors, this is seldom done. If the ground conditions are different at the construction
stage than anticipated at the design stage, it will be necessary for a new foundation design
to be produced. In practice, however, many of the strip foundations for two-storey houses
are of plain unreinforced concrete. The depth at which such foundations are placed,
usually 0.81.0 metres, is determined on-site following a field inspection of the ground.
Where clay soils exist, the depth will generally be not less than 1.0 metre. This is in
marked contrast to the Romanian requirements. The width of foundations are determined
by normal design code requirements, but a compressible material or void former would
be placed on the vertical side of a foundation as a precaution against heave.
The UK regulations lay emphasis upon the appointment of an engineer to advise on
technical matters where complex ground conditions are anticipated. Some guides and
codes are written in non-technical terms to assist the housebuilder to identify poor
ground, whereupon he will be responsible for taking the appropriate action. If a
housebuilder fails to comply with specific requirements, the ten year warranty could be
withheld. In the case of local authority control, the officers have the power to serve a
notice to expose any building work to ensure compliance. However, where an authority
serves such a notice, the housebuilder is entitled to appoint an independent expert. When
an action is taken through the courts, a housebuilder can be fined for the error and have
imposed an additional fine based at a daily rate on the number of days the defect remains.
In practice, whilst this power through the courts exists for local authorities, it is seldom
pursued. There are two reasons for this. Firstly, each local authority has its own policy
and most prefer to determine solutions to such problems by negotiation rather than by
legal means. Secondly, and significantly, if a local authority undertake to expose building
works in the belief that defects exist and subsequently find that the work is not defective,
then it is responsible for all reinstatement costs.
In Romania, the procedure for checking the concrete in foundations are clearly stated
in specific codes and requirements on the use of materials and components, and in the
Romania Law of Quality [3]. It sets out specific duties, responsibilities and liabilities of
individuals and agencies associated with the construction of any building. These
responsibilities fall upon investors, designer/engineer, contractor, building owner,
building administrators and building users.
When considering a building, an independent chartered design engineer must certify
that the proposal has been designed properly. During construction, the concrete must be
tested at different stages. The contractor must undertake concrete tests. The supplier, too,
must certify concrete strength in accordance with the design. Site tests are required upon
receipt of the concrete and at the point of pouring the element (eg the foundation).
Independent and separate external quality checks may also be undertaken by appointed
inspectors under contract to the government. These inspectors may carry out any test on
the concrete and they also have the authority to check the validity of all the tests executed
by others.
The penalties of non-compliance with the law are specified in the law and these are
dependent upon the contraventions and the consequences of the non-compliance. The
penalties range from small fines for minor contravention to large fines and a term of
Foundations for low rise buildings in Romania and UK: Comparisons and effects 119
A comparison between the UK and Romanian causes of error based upon currently
available data is difficult. Information gained from two sources [4,5], one Romanian and
one from the UK, provides an unrefined comparison between similar sources of error as
shown in Table 1. The Romanian data is based on a survey of 840 cases of building
defects, whereas the UK study is based on a survey of 360 cases of residential properties
known to have inherent structural defects.
Table 1. Comparison of Defect Error Sources.
Defect Source Romanian UK
% %
Design mistakes 16.6 59.2
Construction mistakes and material defects 56.7 6.8
Error in building use 14.4 15.2
External factors 12.3 18.9
An analysis of the data reveals that design mistakes are considerably more prevalent in
the UK study than in the Romanian study. This result demonstrates the apparent detailed
attention paid to design by Romanian engineers compared with UK designers.
Conversely, the data suggests that the execution of the works is more problematic in
Romania than in the UK. One of the major problems in Romania is the lack of high
quality materials, especially for finishes and insulation. The net effect of this is that,
whilst there is adequate and effective supervision on site, the failure often stems from a
premature deterioration of low quality materials and components. Another possible
reason is that it is only on major building projects in Romania that modern sophisticated
plant and equipment is available. In the UK, high quality materials and components are
readily available. These coupled with competent supervision and the wide use of plant
and equipment, is likely to be a reason for lower construction mistakes and material
failure. Errors in use and external factors are similar in each of the survey results. It
Appropriate concrete technology 120
would be unwise at this stage to draw any detailed conclusions from the Nottingham
research project. However, it is worthy of note that the design/construction aspects of
projects produce different outcomes with respect to damage/defects sources.
CONCLUSIONS
The type of housebuilding in Romania and the UK is outlined, together with the problems
of building on clay soils. The design requirements and the control procedures in Romania
are arguably more demanding than in the UK and the design errors from the table support
this view. The construction aspects are difficult at this stage to quantify, but there appears
to be a problem gaining access to, and the use of, high quality building materials and
components in Romania.
In both the UK and Romania, legal and quality procedures exist. The UK model has
been in place for many years and has evolved over time. The Romanian model and the
recently introduced law cannot yet be fully evaluated because insufficient time has
elapsed to judge how and if the control procedures are effective. It can, nevertheless, be
stated that the sanctions available to aggrieved clients and the government are more
onerous in Romania than in the UK.
This paper has not sought to provide a detail of all controls and differences in each of
the countries, but merely to highlight that differences do exist. It is easy for one particular
country to believe that it has the better construction industry compared to another. What
is important is for building professionals to realise that each can benefit from knowing
and understanding how others operate. Only then will this more catholic approach to
knowledge acquisition be beneficial and improvement realised.
Concrete is a universal construction material and Romania and the UK have a wealth
of experience in its design and its use. The sharing of ideas, design techniques and
construction strategies will help to ensure that defects to foundations will be minimised.
It is clear from the evidence that no one country has total knowledge of the subject, but it
might be possible if construction professionals realised that through dialogue and
interaction, improvements can be achieved which can only benefit society at large.
REFERENCES
ABSTRACT. The paper reviews the aspects of suitable training for site
concrete workers. Based on a combined research and construction
experience, a series of example training aids using models and special
graphics are presented. The current development of an integrated
secondary to tertiary education for building trades in New Zealand is
outlined.
INTRODUCTION
The focus of research and development for concrete continues in the direction of
analytical research and design data with an ever increasing accuracy of test performance
to demonstrate the improved concrete properties. While this research growth to evaluate
and improve the performance of concrete structures is a pre-requisite for development in
the 21st Century, how much effort is focused upon the education of the site worker?
Most of these recent developments result in new concretes for the specifier. Does
the site worker recognise this new concrete? Does anyone on site know what concrete
with a Dc of 1013 means? Does anyone actually care? Does anyone know that
workmanship matters can change the Dc of 1013 to Dc10? Many developments will
never achieve their researched potential without some radical upgrading of construction
practice education. Over the past 20 years the NZ Concrete Research Association, now
Cement & Concrete Association of NZ, has had a hands on approach in relation to the
research it carried out because it also had direct access to site workers through the
provision of its construction education programme.
Appropriate concrete technology 122
The key issues are practical training techniques that capture the attention of the site
worker, so that although they may never understand the units of Dc 1013, they do know
the consequences of their actions in terms that can be understood.
During the last 70s and early 80s, matters relating to durability were under serious
discussion. The NZCRA, like many other research agencies, carried out a series of
durability studies. These studies were invariably limited in terms of any statistical
approach, but were designed to test overseas findings. During this period, water
permeability followed by chloride diffusion permeability were test methods used to
evaluate the concrete properties for durability.
In the testing regimes adopted, a component was cast in concrete using standard
formwork and compaction equipment on specimens obtained by coring. This sort of
approach produced a greater variability of the results compared to concrete cylinder
sampling. Different curing regimes clearly showed through on the results. While these
results are no different to many other researchers, it became clear that workmanship on
the construction site was a considerable parameter influencing the real performance of
concrete in the field.
The 1990 attendance at the Dundee Conference was made with a view to considering
the international preferences for the measurement of permeability. In reality the worry of
the size of the workmanship influence became more important than the concerns of
methods of property measurement. Accordingly, the challenge became one of trying to
influence site workmanship to meet both past and present and new requirements for
concrete construction. The challenge was made significantly more difficult certainly in
New Zealand, by the significant loss of independent site supervision, responsibility losses
by sub contracting and ironically by the efficiency of the local ready mixed concrete
industry. In this latter case the skills of understanding concrete became more and more
focused away from the construction site, as the skills for concrete production were
removed from the site worker. In recent days the ready mixed concrete industry has
realised the difficulties that can arise from having its product abused by construction
placement and is actively co-operating in the new training programmes.
In the early 70s a Concrete Construction course was offered by Technical Institutes of
the day, but after several consecutive years of no students, the NZ Vocational Training
Board abolished the course leaving only isolated topics within a number of courses, e.g.
Civil Engineering, Carpentry, Bricklaying etc. None of these courses were suitable for
the site worker. In an effort to redress the balance the correspondence course developed
by the then C&CA (UK) for the London City & Guilds Concrete Technology and
Construction was introduced into New Zealand run by the Concrete Research
Association. Even at this time the course continues to be run. The mainstay of technical
adult education, as opposed to tertiary education, was undertaken through 1 or 2 day
training courses. 500600 people a year were trained in this manner in the early 80s.
Who were the people being trained? A significant proportion were clerks of works and
works foremen for the principal companies. As has already been mentioned the clerk of
works influence has been significantly eroded, meaning that one avenue of the transfer of
Site workers-the challenge 123
knowledge to the site worker has gone. Many companies choose to subcontract the
concrete placement and that transfer of responsibility causes another avenue of
instruction to be lost.
The New Zealand Government has embarked upon an ambitious integrated
programme which allows industries to develop training modules to suit its particular
needs. The training may embrace school, polytechnic and University activity. The
development of such a co-ordinated programme is proving to be a difficult task, not made
any easier by the use of educationalists to write the training achievement requirements.
However, out of this will come an opportunity to provide site workers with some
achievement stepping stones. What has to be guarded against is setting unrealistic
academic achievements in relation to practical skills achievement, i.e. we do not require a
treatise on how to use an internal poker vibrator, we want to have demonstrated the
practical use of the equipment. An overall industry strategy was considered in advance of
the education/vocational training schemes being pursued by Government agencies. It is
illustrated in Table 1.
Table 1 Draft Strategy for Training
Advanced Research
National Certificate Level
3 High High Special Special
2 Mid Mid Mid Mid
1 Basic Basic Basic Basic
Concrete Construction Concrete Testing
Technology Technology Production
A B C D
It is set up in such a way that after some common core topics, there are a number of paths
to suit the avenues of specialisation with the industry to achieve a National Certificate.
However the real concern is not to over emphasise these peoples training at the
expense of the site worker. Hence the teaching/skill achievements etc. in Level 1 are
geared entirely for the concrete worker. The training given at this Level 1 is to be linked
into the system which will allow personal development through the other training
levels, but there is no compulsion element to proceed further. The skills achieved in
Level 1 topics are formally credited.
Each topic is packaged in a presentation of 1 hour which will be discussed later. Special
extension topics can be added to suit particular industry needs. Table 3 shows some of the
extension topics which are also treated as a 1 hour presentation. Many of these topics will
have Level 2 and Level 3 components.
Site workers-the challenge 125
Training Techniques
Up until the late 80s all our training was addressed to people who were substantially
literate. At that time a major New Zealand contractor who had had many of its foremen
trained by the 1 and 2 day courses, asked for a training programme for its site workers
not just the foremen.
This brought about a significant change in approach. Previously it was reasonably easy
to justify why something should be done in a technical sense and follow this with an
overview of how to do the work. The challenge was now to explain the WHY in terms
acceptable to the worker as well as HOW in real terms. This has to be done, not in the
classroom environment, but in the site hut or on-site. It has to be done not totally relying
on the written word, but principally based on a visual and participating approach. The
training video, while useful to reinforce a point, remains impersonal and often too small
scale for visual impact.
Accordingly each topic was treated in a manner that had a mixture of hands-on, model
demonstration and overhead projection animation to fix the WHY in peoples minds.
When some one knows WHY they are doing their job, it is much more likely they do it
without the need for constant supervision of the HOW.
It is obviously not possible within the confines of this paper to cover all the topics, but
a selection of model and overhead projection animation examples are given to illustrate
the techniques used to get the messages accepted within a short time span of 1 hour for
each topic.
In the presentation it is particularly important to emphasise that each action has a real
impact in the final performance or life of the concrete member.
Some of the results of the approach can be gauged from contractor observations as to
what took place after their workers attended the on-site one hour training on the different
topics.
1. High quality finished required on concrete columns: Concrete gang refused to put the
concrete into the formwork because they would see the formwork was going to leak.
The formwork was rebuilt and a job requiring no remedial touch ups was
achieved.
2. Measureable improvement in productivity and quality with a reduction in overseeing
management, i.e. because each worker knew WHY the job he was doing was needed,
there was less need for nagging from others to do the job properly. There was much
less remedial work.
3. Even curing got done! Very important when not doing it can change a Dc1013 to
Dc1010.
Site workers-the challenge 129
SCHOOL/INDUSTRY TRANSITION
The danger of condemning people early in their lives is very real. Often entry to any
form of tertiary education is bedeviled by a long list of academic requirements. This can
often debar people with essentially practical skills from even entering or taking advantage
of tertiary education. The result is that no further training takes place.
Polytechnic courses are often more related to academic skills before getting to grips
with the practical hands on skills.
A pilot scheme is being developed by one New Zealand Polytechnic which permits
entry from school into a pre-employment Building Trades Course. Students can join the
course at different times during the year, which allows for easier transition from
secondary education to tertiary. School leaving ages are generally being raised, but unless
meaningful education is provided, often the extra year becomes a mindless waste of
time, particularly for those people whose future paths lie in practical skills.
Entry to traditional apprentice trades is also becoming shackled with academic success
requirements. The pre-employment general building skills course also provides an
alternative bridge to reached practical skills equivalent to the former apprenticeship
approach. A student can, after attendance at the pre-employment course which covers a
wide range of topics, choose entry into the more specialised areas of construction.
However the real challenge has still to come, because unlike specific apprenticeships
of the past, these students will be Multi Skilled to a significant level in many trades.
The outline of the developing programme is shown in Table 4.
CONCLUSIONS
1. It is important to recognise that workmanship skills are a vital part of concrete
performance chain.
2. Simple training properly addressed to the workers involved can significantly influence
workmanship standards.
3. With research introducing new applications of materials, there is as much need for
providing appropriate instruction at worker level as there is at specifier, designer or
concrete producer levels.
Appropriate concrete technology 130
REFERENCES
1. Concrete on site. Vol. 111. Crowthorne, U.K.: British Cement Association, 1993.
2. Site concreting. Vol. 1 . Porirua, N.Z. : Cement & Concrete Association of New Zealand, 1994
Site workers-the challenge 131
Dr John B Newman is Head of the Concrete Structures Section and Director of CRIC at
Imperial College, London. He has supervised many programmes of research including
investigations of concrete durability, structural behaviour and the development of special
concretes and components. As a consultant and expert witness he is primarily concerned
with the development, acceptance and testing of construction materials and components
and investigations into the integrity, repair and maintenance of buildings and structures.
INTRODUCTION
The correct specification of concrete is a time-consuming and often repetitive task for all,
even the specialist. Recent amendments to cement standards to harmonise with
Euronorms have complicated the topic further for practising engineers. Later this decade,
BS EN 206 : Concreteperformance, production and compliance criteria will be
released as a national standard (anticipated to be released in early 1998) and many
concrete specifiers will find themselves tasked with responsibility for completely re-
drafting their specifications without possessing sufficient knowledge of the new standards
for concreting materials and accepted best practice. The objective of this project is to
develop an effective method of concrete specification covering current and future
standards and regulations using state-of-the-art computing techniques.
In operation, the specification system will be designed to be easily used by practising
concrete specifiers (principally consulting engineers) and to be readily up-datable to
provide for the introduction of future amendments to both UK and European standards
and Regulations. The result of the project is intended to alleviate the burden on concrete
specifiers whilst enabling the efficient production of correct, concise and appropriate
concrete specifications. Benefits in quality assurance will be attained by ensuring that the
current version of the appropriate Standard and Regulation will be used and that the
criteria for decision making will be logged to provide an audit trail.
The project is a collaborative activity that has attracted approximately 50% of its
funding from the Department of the Environment under their Partners in Technolog y
programme. The residual funding is provided by the Ready-mixed Concrete Bureau, the
project proposer and lead partner. Other partners supporting the project are Posford
Duvivier, Consulting Engineers, PD, together with the Concrete Research and Innovation
Centre at Imperial College of Science, Technology and Medicine, CRIC.
Development work commenced in mid-1995 with a planned completion date of
August 1997 to coincide with the projected agreement on a prEN 206 for CEN enquiry.
Consequently, at the time of writing much of the detailed aspects of the programs
operation are unavailable. However, a demonstration system in accordance with BS 5328
is intended to be available for the Congress.
Appropriate concrete technology 134
PROJECT OVERVIEW
Scope
The scope of this project is to produce a computer program to aid the production of
economic, concise and unambiguous concrete specifications in accordance with the
appropriate prevailing standards and regulations. Of particular importance to this project
is the inclusion of BS EN 206 : Concreteperformance, production and compliance
criteria in the program to promote its introduction to UK construction practice.
In addition to providing a means of specifying concrete, it is intended that the system
will also aid the dissemination of background information on key aspects of concrete
technology and practice; eg cement selection and behaviour, concrete mix design,
admixtures, workability/cohesion, placing methods and site quality control etc. This part
of the program will be designed to be accessible from two routes;
as part of an on-line help facility
as an interactive reference guide.
To maximise user acceptability and accessibility, the program will be developed for
deployment on a range of popular computer operating systems. For ease of use the
concrete specifier will communicate with the program through a graphical user interface,
GUI. The adoption of a GUI provides for such items as menu bars, pull down lists, check
boxes, text entry boxes as well as facilities for user feedback/confirmation with
illustrations and photographs.
Methodology
System design involves development in three distinct areas;
the knowledge baseusing information from British and European Standards and
Regulations (eg. BS 5328 1, EN 197 2, prEN 206 3, Specification for Highway Works
4
, etc.) communicating with the GUI as shown in Figure 1.
Computer-based concrete specification 135
Knowledge base
The underlying philosophy for the knowledge base is traditional set theory and pattern
matching. To facilitate development and to minimise demands on memory requirements,
the knowledge base will be produced in a modular manner, consisting of a number of
separate knowledge bases which will call each other as required. The knowledge base is
being built using a set of rules that operate on user-defined and system objects (variables,
strings, flags etc). Rules are developed in an If-Then-Else format and as separate
entities without a predefined processing route. This approach ensures that all applicable
rules are processed and facilitates rapid modification to accord with changes in the
Standards and Regulations.
The illustration in Figure 2 suggests that each knowledge base is plane, with distinct
boundaries interacting only in well defined areas where one overlaps the other, ie. the
relationship between hardened properties and the placing requirements for construction
providing constraints on the fresh properties of the concrete. However, the reality is
somewhat less clear. Whilst each knowledge domain is well defined, the interaction
between knowledge bases is largely dynamic and the knowledge base behaves more like
a group of fuzzy, amorphous masses with both strong and loose links.
selection of cements and aggregates for concrete exposed to industrial chemicals, high-
performance concretes and architectural finishes. The information used to develop this
part of the knowledge base will be based upon information that already has a proven
history of satisfactory use, with the lead reference being the ACI Manual of concrete
practice 6.
Once the most appropriate concrete has been selected for each element in a project, a
rationalization process will be evoked. This routine will look for similarities and
differences between concretes and reduce the number of different concretes used on a
project to an economic minimum. The aim of this routine is to improve construction
economy by providing flexibility for maximising the use of any given concrete.
User interface
A object-driven user interface will be provided, designed for rapid specification
production for both individual concrete elements and projects containing many elements.
A schematic of the interaction between the knowledge base and graphical user interface
is given in Figure 3.
In general, the user will not be required to directly input information such as sulfate class
and exposure conditions. It is intended that as much information of this nature as possible
will be collected using indirect methods. Adoption of this methodology is intended to
ensure accurate interpretation of all available data. It will also permit updates and new
releases of standards and regulations to be supported without the requirement of re-
engineering the program, new data files only will be required. Once administration data
has been collected, the first dialogue routine involves a determination of the type of
project/element being considered and will then suggest much of the environmental
Appropriate concrete technology 138
exposure conditions. For example, if the structure type is a road bridge in the UK, the
system will assume that concrete will be exposed to chloride attack from de-icing salts,
that it will be exposed to freeze/thaw action and that the Specification for Highway
Works 4 and BS 5400 7 will be the lead Regulation and Standard. Naturally, the user will
be offered the opportunity to confirm or amend the suggested data but this approach will
help to ensure that all items required to produce a correct specification will be provided,
consistent with the rules of the appropriate Standard/Regulation.
Similarly, where site investigation information is available, the user will be asked to
enter the details of the chemical analysis rather than just the sulfate class. This approach
has two benefits; it will help to avoid interpretation errors and will ensure that the
available information is interpreted in accordance with the current requirements.
Information relating to section size and envisaged placing method will be used as part of
the rationalization process as well as for checking such items as limits on the heat of
hydration of cement and suitability of aggregate size. Guidance on appropriateness of the
envisaged placing and finishing methods together with workability selection will be
given at this stage. Guidance information will be appended to the specification to be
confirmed or adapted by the purchaser of the concrete (contractor).
Reference guide
The reference guide will provide interactive on-line help and operate as a stand-alone
guide to the philosophy of concrete mix design, materials behaviour, concepts of
Standards and Regulations, specification and site practice together with other items.
Using text, diagrams and photographs the guide will contain background information on
Computer-based concrete specification 139
the concepts of concrete technology. Some topics that are difficult to describe using still
images, for example cohesion, will also be described using video. However, this facility
requires extensions to some operating systems that have yet to be developed and may not
be available on all platforms.
Anticipated benefits
It is envisaged that this system will be a potent method of facilitating the rapid
introduction of
emphasised by the inclusion of Table 2 and Table 2A which have also been extracted
from an actual tender document.
the advances in concrete technology described in UK national and European standards
to aid construction economies. Where adopted the program will enable more efficient use
of materials and will make more efficient use of specialist resources (engineer, producer
and contractor).
By reducing errors and ambiguity in concrete specifications it is predicted that the
number of specification related disputes will fall significantly, together with an associated
decrease in the amount of returned/waste concrete.
ACKNOWLEDGEMENTS
The authors would like to express their thanks for the support of the Construction
Sponsorship Directorate of the Department of the Environment and to Posford Duvivier,
whose involvement in this project is invaluable.
REFERENCES
Leader Paper
The paper lists and discusses the enormous variety of applications where concrete is
advantageously used. It is shown that in infrastructure applications all building materials
have concrete as a competitor but that the reverse is not true. Based on its competitive
advantages it is foreseen that concrete will continue to be the most versatile and used
building material of the next century. To keep this leadership some of its shortcomings
need to be addressed and alleviated.
Keywords: concrete versatility, tools, properties, applications, competitive materials
Dr. Folker H.Wittmann is professor for building materials at Swiss Federal Institute of
Technology, Zurich. Fracture mechanics of cementitious materials and durability of
concrete structures are his main research activities.
Dr. Juraj Gebauer is assistant vice president of Holderbank Management and
Consulting Ltd. He is specialised in the application of cement and concrete technology,
with particular interest in the effect of cement on concrete properties and high
performance concrete.
Dr. Roberto Torrent is head of the Concrete Technology Division of Holderbank
Management and Consulting Ltd. His main research interest includes concrete mix
design, high performance concrete, durability and permeability of concrete.
The versatility of concrete 143
INTRODUCTION
Concrete is the most widely used building material in the world today. We find concrete
in all kind of residential, public and industrial buildings, in underground structures, in all
transportation structures such as pavements, railroads, bridges and tunnels, in structures
for water and sewage treatment, in power stations, in offshore structures, dams, etc. Most
of the achievements of our modern civilisation depend on concrete, forming with steel the
basis for modern structural engineering.
The total world production of concrete in 1995 lies around 10,000 tons, i.e. about 2
tons for every human being. To put this figure in perspective, if all the concrete produced
in 1995 were used to build a conventional two-lane concrete pavement along the equator,
the road will turn about 50 times around the globe.
The reasons for the wide usage of concrete are not difficult to findthe raw materials
are in plentiful supply and available almost everywhere, requires low energy to be
produced, costs relatively little compared to other building materials, a wide range of
production and applications technologies is available and, last but not least, its great
versatility in applications covers a wide range of demanding performance requirements.
Concrete competes with all major building materialstimber, steel, and other metals,
asphalt, rock, plastic, glass, bricks, ceramicsbecause of its versatility in application.
The object of this presentation is to highlight the versatility of concrete. According to
the definition, versatile means interested in and clever at many things. We will delve
into the tools at the disposal of the concrete technologist to achieve that versatility and
the results that have so far been achieved in terms of properties and applications.
Essentially, we will discuss the following questions:
We will consider practically all concrete types and applications, whereby under concrete
we understand a composite material that consists of a binding medium within which are
embedded particles or fragments of aggregate (ACI-definition). The binding medium
can be Portland cement, blended cement, cement mixes, mineral admixtures and chemical
admixtures, special cements and water. It is evident that this rather broad definition
covers a very wide range of materials and applications.
VERSATILITY TOOLS
One of the main tasks of a concrete technologist is to design concrete mixes suitable for
an intended use. In fact, before designing the mix, he has to define the most convenient
process to build the structure and specify the requirements of the mix accordingly.
Once the requirements of the concrete have been specified, the concrete technologist
has to engineer the characteristics of the composite to achieve the expected performance
at the lowest possible cost.
Appropriate concrete technology 144
He has a limited, yet powerful, tool box to engineer the properties of concrete. He can
act basically on four fronts:
Rheology of fresh concrete,
Chemical processes,
Pore structure of the composite,
Fracture mechanics (especially when fibres are added).
Let us look at them in some more detail.
By making good use of those tools we can produce concretes that, on the one hand,
behave almost like liquids, i.e. they flow unaided, filling each and every available empty
space in the forms, without segregation. On the other hand, we are able to produce
concretes that can be demoulded immediately after casting (e.g. precast concrete blocks)
or that are even able to withstand heavy loads in their fresh state (roller compacted
concrete). Depending on its rheology, concrete can be transported by conveyor belts (up
or down hill), flow freely through chutes or tremies, be pumped over long distances and
heights or be projected through a nozzle (shotcrete).
Chemical Processes
The main chemical reactions that take place in a concrete system are those associated
with the hydration of Portland cement; they start as soon as cement is put in contact with
water and continue for years under suitable hydro-thermal conditions.
Both the intensity and the kinetics of the chemical reactions taking place within the
concrete system can be controlled. The tools available are:
cement composition,
mineral additions added to the cement or into the concrete mixer,
fineness of the cement,
cement content of the mix,
use of chemical admixtures, mainly retarders and accelerators,
thermal treatment.
By proper use of these tools it is possible to achieve concretes that set almost
instantaneously to concretes that take hours and even days to set as is the case of the
long-life mortars. Also concretes that gain strength very rapidly like high-early strength
cement concrete or steam-cured concrete or very slowly, e.g. blast furnace slag cement
concrete. Concretes that develop a high or low amount of heat of hydration can be
designed, e.g. for winter concreting and mass concrete, respectively.
By adding certain chemicals it is possible to generate and entrap gases in the mass
(frost resistant concrete and aerated concrete) or to promote certain controlled expansive
reactions to compensate for the drying shrinkage (expansive cements).
Pore Structure
Concrete is a porous material containing a system of pores covering a wide range of pore
sizes.
Pores can be found within the aggregate particles, but also between the particles if the
intergranular voids are not completely filled with cement paste.
The paste itself has a high porosity and at least three different kinds of pores can be
found:
Gel pores, in a size not much larger than a few times the size of a water molecule, i.e. 1
to 10 nm. These pores are responsible for most of the volume changes of concrete
associated with gain or loss of water (drying shrinkage, swelling and part of the
Appropriate concrete technology 146
creep). They are a structural feature of the hydration products and, as such, cannot be
eliminated.
Capillary pores, resulting from the space originally occupied by the water of the mix
that has not been filled with hydration products. For good quality concrete the size of
these pores is in the range 1050 nm. For low quality concrete it can reach 10,000 nm
(i.e. 10 m).
Air micro-bubbles, originating from the use of an air entraining or foaming agent.
These bubbles are typically in the size range of 10100 m.
The main tools to modify the pore structure of concrete are:
water/cement ratio (volume and size of capillary pores)
cement content (volume of gel and capillary pores)
use of micro-fillers (e.g. silica fume)
use of air entrainers (volume and size of micro-bubbles)
use of porous aggregates
use of special grading of aggregates (drain concrete).
By proper engineering of the pore structure of concrete we can modify at will (naturally
between certain limits) properties such as density, strength, modulus of elasticity, thermal
insulation, permeability, radiation absorption, etc.
Fracture Mechanics
One of the characteristics of concrete more difficult to modify significantly is one of its
inherent shortcomings: its brittleness or low fracture energy and its low tensile strength.
A remedy to overcome this limitation has not yet been found within the traditional
range of component materials of concrete. However, one way at least to palliate this
problem has been found through the use of fibres as a further concrete component.
Properly designed concrete containing the right volume of the right type of fibres can
increase its ductility dramatically. These improvements have not been sufficient to avoid
the use of steel reinforcement for structural concrete.
Nevertheless, this is a field where developments are expected in the future,
particularly as fibres of more sophisticated materials become available at relatively lower
prices. For special applications this type of high-technology concrete is produced already
today.
Table I presents the range of properties that can be achieved by a clever use of the tools
available, as discussed in the previous Sections.
Figs. 2 represents the range of properties achievable for different concretes compared
with those corresponding to several materials used in construction (construction steel,
aluminium, timber, plastics and glass).
The versatility of concrete 147
It can be seen that concrete is the most versatile material of the six in terms of density
(together with plastics), E-Modulus and thermal conductivity. Its limited performance in
terms of tensile strength (when unreinforced) is evident as well.
In the previous section we have discussed the wide range of properties that can be
achieved by proper use of the tools at our disposal.
By using the know-how available we are able to produce a complete palette of
products designed to suit specific needs of the construction industry. The extensive usage
of the great number of concrete types in the various constructions using a variety of
production methods is shown in Table II, illustrating the extraordinary versatility of
concrete. None of the competitive building materials reaches such a wide usage in man-
made structures.
It is of interest to show the building materials which are competitive to concrete in the
different fields of application. Table III presents the situation for some typical cases.
Table I Typical values of important properties for
different types of concrete
Property Typical Values Concrete Type
200800 Lightweight (Insulating)
8001400 Lightweight (Insulating and Structural)
3
Density (kg/m ) 14002000 Lightweight (Structural)
20002600 Normal weight
26005000 Heavyweight (Radiation shield)
0.52.0 Flowing Fills
Compressive Strength (MPa) 0.430 Lightweight
1560 Normal Strength
60150 High-Strength
Modulus of Elasticity (GPa) 215 Lightweight
1535 Normal Strength
3050 High-Strength
Thermal Conductivity (W/m.K) 0.11.0 Lightweight
1.02.5 Normal weight
<0.1 High-Strength (>65 MPa)
0.10.5 Good Quality
16 2
Permeability to Air (10 m) 0.52.5 Medium Quality
Appropriate concrete technology 148
Coloured concrete
Structure types
Buildings (residential, public, industrial, slabs, columns, decks, floors, partition walls, fundaments,
silos, pools)
Transportation structures (rail or roads, roads, tunnels, bridges, pipelines)
Sanitary structures (water and sewage treatment)
Offshore structures
Military structures
Hydro structures (dams. water ducts, embankments)
Power stations
Hazardous waste containments
Table IV presents some of the many different types of concrete, the tools used to achieve
their specific characteristics required and various applications.
Appropriate concrete technology 150
mortars
Pumped concrete 1, 3 Workability and cohesion suitable for All kind of concrete
pumping, retarders to delay setting constructions
Precast concrete 1, 2, 3 Normally stiff mixes, fast setting and Concrete elements, Pre-
high early strength. Usually thermal stressed bridges
curing
Concrete for 2, 3 Low permeability (low w/c ratio), Structures exposed to
aggressive components that are resistant to the aggressive media
environments aggressive media, surface treatments
Fibre concrete 4 Fibres (steel, synthetic, glass) are Industrial floors,
added to improve the cracking impermeability, impact,
resistance, durability explosions
High 4 Low w/c ratio, silica fume or other fine High performance structural
performance silicas, special aggregates, fibres members, wear parts,
concrete machine parts
(*) 1 : Rheology of fresh concrete, 2: Chemical Processes, 3: Pore Structure, 4: Fracture Mechanics
Two extreme examples should illustrate and complement the information given in table
IV.:
Flowable Fill
This is a material intended to be used to backfill excavations (for example to fill the gaps
between the excavated soil and pipelines, etc.). Traditionally this have been done by
placing and compacting granular materials in layers. In many cases the material was
dumped full depth into the trench, never tamped and therefore not adequately compacted.
The idea was to replace that traditional technique with another, allowing fast filling of
all voids, easy re-excavation and fast re-opening of the road.
The properties required are:
self-levelling (i.e. no compaction should be required),
no shrinkage and
very low strength (sufficient to support the construction of a pavement on top but low
enough to ensure easy re-excavation when required).
The recipe of this kind of material varies, but as a guideline the following composition
can do the job.
Portland Cement 25 kg/m3
Water 225 kg/m3
Sand 1105 kg/m3
14 mm Stone 560 kg/m3
20 mm Stone 387 kg/m3
Appropriate concrete technology 152
Supplementary cementitious materials such as fly ash, slag, etc. can also be used.
Strengths not exceeding 0.5 MPa at 28 days are achieved, mostly through expulsion of
mixing water (cement acts mainly as a lubricant) and self-compacting. The low amount
of cement guarantees the non-shrinking behaviour.
The recent developments in the area of ultra high strength cementitious systems al-lows
to produce materials having properties approaching those of steel and other metals or
ceramics. One example are the macro defect free materials (MDF).
The Reactive Powder Concrete using a special concrete mix design with Portland
cement, silica fume or other fine silicas, superplasticiser, well graded fine sand and steel
fibres can be used for high performance structural membranes. The Reactive Powder
The versatility of concrete 153
Concrete exhibits ultra high strength up to 800 MPa and high ductility at the same time.
Another high performance material used in security products (safes), wear components,
machine parts and other structural application is Densit, based on Portland cement, silica
fume and superplasticiser.
FINAL REMARKS
At the end of this tour, that took us around the present of concrete as the most versatile
and widely used building material of modern times, we should have a look into the
future. Will concrete still have such a predominant role in the future? Is this
predominance assured for the years to come?
The answer is difficult. History proves that no single human development remains
predominant forever and, due to depletion of resources coupled to technological
development, it is possible that concrete will be superseded in the long-term by another
material or other materials which could possibly be unknown today.
However, the raw material reserves for making concrete are assured for many years to
come and it is unlikely that some other material will take over in the foreseeable medium
to long-term.
Moreover the full potential of concrete as a structural material has not yet be realised.
The obvious drawbacks of concrete such as low ductility, low tensile strength, limited
durability in certain situations, poor aesthetic and environmental image constitute a
challenge for the concrete specialists to find solutions and so even increase the share of
concrete on the building materials market. In particular the further reduction of CO2-
emission and energy consumption in the production of cement and concrete, recycling of
industrial by-products and aggregates form demolished concrete shall help to make
concrete more competitive and versatile. Overcoming the apparent contradiction between
the low cost image and the high performance potential is another challenge for the
concrete community.
Normal concrete will continue to be the major building material for quite some while.
High technology concretes, as they can be produced today need much more stringent
quality control and their application will always be based on the development of more
sophisticated engineering projects.
Concrete is even considered as the possible material for construction in the outer space
for the same reasons it has found such acceptance on earth.
We can confidently state that none of us here today will witness during our lives the
replacement of concrete as the most versatile building material.
CONCRETE MASONRY: A DIFFERENT
DIMENSION FOR HOUSING IN
DEVELOPING COUNTRIES. THE 40
YEARS EXPERIENCE IN COLOMBIA
G G Madrid
Columbian Cement Institute
L G Pelaez
Cementos Nare
Colombia
INTRODUCTION
Governmental low-cost housing programs, started in Colombia during the late 50s,
became of a capital importance during the early 80s, due to the persisting immigration to
cities from rural areas. This phenomenon, originated in the political violence of the early
50s and in the historical social imbalance, made Colombia a very urban country (the
ratio of urban to rural population (in %) changed from 43/57 in 1951 to 65/35 in 1985
[2]), even though the country has been agriculture dependent (coffee, flowers, bananas)
until recently; so most of the information given in this paper deals with urban housing
projects.
CB is the second wall material in Colombia after fired-clay units (bricks and blocks).
There is no industrialized construction of wooden houses, and there is almost no
production of panels for partitions, just some fiber-cement and concrete panels for farm
houses or cabins (people consider a house to be strong only if it is made with masonry or
thick walls). The use of earth walls is almost forgotten, except in distant, underdeveloped
areas or for landscape purposes. Steel is just starting to appear in frames for buildings
(there is the production of reinforcing steel, small structural sections and laminated-steel
roofing). Clay is not available everywhere, or with quality enough to produce units for
load-bearing and reinforced masonry, like in Colombias Atlantic coast, where CM is
almost the only possible or available material for walls and partitions.
In most of the country (specially in large cities), competition is strong among clay and
concrete; but clay masonry is more widely known and accepted by construction workers
and users, since it is easy to handle (light weight, easy to cut) and very convenient to
produce a good result with minimum technical knowledge. It makes that, unless for
economic reasons (and recently due to the reinforcing requirements to attend
earthquakes), clay masonry be the first-choice for new houses or for additions to old
ones. In rural areas, houses are built with the available material; but a point in favor or
CM is the possibility to be produced in small-scale plants, something technically
impossible for fired-clay units. This paper presents the scenario of low-cost CM in
Colombia under its main physical, technological, economic and social characteristics.
Appropriate concrete technology 156
PHYSICAL CHARACTERISTICS
Typologies
Solid walls are provided when the house is sold. Double lined walls are suggested for
future expansion of the house. Project H is the start of a rural house with open space
around.
In Project E to G, E is the one-floor house as sold; F is the 1st. floor of the two-floor
house or the modification when expanded; G is the 2nd. floor of the two-floor house or
Concrete masonry: a different dimension for housing in developing countries 157
the modification when expanded. Urban projects only have a backyard as area for
expansion (projects A, D).
Even though it is necessary to get a permit from the city planning office for any
modification, the requirements are often violated, with an intricate mix of materials and
structural types. It could be found that over CM, clay masonry or concrete walls first
floors, second and third floors are made with concrete or clay masonry, or even with a
lightweight panel house, creating a dangerous situation in seismic regions, in steep-
sloped terrain and in technically poor environments. These violations and the non-
reported modifications are such that the yearly figure for the non-registered area for new
or remodeled construction is between 35 and 100% of the registered one, depending of
the controls by local authorities, with an average around 55% [3].
This is a delicate topic from the point of view of planning and safety. Responsible
promoters design according with planning regulations and make it clear to potential
owners, the restrictions for future expansions. But, once the houses are sold, those
requirements are forgotten, and after ten years, it is difficult to recognize the original
houses in a block (Figures 2, 3). It is advisable to be preventive, studying the local
traditions in order to design with enough structural capacity to be used in the future, since
planning offices are not capable to control every construction.
As an example, some of the planning regulations, in Medelln, for one, two and three
dwelling low-cost housing are: Minimum areas for the lot, 54, 60 and 72 m2, respectively,
with frontal lot widths of 4,5 m for the first one and 6 m for the second and third ones.
The minimum embryonic unit is 18 m2, and should have one multiple space (living-
dining-bedroom) a kitchenette and a bathroom (Figures 1A, 7). The minimum regular
unit is the embryonic one with one or two bedrooms. The minimum area possible to be
built must be 40 m2 (for second or third-floor projects, a finished slab must be provided
additionally to the embryonic area [4]).
Appropriate concrete technology 158
Complete houses or apartments have united living and dining rooms, connected to the
kitchen; a laundry area (exterior but covered in patios or terraces); one to three bedrooms
and one or two bathrooms (sink, toilet and shower). First-floor houses always have a
patio. Second and third floor houses have a terraces. One dwelling three-floor houses are
sold with one unfinished floor (bare slab and walls) or with the possibility for the
construction of one or two rooms (Figure 1).
Two indexes are used to evaluate the size of inner spaces and the space occupied by
walls [5], and to calculate the consumption of masonry. For houses the average are: 0,598
m/m2 (wall length/habitable (net) area) and 0,077 m2/m2 (area occupied by walls/net
area). The second index divided by the first one gives the average wall thickness, in this
case 130 mm. Values have been found from 0,512 to 0,666 m/m2 for the first index and
from 0,059 to 0,092 m2/m2 for the second one, with average wall thickness from 115 to
150 mm.
Rural Houses
It is uncommon to find rural-housing programs, except for reconstruction after natural
disasters. Those programs, executed by responsible, mostly non-governmental
institutions (NGOs), tend to be more complete, in services, than those developed for
commercial purpose in urban areas. A church, a school, a soccer field or a concrete
multiple-sport slab, a health center, a communal building and some workshops or
commercial spaces are often provided. It is intended that typologies do not deviate form
traditional ones, but have better services (Figures 1H, 5).
Urban Design
Regulations for urban design (area/dwelling, densities, distance to streets, constructions
and streams, road or walking-path design) differ a lot from city to city. In Medelln, for
low-cost projects, the main road section (bus route and emergency access) is 10 m (a 7 m
pavement, two 1,5 m sidewalks), non including an optional frontal green area. Most of
the houses are built in front of 6 m wide walking paths, with a 2 m central sidewalk and
two 2 m green areas; or two 1 m sidewalks and one 4 m wide green area [4] (Figure 8).
The walking paths should not be more than 150 m long when connecting two access
roads or a road and a green area (at least 500 m2). In both cases they should be cut by
another walking path to reduce the block length to 75 m. Walking paths and sidewalks
are a continues surface for slopes up to a 16 % or should have steps if slope is higher [4].
The slope is limited just by feasibility (Figure 13).
Appropriate concrete technology 162
TECHNICAL CHARACTERISTICS
Several factors affect the technical characteristics of low-cost houses and buildings,
generating a considerable amount of changes from original fast-thought designs:
Block Types
Most of low-cost CM housing is built with standard 400200 mm blocks, half and U
units, very seldom using architectural ones. The Structural Code allows 125 mm wide
blocks for load bearing purposes; 150 mm ones are common; 100 mm units are used for
partitions. Something new is the use of 250 mm tall units; preliminary results indicate
that there is an improvement in strength of the masonry prisms (Figure 10). This change
has gained acceptance since: There is no additional effort during production, reduce one
layer of mortar per meter in height and, since block laying is mostly paid by units, there
is a reduction of 20% in labor costs and some savings in mortars, more expensive than
blocks (Figures 10, 11, 13). Some contractors producing block on-site, including special
units like U, T, mid-height and tall blocks (Figure 14).
ECONOMIC CHARACTERISTICS
Costs Structure
Commercial construction
Official policies determined that all construction material for housing is excepted from
the 14% Added Value Tax (AVT), and there has been a subsidy for low-cost housing. In
1939 the Territorial Credit InstituteICT was created to build low to medium cost
housing. Because of an accumulative deficit due to unpaid quotas, bureaucracy and
complains for bad quality, ICT was liquidated and the INURBE began providing a direct
subsidy in 1992. The subsidy has three scales depending on the population (Table 1). The
top price of a dwelling must be US $ 12 590, 15 108 or 16 997 respectively, to be eligible
for the subsidy, being it equivalent to 100, 120 and 135 Minimum Monthly Salaries
MMS. The subsidy is given as a part of the down payment. The rest is paid in 15 or 20
years in Constant-Acquisitive-Power-UnitsUPACs. The units increase their value
Appropriate concrete technology 166
according to inflation. In most cases, subsidies are higher if given through a Social
Compensation Box, a system to benefit the lower-paid people, with founds coming from
a 4% of all the salaries.
For a builder of low-income CM housing, the cost structures is something like: Land
17%, construction 54%, financial 6%, indirect 13% (design honorariums, promotion and
sales, le-gal, registration and insurance), utility 10% for a retail price of some US $
225/m2, of which the under-structure and services costs are US $ 45/m2.
the labor on the site do not get full salaries and benefits. They also argue pollution and
tax disadvantages of on-site production, but it is equivalent to make concrete on-site,
instead to buy ready-mixed concrete. Arguments are endless. Contractors report to make
blocks for 80% the factor price and to have the additional savings of the hauling,
equivalent to near 22% of this amount; so the difference is near 30%. The factory prices
before AVT for 200400 mm blocks are: 0,31; 0,40; 0,48 and 0,63 US Dollars for 100,
120, 150 and 200 mm wide units. Those figures look attractive but beware of the hidden
costs contractor usually do not consider, the relative costs of materials, hauling, labor,
taxes, planning and environmental restrictions, trade policies, available on-site space,
adequate down-scaled and cheap equipment and molds, etc.
CONCLUSIONS
The Colombian experience demonstrates that concrete masonry is not only an alternative
but a very versatile, economical and technically efficient system to build low-income
housing in developing countries that should be studied for the benefit of the population.
Table 1. Subsidies for low-cost hosing in Colombia
(US Dollars) (as August, 1995)
TOWN POPULATION, thousands
INSTITUTION CONSTRUCTION <100 100 to 500 >500
INURBE IMPROVEMENT 1638 (210*) 1 638 (210*) 1 638 (210*)
NEW 1 638 (210*) 1 950 (250*) 2 340 (300*)
SOC. COMP. BOXES IMPROVEMENT 1 950 (250*) 1 950 (250*) 1 950 (250*)
NEW 2 730 (350*) 3 120 (400*) 3 510 (450*)
COMMERCIAL TOP PRICE 12 590 15 108 16 997
(100**) (120**) (135**)
Exchange rate: 949.5 Colombian Pesos/US Dollar. * UPACConstant-Acquisitive-Power
Unit7,8 US Dollars. ** MMSMinimum Monthly Salary126 US Dollars.
REFERENCES
INTRODUCTION
In the building of his shelter primitive man faces one supreme and absolute limitation:
the impact of the environment in which he finds himself must be met by the materials
which that environment affords. The environment is scarcely ever genial, and the
building materials are often appallingly meagre in quantity or restricted in kind.(Fitch
and Branch, 1960)
Emerging countries face similar problems with regard to materials of construction.
It has been suggested that as much as eighty percent of the world is underdeveloped or
developing and it is said that the growth in the population of the world between now and
2050 could be as much as twice the total population (4, 5109) of the world in 1980
(Preston and Rees, 1994). It is certain that the demand for shelter is going to be very
difficult to meet. The backlog of housing in South Africa has been variously estimated as
more than one million units and it is likely to grow before the implementation of the
national housing programme begins to take effect. In India the housing shortage in 1991
was estimated at 31 million units (National Housing Policy, Government of India, 1992).
Similar scenarios can be cited for the countries of sub-Saharan Africa and South and
Central America.
The provision of housing is governed by the need for shelter, the availability of
materials, the appropriateness (economic and physical) of the materials to the
environmental demands and by the social and other customs of the housed.
Since the greatest demand for shelter is from the under-employed and the unemployed,
the cost of the housing and the potential for job creation are of paramount importance. It
has been argued that the creation of employment should take precedence since people
with an income can and will make their own provision for shelter (Morris, 1994).
In South Africa the greatest stumbling blocks to the provision of acceptable shelter
have been legal (Groups Areas Act), lack of proclaimed and serviced land and the
unavailability of appropriate finance. Nevertheless the technology employed does play a
major role not only because different techniques have different immediate cost
implications but also because they involve the community differently.
What is sought after, therefore, are building systems that are relatively cheap but
durable and thermally acceptable as well as allowing participation by the community and
thereby improving the cash economy of that community.
The South African experience with the provision of housing on a massive scale in the
decade from 1950 to 1960 was only partially successful. It provided shelter, at low cost
and of very basic standard, in what were planned and executed as temporary dormitories
in accordance with the political ethos of the time. The townships were not intended to
be viable as towns or suburbs and lacked all community amenities. A repetition of this
experience could not and should not be tolerated. However, the need is such that it may
be necessary to build mass-housing in certain areas, using large contracting companies
and even industrialised construction systems. Neither of these would meet the criteria of
involving the community in the construction process and of keeping the money involved
within the community.
Much is made of the potential for job creation of massive housing programmes but it
might be argued that the reverse is also true, namely, that if unemployment can be
reduced it will lead to the provision of housing for themselves by the newly employed.
Material matters in affordable housing 171
(Morris, 1994) To give the unemployed housing that they cannot afford is to trap them in
sheltered starvation.
The official approach to alleviating the South African housing shortage has undergone
dramatic changes since the change in government in 1994. Initially reformist fervour led
to unrealisable promises of relatively large, expensive houses for all. Gradually reality
imposed itself on the authorities and the current policy is that all available means must be
explored to provide shelter, whether this entails mass production of formal, though
small, houses; small projects undertaken by emerging builders with or without
assistance in the management of the project or self-build housing on sites that are
provided with minimal services.
Much attention has been paid to creating a financial climate within which large or
small loans can be made available to those who have a minimum of collateral. At the
lowest end of the economic scale, outright grants are being made to provide the
unemployed or indigent with a site that has been serviced (electricity, water and waste
disposal) and on which he /she can erect a shelter that may be rebuilt or upgraded as the
occupants financial position improves (Cleobury J, 1995).
CONSTRUCTION TECHNOLOGY
During the era of mass-housing construction in 19501960, most of the houses were
constructed of clay brick or concrete block walls on concrete strip foundations with
galvanised steel or asbestos cement roofing (NBRI, 1987). There were no floors and no
internal doors, nor was there reticulated water, water borne sewerage or electricity
provided.
This approach would not be acceptable now but the way in which the houses were
constructed might. Although the labour content was high it was organised to
industrialise the process. Teams of unskilled workers were trained to carry out specific
tasks, for example, one team would dig the foundations, another would cast the
foundations, and a third would build up the corners. In this way the assembly line
workers moved along the assembly line of houses.
Various attempts have been made to construct mass-housing in South Africa using
industrialised systems of heavy-weight as well as with light-weight concrete panels
without notable success. It is not clear whether the scale on which these projects were
based was too small to justify the capital invested or whether the difficulties of
maintaining dimensional accuracy were beyond the capability of the protagonists but it is
likely that in the new circumstances there will be new attempts to use industrialised
systems to meet the demand for housing.
What is certain is that because of the lack of an industry and construction tradition
based on indigenous timber, housing construction will rely heavily on cement and
concrete in one form or another. This applies to the whole of Southern Africa. Despite the
very extensive plantations of exotic timber, South Africa cannot meet its own need for
construction timber and the indigenous timber is both unsuitable and scarce. Malawi is
facing such massive depletion of its timber resources resulting from the insatiable
demand for fuel that it might have to follow Indias lead and ban the use of timber in the
construction of governmental buildings. Zambia and Zimbabwe are in much the same
Appropriate concrete technology 172
overlie extensive dolomitic deposits and the formation of sink-holes is not uncommon. A
complete reduction plant at a gold mine in the Western Transvaal disappeared into such a
sink hole some forty years ago.
The consequence is that much attention is being paid to the foundations needed to
overcome the problems caused by, for example, heaving clay. A variety of raft-like
foundation designs have been developed and so-called waffle-rafts can be produced at
costs that compare well with the more traditional strip foundations (Williams A A B and
Donaldson G W, 1980, Harrison N C, 1994). They are purpose designed in reinforced in-
situ concrete and ensure that the superstructure will not crack disastrously.
Innovative approaches to the construction of walls include the use of in-situ cast walls
of sand/cement concrete. The NBRI developed this system under the name of BRISC
(Building Research In-Situ Concrete) for use at Kabokweni in the eastern Transvaal and
showed that acceptable accommodation could be provided at an affordable price (Scott T
W, 1983). The design used plywood form work designed to be manhandled and requiring
no cranage or other mechanical manipulation.
The relatively dry mortar was then tamped into the forms to produce monolithic walls
with window and doorframes pre-positioned and built in.
This design had an interesting sequel when the NBRI was called on to help the
government of the Comoros to develop a building system that would avoid the
destruction of the coral and the beaches that constitute a major source of tourist income.
The custom had been for coral to be calcined to produce quicklime and combined with
beach sand to produce an indigenous hydraulic cement. The islands are volcanic and
there is little other building material available. The solution proposed was that cement
should be imported and combined with volcanic gravel to form a coarse mortar.
Formwork, similar to that used in the BRISC system but even easier to handle, was
designed. Large lumps of volcanic rock were placed in the forms by hand and the coarse
mortar was tamped around these plums to form the walls. Once the formwork had been
stripped, a thin rendering of cement mortar provided a surface finish that was acceptably
smooth and easy to paint.
Evaluation of this technique showed that it used only half as much cement as would
have been used had the same house been built of concrete blocks and mortar (NBRI,
1986).
This development was a good example of using the materials available, in this case
volcanic gravel and broken rock, (avoiding the need for materials that are scarce, and
reducing the need for expensive imported products) and allowing the use of local labour
instead of sophisticated equipment. Following this theme and remembering the statement
by Fitch and Branch that the need for shelter must be met by the materials that the
environment affords, one needs to examine the use of soil as a building material.
Soil has been used as a building material throughout the history of humankind and it is
still being used as adobe or pis-de-terre (rammed earth) and there are examples of such
use in many parts of Africa. One of the disadvantages is that sun-dried soil blocks or
mortar are not resistant to water. For many in Africa it is standard practice to repair their
homes or other structures annually to counteract the damage caused by rain. In Malawi
the thatched roofs that are typical of traditional home building, are extended to almost a
metre beyond the perimeter walls to protect them from direct rain induced erosion.
Appropriate concrete technology 174
If, however, the soil-based building elements, whether they be of rammed earth, sun
dried blocks or soil-and-cowdung mortars, could be stabilised against the effects of water,
the criteria set by Fitch and Branch may be met.
Much work has already been done on the production of soil-based building materials
and the Building Research Establishment (BRE) in Britain and various researchers in
Australia, among many others, have published results of their investigations. BRE
developed a manually operated block-making machine for stabilised-soil blocks
(BREPAK) (BRE, 1987) and Middleton (1975) in Australia developed a device for
testing the resistance of soil based building materials to standard conditions of water
spray. Internationally several manufacturers have produced more or less sophisticated
machines for producing soil or soil-cement masonry units and South Africa is no
exception. There are at least two manufacturers of machines for producing building
blocks currently in the market.
In the laboratories of the University of the Witwatersrand the production of adequately
strong and water resistant walls of stabilised earth in the form of rammed earth and of
hydraulically compressed blocks has been investigated. The soils were characterised by
the usual techniques of soils analysis and the range of cement concentrations that would
ensure durability determined. The results indicate that at relatively low levels of cement
content (between 6 and 8% of an 85% OPC/15% PFA blend known as PC15PFA), a wide
range of soils could be satisfactorily stabilised. (Roberg C W, 1993; Blight G F, 1994).
The potential therefore exists for cement-stabilised soil to play a major role in the
alleviation of the housing shortage in Africa since the equipment is relatively inexpensive
and simple to operate, the production of masonry units or rammed earth walls labour
intensive and the basic materials readily available.
A development that has recently come to our attention is a simple hand-operated
machine for the production of concrete roof tiles. The use of concrete roof tiles
manufactured in sophisticated factories is well established and provides attractive and
durable roofing but the advent of a simple device that allows rural development of small
scale manufacturing facilities could be valuable in its creation of employment and
entrepreneurship. The device has been used extensively in Malawi and Zimbabwe and the
quality of the product appears to meet the needs of the users. Nevertheless, the South
African manufacturer of the tile making machine has submitted tiles to the South African
Bureau of Standards (SABS) for evaluation to ensure that the final product will resist the
often destructive hailstorms that strike the inland areas of Southern Africa.
There are many other innovative techniques that are being actively promoted and most
of them rely on the use of OPC or OPC/PFA blends. Because South Africa derives about
90% of its electrical energy from coal-fired thermal power stations and the ash content of
the coal is high (>30%), the amount of PFA produced annual is measured in tens of
millions of tonnes. The incorporation of PFA in cement blends has become almost
universal. In addition the use of coarser grades of ash as fillers or fine aggregate is
becoming widespread.
An example of this in the housing field is provided by the commercially developed
version of wattle-and-daub construction that uses expanded steel mesh to replace the
lathes of wattle and a cement and PFA mortar to create the wall. The cementitious
material is applied to the metal on site by hand or it may be sprayed on. The wall panels
thus formed are nominally 4 metres by 2,4 metres in size and the claim is made that a
Material matters in affordable housing 175
house of modest size can be constructed in one day. The structure is founded on steel
rods driven into the soil and attached to horizontal steel U-channels that support the
walls. The PFA is necessary as it aids pumpability for Gunniting and reduces that cost
of the cementitious mortar.
It has been assumed that concrete block or ash block construction is the cheapest way
to erect low cost single storey homes and much has been made of the labour intensive
local production of such blocks by the prospective home owner/builders as a means of
empowering the homeless to contribute to their housing and to do it in such a way that
the money involved is circulated within the local community. By using suitable, screened
clinker, sand, coarse PFA and classified fly ash with cement as the binder it is possible to
produce blocks of adequate strength and durability that are 95% ash-based. It is claimed
by the protagonists of this wattle-and-daub system that it can produce walls at a price
that is only two thirds the cost per square metre of concrete block walls.
Other approaches to the provision of housing include renewed interest in foamed
concrete. The Department of civil Engineering at the University of Pretoria recently
reported the construction of a 24m2 house on the basis of casting 6m4m foamed
concrete (800 to 1400 kg.m3), including doors and windows where required, and tilting
the cured slabs into position. The concept of tilt-up construction is not new but we
believe the use of foamed concrete in this form of construction may well be (The Civil
Engineering Contractor, 1995).
CONCLUSIONS
The need for housing throughout the developing world is great and there is no simple
solution to providing such housing. The environmental requirements differ according to
location and climate, there is no single construction technique that will meet all needs and
the appropriate solutions will be as varied as the circumstances. Nevertheless it is clear
that cement and cementitious materials including pulverised fuel ash will play a major
role in addressing the shortage of housing.
In South Africa many different approaches will be adopted and used in parallel and
there is no reason to believe that it will be different in other developing countries but the
participation in the provision of housing by the those who need the housing will be
crucial to the success of any endeavour.
REFERENCES
Keywords: Soil, stabilisation, low cost housing, ordinary portland cement (OPC),
pulverised fuel ash (PFA), construction, appropriate technology.
Professor John Morris is the PPC Professor of Building Science in the Department of
Building and Quantity Surveying of the University of the Witwatersrand, Johannesburg.
Formerly Chief Director of the South African National Building Research Institute
(NBRI) of the Council for Scientific and Industrial Research, he has been involved in
research into building materials for many years. He has been an honorary member of the
Institute for Housing of South Africa since 1987 and was recently Chairman of the
Transvaal Branch of the HSA.
Geoffrey F Blight is a final year BSc (Building) student in the Department of Building
and Quantity Surveying of theUniversity of the Witwatersrand, Johannesburg.
Appropriate concrete technology 178
INTRODUCTION
It has been suggested that as much as 80% of the world is underdeveloped or developing
and it is said that the growth in the population of the world between now and 2050 could
be as much as twice the total population (4,5109) in 1980 (Preston and Rees, 1994). It is
certain that the demand for shelter is going to be very difficult to meet. The backlog of
housing in South Africa has been variously estimated as more than one million units and
it is likely to grow before the implementation of the national housing programme begins
to take effect. In India the housing shortage in 1991 was estimated at 31 million units
(National Housing Policy, Government of India, 1992). Similar scenarios can be cited for
the countries of sub-Saharan Africa and South and Central America.
Since the greatest demand for shelter is from the under-employed and the unemployed,
the cost of the housing and the potential for job creation are of paramount importance. It
has been argued that the creation of employment should take precedence since people
with an income can and will make their own provision for shelter (Morris, 1994).
In South Africa the greatest stumbling blocks to the provision of acceptable shelter
have been legal/political (Group Areas Act), lack of proclaimed and serviced land and the
unavailability of appropriate finance. Nevertheless the technology employed does play a
major role not only because different techniques have different immediate cost
implications but also because they involve the community differently. In addition serious
concerns have been expressed about the ability of the local materials industry to meet the
expected demand for products.
What are sought after, therefore, are building systems that are relatively cheap but
durable and thermally acceptable as well as allowing participation by the community and
thereby improving the cash economy of that community.
The use of soil as a building material is as old as humanity itself and sun-dried bricks
were used in Ancient Egypt and Mesopotamia. It was used in the south western United
States by the cliff-dwellers before the arrival of Europeans. In Malawi the use of sun-
dried clay building blocks is fairly common and the traditional practice of extending the
thatched roof on all sides to provide an eave-overhang of about a metre is a pragmatic
solution to the problem of rain-induced erosion of the clay-block walls. It incidentally
also improves the thermal performance of the building as it shades the walls from the
tropical sun.
In southern Africa sun-dried clay bricks or cement stabilised pis-de-terre (rammed
earth) appear to have been stigmatised as unacceptably primitive materials whereas in
parts of the south western United States (Arizona and New Mexico) adobe houses are
sought after, mainly because of the excellent thermal comfort adobe construction
provides. However, research into the use of clay as a building material has been
widespread and there are a number of initiatives aimed at re-establishing the use of clay
as a building material (McHenry, 1984, United Nations Centre for Human Settlements,
1984, Roberg, 1993).
Cement stabilisation of soil for the production of building blocks 179
The problem of water resistance is likely to restrict the use of unfired clay building
materials seriously and it provided the motivation for the research that was undertaken by
the Department of Building and Quantity Surveying of the University of the
Witwatersrand.
Wet/dry durability
Water absorptivity
Whenever possible unstabilised blocks (zero % cement) were included in the tests but
since they disintegrated upon exposure to water they could not be tested for standard
penetration, wet/dry durability or water absorptivity.
The blocks were allowed to cure under plastic sheeting for three days, thereafter the
unstabilised blocks were allowed to dry out, uncovered, while the stabilised blocks were
kept under polyethylene sheeting until they were tested for unconfined compressive
strength after four days.
Despite some variations in the rate and manner of strength development depending on
the type of soil, Figure 1 and Figure 2 show that all the stabilised blocks, irrespective of
the extent of stabilisation, reached acceptable strengths (3MPa) within seven days of
manufacture. It may be concluded that the degree of stabilisation has a direct linear
effect on the strength of blocks made of soils similar to those used in this study. Blocks
which are stabilised by the addition of 4% PC15PFA will reach higher 28 days strengths
than those which are unstabilised (Blight, 1994).
The accelerated water erosion test used a device designed by the Building Research
Division of the CSIRO (Australia) and described by Middleton (1987). In this test a spray
of water (at a prescribed volume per minute and prescribed pressure) through a standard
nozzle is directed at the test object at a prescribed distance for one hour. The test is
interrupted briefly every fifteen minutes to examine the test object for erosion.
The results indicate that unstabilised blocks of either soil type are unsuitable for use
where they might be exposed to rain whereas the addition of as little as 4% of PC15PFA
improves the performance of the blocks remarkably. Such deterioration as occurred was
limited to spot erosion that did not develop further and It would seem that unless a high
Appropriate concrete technology 184
standard of finish was demanded, the addition of 4% PC15PFA by dry mass to most
soils, will provide a sufficiently durable block (Blight, 1994).
The wet/dry durability test was based on a test for engineering materials (TMH1,
1979). The blocks were cured for seven days and then subjected to twelve cycles of wire
brushing, wetting for five hours and drying for 42 hours before repeating the cycle. By
recording the mass and moisture content of the block before starting the test and again
after the twelfth cycle the amount of material lost can be determined.
Not unexpectedly there is an inverse relationship between the amount of material lost
and the degree of stabilisation and as with some of the other tests there was a difference
between the performance of the two different soils. Since the test is difficult to relate to
the demands likely to be placed on masonry units in home building, only the generalised
conclusion that higher degrees of stabilisation will provide greater abrasion resistance
and that coarser soils tend to be more abrasion resistant can be drawn from this test.
The water absorptivity tests were not very successful for various reasons but it was
evident that the coarser material (decomposed granite) absorbed more water than the finer
material and that after two hours the rate of water absorption dropped sharply in both
cases reaching values of between 2% and 4,5%. Increasing the degree of stabilisation also
decreased the amount of water absorbed.
The overall conclusions of this set of investigations were that the addition of as little
as 4% PC15PFA to most soils would allow the production of hydraulically compressed
masonry elements of adequate strength and durability for the construction of single storey
domestic dwellings.
Thermal performance
Since the great majority of the housed and unhoused South Africans live on the high
interior plateau (1000 metres above mean sea level) known as the Highveld where
diurnal ambient temperature fluctuations often exceed 20C the thermal mass of the
construction materials becomes important. Light weight materials have little thermal
inertia and the temperatures of houses built of them are likely not only to follow the
variations in ambient temperature but to exceed them. In contrast heavy weight materials
such as brick-and-mortar or soil-cement blocks have a considerable capacity for
absorbing heat during the day and releasing it at night thereby acting as a thermal
flywheel and reducing the fluctations in the internal temperatures of the houses (Van
Straaten and Lotz, 1972, Wentzel, 1981).
Since low income and no-income families can in any case ill-afford the cost of heating
their homes in winter, it is of considerable importance that the hundreds of thousands of
houses that are to be built under the South African Governments new housing initiative
be as energy efficient as possible.
The properties of soil/cement masonry units, manufactured on site are such that they
should be seriously considered as one of the ways to meet the housing needs of
developing countries. A system such as that examined in our laboratories might offer an
approach that, because of its simplicity, would allow relatively unskilled workers to erect
durable, thermally efficient homes at reasonable cost, thereby developing their skills and
contributing mterially to relieving the housing shortage in the developing world.
Cement stabilisation of soil for the production of building blocks 185
REFERENCES
Keywords: Labour-intensive, Uncut stone masonry, Cement mortar, Arch bridge, Cube
strength, Finite element analysis.
Roderick G D Rankine is Lecturer of Construction in the Department of Building, at the
University of the Witwatersrand, Gauteng, South Africa.
Professor Geoff J Krige is a past Head of Structural Engineering, University of the
Witwatersrand, Gauteng, South Africa.
An innovative labour-intensive method for construction of arch bridges 187
INTRODUCTION
Once the foundations are in place, sheets of corrugated iron are pre-cranked to the shape
of the intrados and positioned over the foundations to provide temporary support to the
arch span. In the case of large spans, these sheets are propped with gum poles to help
bear the weight of the masonry until it becomes self-supporting (Fig 1). In some areas
alternative centering using 50mm saplings, covered by old cement packets, and supported
by temporary mud brick walls has been successfully deployed(1).
Appropriate concrete technology 188
Workers gather stones (typically weighing anywhere from 1040kg), from the fields
nearby. Sand from the river bed nearby is used to make a cement mortar. The cement:
sand ratio (by volume) varies from 1:4 to 1:3. The manual assembly begins by
embedding individual stones into a thick layer of mortar. The stones are not cut or
dressed in any way, and are placed dry in intimate contact with one another. This
procedure is followed; horizontal layer upon horizontal layer, until a solid structure
results. As a precaution, a grid of high tensile reinforcing steel has been placed
horizontally in both directions, immediately above the crown of the arch in the initial
structures. This steel is placed in a thick bed of cement mortar, of a 1:3 cement: sand mix.
Another bed of mortar (or preferably concrete) is used to even out the bridge deck to
form a trafficable surface.
Cube Tests
As traditional 100mm and 150mm concrete testing cubes are too small to accommodate
even the smaller stones used in this material, an alternative test was developed by
Rankine(6). For ease of fabrication, the traditional cube shape was maintained, but the
overall dimensions were increased to accommodate the large stones. Initially cubes of
300mm, 400mm and 500mm were chosen as a means of exploring different possible
An innovative labour-intensive method for construction of arch bridges 189
sizes. The 300mm size proved too small to accommodate all but the very smallest stones
and the 400mm size only permitted one, or at most two, of the larger stones, Thus, it was
decided that the 500mm size, shown in Fig 2, was the most representative of the material
and has subsequently been adopted for further testing.
Cube stresses
(7)
Rankine et al describe some large cube tests which were carried out as a preliminary
indicator of the strength of the materials used for these bridges. The cubes were made on
site, using weathered hand gathered stones and the same procedures as were used for the
bridges themselves. A subsequent series of tests(8) using quarried dolomite rock (graded
in various combinations using individual stones weighing between 1 40Kg), and the same
strength mortar has produced a stronger material. The 28 day compressive strength of
these cubes is presented in Table 1. The weathered stones with a friable chalky surface do
not appear to provide as good a substrate for the bonding of cement mortar, as do
quarried rocks with fresh faces.
Table 1 Results of compressive tests.
Cube size (mm) Mass (Kg) Failure load (KN) Failure stress (MPa)
300 60 1260 14, 0
400 149 1435 9, 0
500 2406 9, 6
Appropriate concrete technology 190
The first three specimens contained hand gathered stones with weathered faces.
Specimens prefixed with an asterisk represent an average of three test cubes containing
quarried dolomite (with fresh clean surfaces). The 28 day 100mm cube mortar strength
for all specimens was between 14 and 15 MPa.
Measurements in-situ
To gain an initial understanding of the magnitude and nature of the stresses and strains
within the arch structure, measuring devices were made by fixing electronic strain gauges
onto concrete prisms as shown in Fig 3. These were then built into arches during
construction. At a stage when the bridge became trafficable, a truck with an axle load of
100 KN, was driven across the bridge (see Fig 4), to record the stresses in each device at
various axle positions. The first series of stress measurements confirmed that, the live-
load-induced strains, and hence stresses were very low. The measured tensile and
compressive stresses, induced by this live- load did not exceed 0,074 MPa. It was evident
that very sensitive instrumentation was needed to provide more reliable data.
Subsequently, a mechanical strain muliplying device has been developed(9) in an effort to
amplify the the strain before it is recorded. This results in a more sensitive signal, without
any sacrifice in clarity and without generating more noise. The data gathered will be used
for validation of further theoretical models.
An innovative labour-intensive method for construction of arch bridges 191
The low measured and predicted stresses within these structures, compared with the
measured strength of the material, provides some temptation for speculation that an
adequate margin of safety against compression failure exists. Preliminary indications
suggest that allowable stresses should be limited to below one MPa, since traditionally,
masonry arches have been assigned very high factors of safety against crushing of
between 10 and 40(10). It is also prudent to remember that the greatest contributor of stress
in these structures appears to be the theoretically predicted self-weight-induced stress
(which presently remains to be validated).
CONCLUSION
REFERENCES
1 GPOBLER, L.J., PIENAAR, P.A., MALOKA, H.M., AND PIENAAR, A.J. Community based
construction of drainage structures; The Bolobedu case study. Annual Transport Convention,
Session: Community Consultation RDP Implementation, Pretoria, 1995.
2 WOOTON, P.M., and STEPHENS, R.D. Planning and provision of appropriate drainage
structures for low cost rural roads. Annual Transport Convention, Vol 1B Bridgets CSIR,
Pretoria, 1987.
3 MATNWARING, P., and HASLUCK, J.P. An approach to the provision of low-cost roads in a
developing country. Proc Annual Transport Convention, Session: Transportation in Developing
Countries, S350 (Vol 8-TC), Pretoria, 1985.
4 THE RECONSTRUCTION AND DEVELOPMENT PROGRAM, A POLICY FRAMEWORK,
Johannesburg: African National Congress, 1994,
5 NPWP, 1994. National employment. creation program for the provision of public infrastruture
using labourintensive methods. Report prepared for the National Economic Forum technical
committee on a public works program, June 1994.
6 RANKINE, R.G.D. Development of a standardised strength test for rubble masonry. Paper to be
presented to the Journal of The South African Institution of Civil Engineers.
An innovative labour-intensive method for construction of arch bridges 193
7 RANKINE, R.G.D., KRIGE, G.J., TESHOME, D., AND GROBLER, L.J. Structural aspects of
labour intensively constructed, uncut stone masomy arch bridges, The Journal of The South
African Institution of Civil Engineers, Vo1 37, No 3, Third quarter 1995, pp 1217.
8 LINDQUIST, E.S., and GOODMAN, R.E.Strength and deformation properties of a physical
model melange, Proceedings of the 1st North American Rock Mechanics Symposium/The
University of Texas at Austin/13 Jun 1994, pp 843850.
9 RANKINE, R.G.D. A mechanical method of sensitising electronic strain measuring devices in
rubble masonry structures. Paper to be presented to the 1st South African Conference on
Applied Mechanics. Topic 4, Strain, Applications of strain measurement, and strain
measurement devices, July 1996,
10 TELLET, J. A Review of the literature on brickwork arches, Stoke-on-Trent, BCeramRA, Tech.
Note 338, 1982.
MANPOWER MOTIVATION
IMPROVES QUALITY IN
CONSTRUCTION
M S Chishty
M A Choudhry
University of Engineering & Technology
Pakistan
INTRODUCTION
RESEARCH METHODOLOGY
The data for this study were collected at 24 different Government sponsored projects,
ranging in cost from Pak. Rs. 100 millions to Rs. 750 millions approximately. The
projects selected for survey work were of different types and of different geographical
regions of the country. The range for stage of completion was different. Some were on
Appropriate concrete technology 196
going whereas maximum percentage of these projects was already completed. The data
collected at each of the project consisted questionnaire survey and the interviews of the
personnel at the site. The questionnaire was prepared after consultation with experienced
personnel attached with the construction work. A team of the undergraduate students
were trained for the survey work and the interview work was properly demonstrated at
site. The questionnaire mainly related the manpower motivational tools, quality assurance
techniques and time period.
The survey team faced great difficulties in the collection of data. The response was not
encouraging, as at most of the sites the management hesitated to provide required
information. The relative information at some of the sites, was obtained from the lower
staff as proper management was not available whereas other sites were not even aware of
such motivational plans. At certain organized project sites, this type of survey was very
much appreciated and even some additional methods/approaches were conveyed which
were ultimately incorporated in the questionnaire.
In the processing of the collected data, the major interest was to identify the factors
influencing motivation of manpower and their productive ability. Quality of construction
projects is related to motivation, performance and satisfaction of the manpower. Graphs
regarding the factors influencing the quality and efficiency of construction project are
shown in figures 1 through 7.
FIG. 2: REGULAR
MANAGEMENT CELL IN
VARIOUS COMPANIES
An inspection of the Figures 1 and 2 reveal that management practices are observed by a
very small number of companies; only 13% of the companies have qualified, experienced
and trained managers and 29% have regular management cell. Similarly, education and
other training programs which impart a sense of confidence and result in higher level of
performance and motivation in the workers, are also rare; only 21% of the companies
surveyed provide educational training to their workers, see fig. 3.
Another very important factor that usually results in higher levels of performance is the
safe working conditions and the presence of some incentive plan
further enhances it. However, the conditions regarding safety and motivational plans
were not found proper and encouraging. From figures 4 and 5 it is clear that conditions of
safety measures at construction sites are extremely poor; only one out of all twenty four
companies surveyed has its own safety manual, and same company has motivational
plans for its employees.
Fig. 6 and Fig. 7 show the time and cost overrun for various projects. The projects which
suffered maximum delay in completion and increase in cost are those which had a
combination of the factors discussed above.
Some of the other factors resulted in lower performance level include frequent Hiring and
Firing, and Lack of Security against employment to workers. Similarly, on-site conditions
like improper Site Layouts, Crew Interference, Unavailability of Required Materials at
proper time affected severely the morale and motivation of workers. Frequent Design
Changes and Untimely Quality Inspections resulted in poor workmanship.
On the other hand, on projects which were completed on time and within initial
estimated costs, least amount of rework was reported. This was mainly due to the
presence of Qualified Construction Manager on the site, and trained workers. Previous
experience of workers on similar projects further enhanced the productivity and quality.
On these projects, most of the equipment was owned by the owner and proper alternative
arrangements in case of failure of an equipment were available. Similarly, safe working
conditions and presence of incentives resulted in higher level of performance,
productivity and hence improved quality.
Based upon the collected data, the authors have reached at following conclusions and
recommendations:
1. At most of the projects, the quality problems were mainly due to the factors
affecting morale and motivation of the workers. Absence of qualified construction
manager in the team was the most obvious one. To enhance the performance level of
workers, inclusion of qualified project manager in the construction team must be a
prerequisite. This may be included in the conditions of pre-qualification of the
contractors, and minimum qualification of the manager must be specified at the time of
award of the project.
2. Lack of education and training of workers resulted in poor quality work. In an
industry like Construction, the problem may be solved, to a large extent, by issuing
instructions about that days work to Foremen and Supervisors every morning in 1520
minutes before the start of the work. These instructions should include the scope of work
for the day. These First-Line managers then may be advised to convey the same to the
workers.
3. Lack or absence of incentives and motivational plans is another factor that affects
quality/efficiency in construction. In this regard, it may prove to be quite encouraging
that at the end of each milestone, those who have shown best performance or have
improved during that time, be rewarded. These rewards may include financial or non-
financial rewards or a combination of the two.
4. Unsafe working conditions lead to the feeling of fear because of the impending
danger. This results in decreased concentration to work and hence low productivity and
poor quality work. In order to enhance quality in construction, it is suggested that proper
safety measures be ensured as per conditions of the project. The prevailing situation may
be improved by educating the workers about safety everyday before the start of the work.
Manpower motivation improves quality in construction 201
REFERENCES
1. Ashford J.L. 1989 The Management of Quality in construction, E. & F.N.Spon Ltd. 29 West
Street N.Y.
2. Chishty M.S., Improving Construction Quality Through Improving Motivation, Scholarly
Paper submitted to Department of Civil Engineering, University of Maryland, U.S.A, in partial
fulfilment of requirements for degree of Master of Science, 1992
3. Chishty M.S., Choudhary M.A., Risk Management Ensures Quality in Construction Projects,
International Conference on The Concrete Future 911 Feb. 1993, Kualalumpur, Malysia.
4. Choudhry M.A. et. al. Trained Manpower Efficiency in Construction Projects, International
Symposium on Property Maintenance and Modernisation, 79 March, (1990), Singapore.
5. Choudhry M.A. et. al. Effect of Time Requirement upon Economy of Construction projects
(Survey Report), 10th International Conference CIB 86 Congress on advancing Building
Technology, Sept. 2126, (1986), Washington, USA.
6. Laufer Alexander & Jenkins Douglas G., Dec. 1982. Motivating the Construction Workers,
Journal of Construction Division, ASCE, Vol. 108, No. CO4, (531545)
7. Maloney William F., Employee Involvement in Constuction Graduate course Handout ENCE
667. Spring 1991.
8. Maloney William F., James M.McFillen, Motivation Implications of Construction Work,
Journal of Construction Engineering & Management, Vol.112. No.1. March 1986
9. Maloney William F., James M.McFillen, Motivation in Unionized Construction, Journal of
Construction Engineering & Management, Vol. 112. No.1. March 1986
APPRECIATING LOW-STRENGTH
CONCRETE
M A Abdul-Salam
University of Qatar
Qatar
INTRODUCTION
Strength -in addition to workability- has for many years been the main property of
structural concrete looked after and pursued by those involved in the construction
industry. The striving for an even stronger concrete has continued, mainly because of the
need of the construction industry in the industrialized world in particular to increase the
strength per weight ratio in the competition with other structural materials, namely,
structural steel.
The cost of improving compressive strength of concrete is usually higher, the higher
the strength and the extra cost has to be weighed against the benefits gained. The cost
effective range of strength is usually determined from experience in a certain locality.
Construction industries in some parts of the world are not well equipped to produce high
strength concrete, but because of the international trend, engineers in these parts might
occupy themselves with strength more than needed, concentrating less on other properties
related to durability. This is particularly wrong in a rather severe environment like that in
parts of the Middle-East where durability is turning into an obstacle that hinders the role
of concrete as a structural material.
A number of investigations were carried out on concrete being the main structural
material for ordinary houses and small buildings in two countries in the Middle-East, the
findings of which will be presented in brief.
Over a period of five years, 1985 to 1990, a comprehensive study was performed to find
out the reasons for structural problems in a number of buildings, mainly residential four
or five stories buildings and one or two stories school buildings in the suburbs of the city
of Damascus, Syria. In the investigations, destructive and nondestructive methods were
used to assess the quality of concrete. The major finding regarding strength was that the
value in these buildings was between 12 and 16 MPa, much less than the filed
construction cube samples strength of 20 to 25 MPa. But, in spite of this low value,
strength was the main cause of the observed structural problems only when:
1 Extra loads were applied in excess of design loads (adding a story or two in addition to
the number of stories considered in design).
2 Strength was less than 10 MPa.
The main reason for structural problems was differential settlement of foundations. All
roofs were well protected from rain and corrosion was rarely observed.
Out of hundreds of buildings around, a limited number had problems. It can be
assumed that most of these buildings have strength of no more than 16 MPa and are still
in acceptable condition. As far as, the design loads are not exceeded and no differential
Appropriate concrete technology 204
settlement occurred. It is known that designers in that area incorporate small service
stresses in their designs.
Qatar is a semi-island dominated by desert. Iron ores do not exist, and wood does not
exist in a scale sufficient to supply the construction in wood. In contrast, limestone is
available in large quantities and so is the clay with the quality needed for the
manufacturing of cement. Consequently, concrete emerges as the only reliable structural
material available locally. Only for special structures do other structural materials prove
to be more competitive.
The achievement of high strength concrete is relatively costly because of the following
factors:
1 Although good limestone aggregate is available, the properties of the stone layers varies
considerably so that a careful selection should be made if the best quality aggregate is
to be extracted.
2 Curing is a major problem because of the high temperature and the scarcity of clean
potable water.
3 Site control is tough and is not always easy to provide.
On the other hand, durability of concrete has been causing a lot of worry about the
efficiency and reliability of the material. Many engineers had become satisfied with
limiting the life of buildings to 15 to 20 years. Some argue that the solution comes with
importing aggregate from neighboring countries. Many are dragged by the international
trend into seeking higher strength properties to solve the prevailing problems.
It has become clear that the main problem for durability is corrosion caused by
chloride contamination. However, when good practice is exercised, i.e., selection of good
aggregate (type A), using cold water in mixing and efficient curing with clean water for a
week, the concrete is good in durability and strength. This is achieved by well established
local or foreign companies at a relatively high price. For ordinary commercial and
residential building, such a high cost is not acceptable.
An examination of the concrete actually used in ordinary housing projects was
necessary to shed a light on the real concrete properties needed.
The average strength of twenty-five houses under construction was found to be 14.8
MPa with a min. value of 11.7 MPa . For comparison, three private houses in the same
areas were investigated and the average strength for these was found to be 25.0 MPa.
It was noticed that contractors who participate in public housing tenders are different
from those involved in the building of private houses and villas. Using the same source
and grade of ready-mixed concrete, private houses contractors achieved an average
concrete strength in the structures which is 1.5 times that achieved by public housing
contractors.
Six old houses set for demolition were also examined. They were about 25 years old.
The strength was as seen in Table 1 with an average of 13.7 MPa and a minimum value
of 9.9 MPa. The chloride & sulphate contents are also shown in Table 1, high values are
evident. It was constantly found, however, that corrosion occurred in the roof slab
because of rain water penetration. Water proofing was damaged and steel cover was very
small. Other parts of the structure did not show signs of corrosion in-spite of the high
chloride contamination. Other causes of deterioration were foundations settlement and
temperature-caused volume changes.
Table 1 Concrete in 25 years old houses
House Average Compressive Chloride Content per Sulphate Content per
No. Strength MPa Weight of Cement % Weight of Cement %
1 10.0 0.82 11.9
2 29.7 1.76 7.1
3 9.9 1.42 5.2
4 10.7 1.14 7.2
5 11.8 0.83 3.0
6 10.3 0.48 6.2
In Qatar, the Ministry of Municipal Affairs supervises the building of both private and
public housing. Cube samples from casts in these houses are tested and recorded
strengths are kept in files. These files were surveyed in full starting from the year 1978
till 1994. The maximum, average and minimum vales of 28 days strength are found and
are shown in Figure 1. Although the average values seem to be good, the minimum
values are clearly low and are similar to the values found in the public housing
investigated.
Appropriate concrete technology 206
How low the strength of concrete in an existing structure could be? Is a question that
might be raised having shown the low values found in the mentioned investigations.
Using an Ultrasonic Pulse Velocity tester PUNDIT, a correlation between concrete
cube strength and UP velocity was developed over a number of years [2]. The correlation
curve is shown in Figure 2. It is suggested that the shown minimum strength is the lowest
that can be found in an existing structure.
Having exploited the major cause of deterioration in the severe environment, protection
from corrosion should be the major factor to be considered to achieve an acceptable
durability. Design plays a role. Knowing that low strength is more likely to be achieved
by construction, a practical design should be based on the expected values. Controlling
deflection and crack width should be a major objective of design. Eventually a proper
protection to steel bars will be accomplished.
To see how design can achieve the mentioned objectives, a design procedure based on
the selection of steel reinforcement ratio using the methods of ACI-318 [3] was carried
out to design a rectangular beam cross-section for an ultimate applied bending moment
Mu= 340 kN.m. Four categories of materials grades are chosen, materials A, B, C and D
as shown in Figure 3.
The resulting cross-sections for =0.50max are given in Table 2. The maximum
instantaneous deflection versus ratio of reinforcement is plotted in Figure 3. Lower
concrete strength will result in a larger cross-section, hence, smaller deflection . Steel
strength has the opposite effect because of the design code assumptions.
The calculated average crack widths shown in Table 2 (ACI formula) show that lower
concrete and steel strengths reduce the values and proper detailing causes a further
reduction. A crack-width reduction of 30% is reached by using a higher percentage of
concrete (15% by volume of lower grade concrete) and higher percentage of steel (15 %
by area of lower grade steel). The depth increase is not usually a problem in low rise
buildings.
Table 2 Design of a Cross-Section
Materials =0.50max b h As W1 W2 W3
(mm) (mm) (mm2) 3 bars in a 6 bars in a 6 bars, 4 in the
row row first row
(mm) (mm) (mm)
A 0.01204 250 745 2059 0.200 0.177 0.170
B 0.00813 250 780 1457 0.260 0.266 0.218
C 0.01806 250 620 2524 0.203 0.181 0.173
D 0.01219 250 645 1786 0.264 0.231 0.222
Appropriate concrete technology 208
CONCLUSIONS
1. High concrete strength does not guarantee great durability.
2. In corrosive environment, protection of concrete may be more important for durability
than great strength.
3. Foundation stability is very important for durability of structures.
4. Significant differences exist between contractors in concrete quality, even when
beginning with the same ready mixed concrete.
5. Design and construction of structures should include recognition of low-strength value
concrete in constructing durable structures.
REFERENCES
1. Al-Mahmoud, Q, Awad, A. Survey of concrete quality & strength. Graduation project, College
of Engineering, Qatar University, 1995.
2. Abdul-Salam, M A. Ultrasonic pulse velocity versus strength of concrete in Qatar. Engineering
Journal of Qatar University, Vol. 5, 1992, pp 8793.
3. American Concrete Institute. Building code requirement for reinforced concrete. ACI318R-89,
1992.
BEHAVIOUR IN EXTREME CLIMATES
OF CONCRETE MADE WITH
DIFFERENT TYPES OF CEMENT
C C Videla
Pontificia Universidad Catolica de Chile
J P T Covarrubias
Chilean Cement and Concrete Institute
J M D Pascual
Pontificia Universidad Catolica de Chile
Chile
INTRODUCTION
EXPERIMENTAL PROGRAMME
Tacking into account the definitions for cold and hot weather concreting given in the
Chilean standard [2] and considering the extreme temperatures that are normally
encountered on construction jobs, the minimum concrete temperature on cold weather
was defined as 5 C and the maximum temperature on hot weather as 30 C. With
these definitions and considering a reference temperature of 20 C, a study of the
influence of temperature variations on concrete behaviour was carried out from the
following points of view: effect of its magnitude (5, 20 and 30 C) and effect of the
exposition period or curing time under those temperatures (particularly, curing times of 5
hours, 7 and 28 days were studied).
Behaviour in extreme climates of concrete made with different types of cement 211
Mix Proportions
The mix design was performed according to the 1988 British method [8], adjusted to
Chilean conditions. The aggregates were combined to have a grading between curves 2
and 3 of Road Note N4 [3]. To achieve the different W/C ratios the cement content was
adjusted. The mix proportions of the studied concretes are shown in Table 1.
Appropriate concrete technology 212
In Figures 3, 4 and 5 the relative compressive strength (ratio of strength of concrete cured
at different temperatures and ages to the strength of concrete cured at 20 C) versus time
are shown.
In general, for all cement types a tendency to have higher long term compressive
strengths when cured at 5 C than cured at 30 C is observed. This can be clearly seen in
the relative compressive strength charts (see Figure 3). For slower strength gain cements
this effect is very noticeable, as can be seen comparing Figures 4 and 5. With these
cements the curves intersect and at an age of 90 days the strength of samples cured at 5
Appropriate concrete technology 216
C have surpassed the ones cured at 20 C. The samples cured at 30 C have similar or
higher strengths, but the rates of strength gain are descending. High early strength
cements show a similar tendency, but less marked. For OPC cements the same effect can
be seen (Figure 5).
It can be observed in these figures that blended cements show ascending curves with
temperature; i.e. the compressive strength is higher the higher the temperature, even for
curing time of 28 days. This behaviour is different to what is shown in the technical
literature about concrete. For OPC the effect is different (Figure 6); in this case the
tendency is similar up to curing time of 7 days, but for concrete samples cured at a given
temperature up to 28 days, the higher compressive strengths are found for lower curing
temperatures. This behaviour agrees with what is stated in the literature.
This is a very important difference in behaviour between concrete with OPC and
Blended Portland Cements, and which this research helped to show. The latter type of
cements are the most widely used in Chile. The authors also think that this effect is also
valid for concrete mixes with PFA, and other types of pozzolanic additions.
The strength-maturity curves for cements with additions show different slopes depending
on the curing temperature (Figure 10); i.e. the lower the curing temperature the larger the
slope of the strength-maturity relation of the concrete and for maturities larger than
approximately 20.000 C-hour the concrete present higher strengths when cured at low
temperatures than cured at high temperatures. For maturities smaller than 10.000 C-hour
the opposite behaviour is observed. It is important to note that concretes cured at 30 C
Behaviour in extreme climates of concrete made with different types of cement 221
during 28 days show the smallest slope (Figure 10). For OPC the effect of the
temperature on the strength-maturity relation is the same but the benefit of lower curing
temperatures is more noticeable even for smaller maturities (Figure 9).
Figure 11 shows that the freezing temperature of concrete and water were of the order of
0,4 and 1,5 C and fresh concrete freezes at approximately the same temperature as
water independent of the type of cement used. Therefore, it was apparent that the soluble
salts of the cement do not alter the freezing point of the water in the concrete, as it was
expected. The results also show that the cooling rate of fresh concrete is independent of
the type of cement used, but is smaller than the cooling rate of water alone. This is
probably due to the test conditions and because at very early ages the cements do not
liberate a sufficient amount of heat allowing to differentiate its cooling rates. Figure 11
shows that a decrease of the W/C ratio delays the time the concrete takes to freeze
(probably because of the larger proportion of cement contained in a concrete with a
smaller W/C ratio), but the freezing temperature does not change.
Appropriate concrete technology 222
CONCLUSIONS
The observed relation between curing temperature and mechanical behaviour of concrete
made with Portland Cement agrees well with results reported on technical literature.
However, for cements with additions the relation is different.
It was concluded that for Portland Cement the effect of the curing temperature on the
compressive strength shows an optimum at 20 C. For Blended Cements, the results
appears to indicate that the maximum strength is obtained for a higher curing
temperature. In the latter case, a higher curing temperature gives a higher compressive
strength for the same maturity, both for short and long term. The results are similar for
the two water/cement ratios analysed (0.45 and 0.60).
The maturity-compressive strength relations are lineal in a semi-logarithmic chart and
are particular for each concrete and cement utilized.
It was shown that fresh concrete freezes at the same temperature as water and that the
minimum temperature the concrete can reach is 0 C. Therefore, the minimum
temperature allowed for the fresh concrete in cold weather construction should be 5 C,
before applying the protection needed (insulation or heat). This will allow to obtain
higher long term compressive strengths and to avoid concrete freezing. Also, it should be
avoided to saturate the concrete during the early life; curing procedures should be
directed to prevent the loss of water by evaporation rather than wetting.
For hot weather concrete construction, the results show that the concrete placing
temperature should be as low as possible and the convenience to maintain the concrete at
low temperature during the early life of the concrete.
ACKNOWLEDGEMENTS
The authors would like to express their gratitude for the grant made by the Council for
Technological and Scientific Research (Project N 815) to carry out the reported work.
REFERENCES
1. AMERICAN CONCRETE INSTITUTE. Manual of Concrete Practice, Part 2. ACI, USA, 1985,
pp. 305R-1305R-16 and 306R-1306R-22.
2. INSTITUTO NACIONAL DE NORMALIZACION. HormignRequisites Generales, NCh.,
Santiago, Chile, 1985, pp 55.
3. NEVILLE, A M. Properties of Concrete, Pitman, London, 1983, pp 779.
4. PINK, V A. Winter Concreting, C&CA, London, 1967, pp 19.
5. VERBECH, G J AND HELMUTH, R A. Structures and Physical Properties of Cement Paste,
Proceedings 5th International Symposium on the Chemistry of Cement, Part III, Tokyo, 1968,
pp 132.
6. VIDELA, C. AND PARADA, O. Prediccin de la Resistencia Potencial del Hormign, Revista
de Ingeniera de Construction, 1988, Dec., N 5, pp 1734.
7. INSTITUTO NACIONAL DE NORMALIZACION. CementoTerminologa, Clasificacin y
Especificaciones Generales, NCh, Santiago, Chile, 1968, pp 20.
8. NEVILLE, A M. AND BROOKS, J J. Concrete Technology, Longman, London, 1990, pp 438.
PROBLEMS ABOUT OPTIMIZATION
OF POROSITY AND PROPERTIES OF
AERATED CONCRETE
R Cabrillac
Z Malou
IUSI University of Cergy Pontoise
France
INTRODUCTION
Generally the mechanical and thermal properties of porous materials, and particularly of
aerated concretes, depend directly on porosity. In the case of aerated concretes porosity is
created by a chemical process[1]. It consists in adding aluminium powder to the
constituents of the mortar (cement, lime, sand and water), which produces a gazeous
release which causes fresh mortar to expand. The quantity and the geometrical
configuration of pores depend on the quantity and on the nature of the constituents which
enter into the composition of the material [2].
The manufacturing process leads to the obtention of anisotropic materials which are
constituted by uniaxially flattened pores which are perpendicular to the direction of the
expansion. This anisotropy may be an advantage for the optimisation of mechanical and
thermal properties of aerated concretes[3] [4].
The primary aim of this work is to study the lightening phenomenon and the influence
of the introduced porosity on the mechanical anisotropy of the aerated concretes. The
introduced porosity depends on the parameters of composition which also act on
relativity of setting kinetic and expansion kinetic; the second aim of this study is to
correlate the relativity of setting and expansion kinetics with introduced porosity and
mechanical anisotropy.
For autoclaved aerated concretes, the autoclavage improves the mechnical strengths of
the matrix but it takes place in the process after the creation of the pores and it does not
have any effect on the introduced porosity. Thereby lightening phenomenon and
mechanical anisotropy due to geometry of pores are the same for autoclaved or not
autoclaved aereted concretes though mechanical strengths are different.
The principle of the study consists in the realisation of different not autoclaved aerated
concretes with different compositions and the corresponding basic matrix. We have
quantified the introduced porosity (PI) with apparent densities of expanded mixtures
(MVAe) and corresponding apparent densities of the basic matrix (MVAb): PI=(MVAb-
MVAe)/MVAb. The compressive strengths (RC) have been measured in the two
perpendicular directions to display the mechanical anisotropy. The compositions of the
basic matrix are given in the table. 1. The ratios of cement (C), lime (X) ,water (E) and
sand (S) are expressed comparatively with the weight of binder (L). For aereted concretes
the aluminium powder (AL) quantities added to basic matrix and expressed by ratios
comparatively with the weight of binder are : 0.01; 0.025; 0.05; 0.1; 0.2; 0.3; 0.4; 0.5;
0.75; 1; 1.5; 2.
Problems about optimization of porosity and properties of aerated concrete 225
Figure 1 shows the evolution of compressive strengths in accordance with the introduced
porosity. The compressive strengths are noted RCij ; the indice (i) corresponds to the
direction of the measure and the indice (j) to the composition. The evolution of
compressive strengths in accordance with porosity shows that the strengths measured in
the direction of the expansion of the material are lower than the strengths measured in the
perpandicular direction.
This difference of strengths which shows the mechanical anisotropy begins for the
first values of the porosity and the difference increases with porosities between 0.3 and
0.45
For the porosities greater than 0.45 the difference decreases and the material tends to
be isotropic.
Methodology
In order to characterize the kinetics of setting and expansion, we have measured the
height of swelling (G), and evaluated the setting by the driving (E) of the Vicats needle
(d=13mm), in accordance with the time for some compositions which have the same
matrix but different expansive agent dosages. To characterize the relativity of the two
kinetics, we have considered the time TO corresponding with the point of intersection of
the two curves obtained. After that, we have investigated for two correlations: the first is
between the mechanical anisotropy characterized by the ratio RC2/RC1 and the relativity
of kinetics of setting and expansion characterised by the time TO; the second is between
the introduced porosity and the relativity of kinetics of setting and expansion.
On figure 6 we see that the introduced porosity increases very quickly for the weak
expansive agent dosages, and it is stabilized for dosages about 0.75 %. The figure 7
shows that the mechanical anisotropy increases also with the expansive agent dosage but
it decreases for the expansive agent dosages greater than 0.75%.
Figure 8 shows the relativity of kinetics of setting and expansion in accordance with
the expansive agent dosage.
The results obtained show that when the expansive agent dosage increases, the time
TO characterizing the relativity of kinetics of setting and expansion decreases. This
phenomenon is explained by the swelling speed which increases when the expansive
agent dosage increases. The figures 9,10 respectively show the evolution of the
introduced porosity and of mechanical anisotropy in accordance with the relativity
kinetics of setting and expansion.
The evolution of introduced porosity in accordance with the time T0 shows that when it
increases the introduced porosity decreases linearly. This phenomenon is due to the
slowness of the expansion comparatively with the quickness of setting and that prevens
the development of the porosity. Concerning the mechanical anisotropy characterized by
the ratio RC2/RC1 we observe an optimal time T0 of about 120mn which permits to
obtain for this composition a maximal anisotropy.
We observe that the kinetics relativity which is according to the maximal anisotropy
doesnt give the maximal porosity.
Indeed the evolution of the mechanical anisotropy in accordance with the porosity
(Figure 11) shows that the maximal anisotropy is obtained for porosities about 0.3.
For greater porosities, the anisotropy decreases and the material tends to be isotropic.
Finally the introduced porosity and the mechanical anisotropy in aerated concretes are
very dependent and their simultaneous optimization requires the mastery of the relativity
of kinetics of setting and expansion.
CONCLUSIONS
This study clearly showed the lightening phenomenon which is shown by the evolution of
mechanical properties in accordance with porosity. In the same way it showed the
relation existing between the quantitative and qualitative aspects of the introduced
porosity; indeed the maximum mechanical anisotropy due to the flattening of the pores is
obtained for porosities between 0.2 and 0.3. Moreover we established that the
phenomenon of the mechanical anisotropy depends effectively on the relativity of
kinetics of setting and expansion and that there is an optimal value of this one which
permits to obtain a maximal anisotropy. On the other hand this optimum does not give the
maximal porosity. This result is in accordance with the evolution of the mechanical
anisotropy with the porosity. The results obtained in this study have highlighted the
importance of the parameter T0. This time which characterises the relativity of kinetic of
setting and expansion determines the developement of the porosity and its geometrical
configuration, and seems to be a fundamental parameter to obtain materials more
performant from a mechanical and thermal point of view.
REFERENCES
[1] RILEM Technical Committes 78-MCA and 51-ALC. Autoclaved Aerated Concrete,
properties, testing and design. E and F N SPON, CHAPMAN and HALL
[2] Z.MALOU, R.CABRILLAC, S.AGGOUN, C.ROCHELLE. Influence des paramtres de
composition sur la porosit et les caractristiques mcaniques des btons
cellulaires.Publication Annales de lInstitut Technique du Btiment et des Travaux Publics Mai
95.
[3] A.BASCOUL, R.CABRILLAC, J.C.MASO. Influence de la forme et de lorientation des pores
sur le comportement mcanique des btons cellulaires. Communication Congrs International
RILEM, VERSAILLES, Septembre 1987
[4] R.CABRILLAC, B.PERRIN. Influence de la forme et de lorientation des pores sur la
conductivit thermique des milieux poreux Communication Congrs International RILEM,
VERSAILLES, Septembre 1987
THE USE OF FOAMCRETE FOR
AFFORDABLE DEVELOPMENT IN
THIRD WORLD COUNTRIES
E P Kearsley
University of Pretoria
South Africa
applications and the use of fiber reinforcing and lightweight concrete for affordable
development.
INTRODUCTION
According to the White Paper on Housing South Africa is characterized by large scale
unemployment in the formal sector of the economy. During 1995 40% of South African
households earned less than R800/month (approximately 150). A relatively small
formal housing stock, slow and progressively decreasing rates of formal and informal
housing delivery have resulted in a massive increase in the number of homeless people.
[3]
It is estimated that the urban housing backlog in 1995 was approximately 1, 5 million
units. Due to the high rate of population growth and the low rate of housing provision, it
is estimated that the housing backlog is presently increasing at a rate of 178 000 units per
annum [3].
The Reconstruction and Development Program currently underway in South Africa
aims at giving all South Africans access to A permanent residential structure with secure
tenure, ensuring privacy and providing adequate protection against the elements and
potable water, adequate sanitation and domestic electricity supply [3].
In order to address the needs of the country, modern specialized technology could be
used to improve the quality of building materials used in under developed areas. By
manufacturing lightweight building blocks, wall panels, roofing sheets, footings and floor
slabs on the outskirts of undeveloped areas, the local communities could use their own
transport to obtain these building materials. Lightweight aggregate or Autoclaved
Aerated Concrete could be used to meet this need, but the cost of the aggregates and the
process exclude these possibilities. Foamcrete however is an affordable, transportable
building material.
Little has been published on the structural behaviour of foamcrete. Since 1992 tests on
the structural properties of foamcrete have been conducted at the University of Pretoria.
The effect of foaming agents, percentage foam and mix proportions of other constituents
are being investigated.
The objective of this paper is to discuss methods of optimizing the mix design of
foamcrete as far as strength is concerned, taking cost into account, while minimizing the
adverse aspects, such as shrinkage, to acceptable levels. Values obtained from
experimental work for different design parameters will be evaluated and possible uses in
the Southern African market will be discussed.
MANUFACTURING REQUIREMENTS
is a function of the volume of foam that is added to the slurry. To ensure that the desired
percentage air is entrained in the mix, pre-foaming, where the foaming agent is aerated
before being added to the mix, is recommended.
During the experiments it was found that the foamcrete is sensitive to the type of
mixer and the mixing speed used. When mixing foamcrete it is recommended that the
foaming agent be added last. The mixing process should be a soft process where the
foam is gently mixed in, to prevent the foam from breaking down when mixed
vigorously.
The density of the foam and the slurry are so different that it is difficult to obtain a
uniform foamcrete mix unless a mixer that forces integration is used. The time required
for mixing should be reduced to a minimum as the foam has a limited lifetime where after
it starts degenerating. If the mixing is too slow the cement paste will not have sufficient
curing time before the foam breaks down leading to insufficient strength for maintaining
the shape of the void formed when the foam breaks down.
MIX DESIGNS
Compressive strength, modulus of rupture and shrinkage tests have been conducted to
determine material properties. Compressive tests were conducted on 100 mm cubes after
7 and 28 days using a special cube press giving readings to the nearest tenth of a kN.
Modulus of rupture tests are conducted on 100 x 100500 mm beams in a closed loop
continuous measuring system. The shrinkage readings were taken on 250 mm shrinkage
beams. To eliminate temperature effects all specimens were kept under plastic covering
at 22 C and 60% humidity in a constant temperature room for 24 hours where after
curing took placed in constant temperature water baths.
To determine the effect of the foaming agent, tests where conducted on samples
consisting of only cement, water and foam The water demand for different percentages of
foam was obtained by slowly increasing the water/cement ratio to the point where no
visual breakdown of foam took place. This ratio was used as a basis and a range of mixes
both dryer and wetter than this mix was tested. For each of the foam contents an optimum
water/cement ratio, resulting in the highest strength, was obtained [9].
Cost is a major concern in the development of any infrastructure. The aim when
designing a foamcrete mix was therefore to optimize strength while minimizing cost.
Ungraded ash (PFA) is available at low cost in large areas of South Africa as a cement
replacement, which led to further tests on the use of ash in foamcrete [2].
Using mixes with fixed percentages of foam and ash/cement ratios, the optimum water
demand was determined as before. The optimum water/cement ratio for various
percentages of foam is shown as a function of ash/cement ratio in Figure 2. From Figure
2 it can be seen that the optimum water/cement ratio for all the different foam contents is
in the region of 0.38 when no ash is used. The required water/cement ratio increases as
the ash/cement ratio increases. This relationship is basically linear for low percentages of
foam, but with high percentages of foam the water demand increases to the extent where
it is not possible to prepare a durable mix with 60% foam and an ash/cement ratio of 3.
Appropriate concrete technology 236
The effect of slagment, chalk, sand, silica fume and superplasticizer on the water demand
of mixes with different percentages of foam was determined. The results obtained from
these tests are, to the extent that the limited space allows, discussed in the succeeding
paragraphs.
STRUCTURAL PROPERTIES
Compressive Strength
To obtain results unaffected by seasonal variation in temperature, all specimens made at
the University were normally cured in water. The results used in this paper are average
values and not characteristic values (fc). No trends in standard deviation was observed.
One would normally assume that wet curing would result in optimum strengths.
However tests conducted at our laboratory indicate that this does not hold true for
foamcrete as can be seen in Figure 3. Using cubes from one mix, groups where placed in
a constant temperature room at 22C in the open and in plastic bags, in water at 22C, in
The use of foamcrete for affordable development in third world countries 237
plastic bags and open in the oven at 50C. Cubes were removed from the oven after 2 and
7 days and placed in the constant temperature room for the remainder of the time. Some
of the cubes cured under water, were dried for 24 hours before crushing to compare dry
results. From these results it can be seen that it would be beneficial to cure foamcrete
products under plastic covers rather than having to use water.
Various materials were used with greater and lesser degree of success but eventually we
decided to use sand as an additional filler. Sand with a moisture content varying between
6% and 14% was initially used without drying. This moisture in the sand was sufficient
not to have to change the water demand of the mixes. The exact water demand of the
mixes were determined by using oven dried sand.
Some of the results obtained when mixing various sand/cement ratios at different foam
contents are shown in Figure 5. From these results it can be seen that the compressive
strength is significantly reduced by adding sand. It is interesting to note that there is no
further significant strength reduction with further increases in the sand/cement ratio.
From the results obtained there appears to be an optimum sand content at a sand/cement
ratio of between 0.5 and 1.
As the arm of the investigation is to produce foamcrete products that can be
manufactured by rural communities, it is essential to determine how sensitive the results
are to small changes in the mix. As the ash and the cement may be obtained in bags,
batching may occur per bag, thereby controlling the volumes and densities. The same can
not however be said for the sand and although a mix be designed for a specific sand
source the moisture content and the exact volume batched may vary.
From Figure 6 it can be seen that a variation in the moisture content of the sand or the
sand/cement ratio has a limited effect on the compressive strength of the mix A variation
in the foam content does however have a significant influence on the compressive
strength and during batching, strict control is required when adding the foam.
The use of foamcrete for affordable development in third world countries 239
Using an ash/cement ratio of 1 and a sand/cement ratio of 1, two thirds of the cement
initially used in the mix has been replaced. Would it now be possible to increase the
strength again by adding small quantities of admixtures? We have conducted tests using
limited volumes of condensed silica fume (CSF). Silica fume to the extent of 2.5%, 5%,
7.5% and 10% of the mass of the cement was used and the results obtained are indicated
in Figure 7. From the results it can be concluded that the silica fume does not have a
marked effect on the strength of the mixes with high percentages of foam, but for mixes
with low percentages of foam adding 7.5% silica fume results in compressive strengths
equal to that obtained before adding any sand.
The compressive strength can be increased even more by adding water reducing agents.
The water demand of a mix with a cement:ash: sand: silica fume ratio of 1:1:1:0.075 can
be reduced by up to 20% by adding super plasticizers. The results indicate an optimum
compressive strength may be obtained for the mixes with 0.75% super plasticizer. These
strengths are higher than that obtained with pure cement slurry mixes (ash/cement ratio=0
and sand/cement ratio=0).
From the results obtained to date there seems to be an optimum mix for a specific
percentage foam added. To find an explanation for the results obtained SEM (Scanning
Electron Microscope) samples have been prepared. Indications are that the foam causes
wall effects and the spherical shape of the foam bubbles results in fewer cement particles
situated next to the void [5]. When the foam disintegrates, a void is left behind allowing
the formation of large calcium hydroxide (Ca(OH)2) crystals. When condensed silica
fume (CSF) is added to the mix, the particle size is so small that the packing distribution
is not affected by the presence of the foam bubbles. When the Ca(OH)2 crystals forms in
the void, the CSF reacts with the Ca(OH)2 to form a much stronger tobermorite gel [4].
Shrinkage
Shrinkage tests conducted by us to date indicate that the shrinkage is directly related to
the percentage foam added to a mix. Foamcrete shrinks between 4 and 10 times as much
as normal concrete. The shrinkage for a specific foam content can be reduced from that
obtained for a simple cement mix by replacing some of the cement with ash and sand and
reducing the water content of the mix (by adding a superplasticizer).
Values obtained during shrinkage tests are indicated in Figure 8. These values indicate
that the shrinkage is directly related to percentage foam added. For 30% and 60% foam
the shrinkage can be significantly reduced by limiting the cement content.
The use of foamcrete for affordable development in third world countries 241
Foamcrete seems to be more suitable to the manufacture of small building elements due
to the relatively high shrinkage and low strength of the material. As the material shrinks
and expands due to variation in moisture content provision should be made for possible
movement or precaution should be taken to prevent changes in moisture content. The
most obvious and general use for foamcrete is for building blocks. Tests have been
conducted and based on initial results walls built using foamcrete blocks act similarly to
those built using ordinary bricks. As blockwork depends on weight when resisting
horizontal forces, care should be taken that blocks are not too light to resist forces. The
loss in weight as a result of drying can be up to 25% of the initial weight, and this lower
density has to be taken into account in design.
Foamcrete does not have a high resistance to abrasion. When using foamcrete in an
abrasive environment it is recommended that composite construction is considered,
where a strong outer layer protects the inner layer. This method could prevent moisture
absorption by the foamcrete and therefore prevent swelling and shrinking. Foamcrete
lends itself to precision moulding. As foamcrete has a high percentage of air entrainment,
it tends to be extremely workable. Where alterations in shape needs to be made after
casting, the material can be sawn and nailed.
A small low cost housing unit, using tilt-up construction, has been developed at the
University. Initial indications are that the material cost for the unit is comparable to that
of corrugated iron. The construction cost should be similar to corrugated iron but the
thermal advantages of the foamcrete could have far reaching implications for the housing
market. The main advantage of this method is that no highly skilled carpenters,
bricklayers and builders are required as the walls are cast on the floor and a normal
straight edge can be used to ensure a neat surface.
Appropriate concrete technology 242
Tests have been conducted on blocks to be used as fire walls. The results obtained
from these tests were encouraging as after two hours in a fire burning at more than 500C
and being hosed down, only minor hairline cracks on the surface of the blocks were
visible. Experiments are currently underway for developing roof tiles, manhole rings,
drainage channels, meter boxes, coolrooms, wall units and building blocks. All these
products could be used in the Reconstruction and Development Program currently
underway in South Africa.
The next phase of the research will include creep and durability tests as well as tests
on the use of reinforcing and the bond thereof. The limited tests conducted to date
indicate that durable precast foamcrete elements, manufactured by rural communities,
could be designed for use not only in houses and infrastructure but as structural members
in double storey schools and clinics.
REFERENCES
1. ADDIS, B J. Fultons Concrete Technology (seventh edition). PCI, South Africa, 1994.
2. CHITHARANJAN, N, RAMAKRISHNAN, V S and GANESAN, S. A New Production
Technology for the Manufacture of Cellular Concrete. Indian Concrete Journal, October 1978,
pp 266271.
3. GOVERNMENT GAZETTE; REPUBLIC OF SOUTH AFRICA. The White Paper on Housing,
December 1994.
4. ILLSTON, J M. Construction Materials: Their Nature and Behaviour (second edition). E & FN
Spon, 1994, pp 91111.
5. NILSEN, U, SANDBERG, Rand FOLLIARD, K. Influence of Mineral Admixtures on the
Transition Zone in Concrete. Proceedings of the International Conference on Interfaces in
Cementitious Composites, October 1992, pp 6570.
6. TAYLOR, W H. The Production, Properties and Uses of Foamed Concrete. Precast Concrete,
February 1974, pp 8396.
7. VAN DEIJK, S. Foam Concrete. Concrete, July/August 1991, pp 4954.
8. VINE-LOTT, K M. Structural Applications of Foamed Concrete. Concrete, February 1981, pp
2425.
9. WIDMANN, H and ENOEKL, V. Foam ConcreteProperties and Production. Concrete
Precasting Plant and Technology, Issue 6/1991, pp 3844.
SOME ASPECTS OF THE DESIGN AND
PRODUCTION OF FOAMED
CONCRETE
D E Wimpenny
SirWilliam Halcrow & Partners
UK
INTRODUCTION
Foam concrete is a relatively light-weight, low strength material formed by entraining air
in a sand-cement mortar. The air is entrained by the addition of either chemical admixture
or preformed foam to the base mortar mix.
In situ foam concrete is used in a number of applications which exploit its high
workability, low density, low strength and high thermal insulation characteristics. The
design of the base mortar and the entrained air can be adjusted to give plastic densities in
a typical range of 400kg/m3 to 1600kg/m3, 28-day compressive strength values between
1N/mm2 and 10N/mm2, and thermal conductivity values between 0.1 and 0.7 W/mC [1].
The main uses of the material are in road reinstatement and underground void filling.
Other applications include as a fill material in ground engineering and as a lightweight,
insulating screed [2].
Until relatively recently the production of foam concrete in the UK was very limited,
with Pioneer Concrete (UK) Ltd being one of only three major concrete suppliers to offer
the product [3]. However, changes to the building regulations in 1990 and the
specification for road reinstatement in 1991 have led to an increased interest in the
material to speed road reinstatement and enhance the thermal insulation of new buildings
[4,5]. Other recent uses of the material in the UK include emergency ground stabilisation
after the Heathrow Express tunnel collapse and as a lightweight backfill to create an
approach embankment to a bridge [6].
This paper summarises an extensive study of the design and production of foamed
concrete undertaken by the author in 1989 and 1990 for Pioneer Concrete (UK) Ltd. The
paper concentrates on the laboratory work, although some practical applications of
foamed concrete are briefly described.
The objective of the study was to examine current methods of design, production and
utilisation of foamed concrete in order to increase the commercial potential of the
material. Within this broad objective it was hoped to give guidance on the optimising the
base mortar mix and the addition of the foam element in order to improve the strength
and cost effectiveness of the material.
A large number of factors were encompassed within the laboratory work, including: 3
foaming systems, 5 foaming admixtures, 7 fine aggregate sources, and 6 aggregate-
cement (a/c) ratios.
In addition to the above, two cementitious types and curing regimes, as well as a large
number of plastic densities and water-cement (w/c) ratios, were investigated. The density
and compressive strength of the foam concrete were the key properties measured.
Limited data on the workability and air content of the fresh concrete were also recorded,
although these data are not reported in detail here.
Some aspects of the design and production of foamed concrete 245
LABORATORY MIXES
MATERIALS
Portland cement complying with BS12 and ggbs complying with BS6699 were used. The
main sources of fine aggregate used were a natural sand from Hogshead Quarry, having a
BS882 limits F grading, and a 50/50 blend of this sand and limestone fines from
Aberduna Quarry, having a BS882 limits M grading.
The main foaming admixtures used were F292 and F294 from Fosroc. Limited work
was also carried out with Hex and Regular Protein from Chubb, and Poro BG from Sika.
All these agent are synthetic surfactants, with the exception of Regular Protein from
Chubb which is based on natural proteins [7]. These agents were used in air and water
pressure driven foaming systems as recommended by the manufacturer and described
below. There was also a limited use of an accelerating admixture, Conplast NC, from
Fosroc.
MIXES
The constituents for the base mortar were batched to within 0.5% and mixed for five
minutes in a 20litre capacity horizontal axis drum mixer. The densities of the base mortar
mix and foam were measured using calibrated containers and from these values the mass
of foam (Fm) to be added to the mix was calculated as,
(1)
where, Bm=base mix mass, Fd=foam density, Td=target density, Bd=base density.
Freshly produced foam was folded in to the base mortar mix using the theoretical dose
as a guide until the approximate target density was attained. Following fresh concrete
tests, two 100mm cubes were cast from each mix for testing at 7 and 28 days. These
cubes were cured in tap water at 20C in accordance with BS1881, Part 111. Where
compressive strength values of less than 2N/mm2 were anticipated a triaxial loading rig
with automatic loading and a 20kN proving ring was used instead of a standard
compression testing machine.
The proportions of the mixes produced have been summarised in terms of the a/c and
w/c ratios. The latter is the nominal free w/c ratio, allowing for surface moisture on the
aggregate, but excluding any moisture in the added foam.
Appropriate concrete technology 246
DISCUSSION
used vary significantly in fineness, with Waddington Fell Coarse meeting limits C of
BS882, and Lytham, St Annes sand being finer than limits F of BS882.
If fineness is summarised in terms of the percentage passing a 600 micron sieve, then
there is a general increase in strength with fineness. For example, at a plastic density of
1000 kg/m3 Waddington Fell Coarse with a percentage passing of 53% yields a strength
of approximately 0.5N/mm2, whereas Hogshead sand with a percentage passing of 96%
produces a strength of over 2N/mm2.
This trend, which appears to be maintained up to a plastic density of around 1400
kg/m3, is consistent with the findings of others [4]. At plastic densities above 1400 kg/m3
coarser graded sands appear to yield strength values equivalent or higher than those
produced by the finer sands.
Whilst the use of percentage passing the 600 micron sieve is a convenient and
commonly used parameter for judging the suitability of sand, it is clear from the results
for Newark and Waddington Fine sand that this approach has limitations.
Taylor has reported grading envelopes for sand to be used in foam concrete production
[4]. These envelopes indicate that the preferred sand would have a percentage passing the
300 and 600 micron sieves of 3580% and 50100% respectively. It is interesting to note
that all the sands used in this study meet this criteria, with the exception of the Lytham
sand which, at the same time, gave rise to the highest strength values.
It can be observed that at 7 days the compressive strength of the mixes containing 50%
ggbs is significantly below that of the 0% ggbs control mixes. For a plastic density of
1450 kg/m3 the strength of the 50% ggbs mix is only 1.5N/mm2 whilst that of the 0%
ggbs mix is approximately 3N/mm2. In contrast, at 28 days the strength position is
reversed. For a plastic density of approximately 1450 kg/m3 the strength of the 50% ggbs
mix is 3.8 N/mm2, whilst that of the control mix is approximately 3.5N/mm2.
The effect upon 7-day compressive strength of using Conplast NC at the
manufacturers recommended dosage rate is shown in Figure 4. It can be observed that
any beneficial effect upon strength of using the accelerator is marginal. The 28-day
compressive strength results also support this finding.
The above results, taken in combination, suggest that modification of the early
strength is best achieved using different mineral admixtures or cementitious types rather
using chemical admixtures. Any admixtures, regardless of type, should of course be
checked for compatibility with the foaming system and agent before use.
PRODUCTION MIXES
GENERAL METHODOLOGY
The Fosroc foaming system and foaming admixture F294 were used throughout the work.
The Fosroc foaming system required a 110V supply. This was provided by a portable
generator which was also used to power electronic scales used for density measurements.
Some aspects of the design and production of foamed concrete 251
A limits M natural sand was used and the base mortar mix was batched to a high
workability, consistent with the laboratory work.
The foam dose was calculated as volume based on the target and base mortar mix
densities in a similar fashion to Equation 1 and then converted into a time using a typical
discharge rate in litres per second for the system. The foam addition typically took
approximately 5 minutes, followed by 5 minutes of mixing to thoroughly entrain the
foam.
The project involved encasing a new gas main in foamed concrete reinforced with steel
fabric. A total of 135m3 of foamed concrete was supplied. Much of the pipe was laid
within the highway using a relatively narrow trench. Concrete was simply placed into the
trench using the mixer truck chute. In order to reduce nuisance spillages onto the
highway, the client provided a purpose made, funnel-shaped extension to the chute.
A portion of the pipe was laid across a school field and this necessitated pumping the
concrete horizontally for over 30m. A very small capacity Sydewinder pump was
successfully able to pump the concrete over the distance required. Due to its flowing
nature it was possible to move large volumes of the foam concrete along the trench using
a spade, thereby reducing the distance for pumping.
The details of the mix are as follows:
a) a/c ratio of 3.0 and w/c ratio of 0.65
b) target density and 28-day compressive strength 1600 kg/m3 and 89N/mm2
c) actual density 1647 kg/m3 (standard deviation 85 kg/m3)
d) actual 28-day compressive strength 9.1 N/mm2 (standard deviation 2.2 N/mm2)
CELLAR FILLING
Approximately 16m3 of foamed concrete was delivered to fill a redundant cellar. The
concrete was discharged into the cellar through a ground level access cover. Foamed
concrete was chosen because of the low load it would impose on the foundations and
walls. The details of the mix are as follows:
a) a/c ratio of 4.0 and w/c ratio of 0.75
b) target density 1400 kg/m3 and minimum 28-day strength 2 N/mm2
c) actual density 1395 kg/m3 and 28-day strength 5.4 N/mm2 (min. 3.5 N/mm2)
CONCLUSIONS
Laboratory studies have shown that the compressive strength at any density can be
optimised by careful selection of the constituent materials and the base mix proportions.
Appropriate concrete technology 252
The water pressure driven foaming system and agents from Fosroc, which were used
for the majority of the laboratory work and in the production mixes, proved reliable.
However, for cooler climates, where ambient temperatures less than 10C may be
encountered, F294 should be used instead of F292. Protein based foaming admixture
yielded slightly higher strength values than synthetic agents.
Higher compressive strength values can generally be obtained using a finely graded
aggregate or relatively high w/c ratio in the base mix. Where strength and cost are the
only criteria, mixes with high a/c ratios and plastic densities will be favoured. The early
strength was found to be influenced more by the use of a mineral admixture (ggbs) than a
chemical admixture (accelerator).
ACKNOWLEDGEMENTS
The research was carried out on behalf of Pioneer Concrete (UK) Ltd and the support
provided by Pioneer Testing Services, Warrington, is gratefully acknowledged.
REFERENCES
1. WALKER B AND CLARKE A. Introducing foam concrete. Concrete Quarterly, 1988, Winter,
pp2425.
2. VAN DEIJK S. Foam Concrete. Concrete, 1991, July/Aug., Vol 25, No 5, pp 4954.
3. MIDDLEBOE S. Fast Foam Filler. NCE., 1988, 22 Sept, pp 4348.
4. TAYLOR R W. Foamed Concrete: For improved trench reinstatements. BCA Publication
46.043, 1991.
5. ANON. Warming up to Change. New Builder, 1989, 16 Nov., pp18.
6. PARKER D. Foam Fitting. NCE. Concrete Supplement, 1995, July, pp46.
7. MOORFIELD G. Filling a gap in the market. Concrete, 1994, May/June, Vol 28, No 3, pp 12
14.
8. GUTMANN P. Bubble characteristics as they pertain to Compressive strength and freeze-thaw
durability. ACI Materials Journal, 1988, Sept-Oct., pp 361366.
BUILDING UNDERGROUNDTHE
CONCRETE CONTRIBUTION
M A Clarke
British Cement Association
UK
INTRODUCTION
This paper briefly reviews developments in the role of concrete underground, drawing on
recent examples. There is renewed interest throughout the world in the possibilities of
building underground roads, utilities, retail and office developments the author hopes at
the conference to stimulate interest in a new forum and databasea British
Underground Space Centre.
Historical Precedent
There is very little that has not been tried in civil engineering under the soil in the past
hundred years.
The means and methodology of construction over this timescale, of course, has
changed dramatically. Take the dawn of the modern tunnelling era, the Thames Tunnel
built by Marc Brunel. It took hundreds of men, horses, winches and hand pumps working
night and day moving at a rate of 7ft a week, it took five years to dig out half the tunnel
length of 600ft. It took a further 10 years to finish it, in between two major breaches of
the tunnel when the Thames poured into it!
What was the point of it all? The tunnel was needed to reduce the journey time and
alleviate congestion of wheeled traffic over London Bridge and also to reduce the number
Appropriate concrete technology 254
of ferry boats milling on the river blocking the progress of cargo clippers and sailing
barges. Around 1820 some 4,000 slow moving, horse driven carts and wagons crossed
London Bridge daily whilst 350 Thames watermen ferried 3,700 people across the river
nearby. Some congestion and some unmentionable pollution to go with it. Why was there
such a build up of traffic?
Rotherhithe on the south side of the Thames was the industrial hub of London,
flourishing with industry and enterprise, packed with mills and wharves, factories and
working people. Unfortunately the heart of Londons docklands was situated at St
Katherines, and at the London and the West India Docks on the north side of the
Thames. The docks took in vast quantities of foreign goods, some of which had to make
the four mile journey by road via London Bridge to get to Rotherhithe, to the factories
and warehouses. The bridge was not wide enough to cope with extra traffic. Building a
new bridge would further disrupt the river traffic and exacerbate road congestion for
quite some time. Brunels tunnel idea was an inspiration.
However, the economic failure of Brunels tunnel, the horrendous time it took to
complete, the catastrophes during construction and its relatively short life thereafter, had
a sobering effect on Londons enthusiasm for digging under the Thames. It was not until
the year 1870 that another passage under the Thames was attempted, when the illustrious
James Greathead, a South African born civil engineer, offered to excavate a tunnel under
the Thames from Tower Hill to Vine Lane for The East London Railway Company for
16,000 in less than a year! A lot less time than Brunels attempt which took 15 years and
614,000. The Greathead shield weighed only 2 tonnes compared to the Brunels shield
which weighed 120 tonnes. It advanced at four and a half feet a day. The tunnel linings
were eighteen inch thick cast iron segments, which were grout packed tight using hand
syringes. The Greathead shield was the forerunner of the modern tunnelling machines.
Developments in subterranean civil engineering today are a worldwide phenomena.
Beneath the skylines of London, Los Angeles, Lisbon, Lille, Leningrad, Lesotho and
Lausanne lies an underworld city of buried rivers and sewers, railways, pipes and
passages, tubes and tunnels. They weave and wind their way to run their urgent errands
carrying people, delivering water, removing sewage, conducting power, piping gas,
sending messages and conveying parcels. They are the arteries of the modern city, an
essential network that imperceptibly controls and maintains the equilibrium and
continuity of an ever-changing and volatile world above ground.
Trenchless technology, cut and cover excavation, the slurry diaphragm wall, the
immersed caisson, augured piling and thrust-bore are just a few of the many methods
available today to create holes in the ground as big as a football pitch or as long as the
M25. Once formed, these holes and caverns have a myriad of uses, but they all have to be
supported in perpetuity, absorbing the external pressures and strains, the differential
subsoil movement, resisting degradation from mineral salts and preventing water ingress.
The materials for construction have evolved to respond to the demand for ease of
assembly, local availability, economic cost and the variable engineering properties, eg
fast setting, retarded set, high early strength, mobility, high durability and toughness.
Concrete in all its versatile forms has become the material of choice in subterranean civil
engineering today.
The world market in trenchless technology in 1992 was conservatively estimated at
$3,650 billion and that excludes the large diameter tunnel market, anything above 2.5m
Building underground: The concrete contribution 255
diameter. For every tunnel contract using tunnel boring machines, it is reasonable to
assume that 25% of the contract sum is the actual cost of the concrete used. The Jubilee
Line Extension for instance has a budget of 3.3 billion over three years. The value to the
concrete industry of this project is around 800 million. On the River Medway immersed
tunnel contract now under construction, over 65% of the total contract sum will be spent
in precasting just the 120m long, 28,000 tonne concrete tunnel units.
Some market, some concrete.
technological development and research investment for such a market has been most
worthwhile.
The Chunnel
Just a little over 2.6 million cubic metres, or 6.2 million tonnes of concrete, enough to fill
Wembley Stadium to the height of the Eiffel Tower, was used in the construction of the
Channel Tunnel. Half a million tunnel lining segments weighing up to 8 tonnes each,
were manufactured on the Isle of Grain and transported to site over a three and a half year
period.
High quality, high strength and fast manufacture were the prime requirements for the
production programme. The tunnel linings were built with a design life of 120 years to
withstand the saline environment under the channel. Granite aggregate from Glensanda in
Scotland were transported to the Isle of Grain, for producing the 60N/mm2 concrete of
40,000 cubic metres of concrete were mixed at Shakespeare Cliff batching plant and
supplied to the tunnels on rail mounted re-mixer drums of 6 and 8 cubic metre capacity.
The concrete was transported up to 22km away, weaving its way on rail mounted
mixer drums. It presented a major challenge to the concrete technologists in designing
mixes to satisfy these onerous working conditions. To start with, all the concrete was
designed with a superplasticiser to keep the water cement ratio to a minimum, some were
over-dosed to retard the initial set if they were transported long distances. The concrete
was designed to meet the varying demands of low water content, slow hydration, high
workability, high early strength, low heat gain.
The crossover tunnels, a vast cavernous space for allowing trains to cross over to the
other track, was excavated using the NATM (New Austrian Tunnelling Method). Once
the cavern shape was excavated by backacter, the tunnel face had to be rapidly lined with
sprayed concrete, before a cylindrical shutter 5.5m long was positioned. The half cylinder
Building underground: The concrete contribution 257
over the tunnel arch was reinforced. Concrete was then placed to fill the void between
shutter and the sprayed concrete liner. The gap between form face and sprayed liner was
over 1 metre thick and needed 200 cubic metres to fill, taking 15 hours to complete. The
cylindrical shutter had to be ready for the next section within 36 hours of casting, so the
concrete had to have a guaranteed minimum strength of 10N/mm2 within 18 hours of
placing. It did not prove to be a problem. What to do with all the excavated spoil? The
5.5 million cubic metres of tunnel debris was used to reclaim thirty hectares from the sea
opposite Shakespeare Cliff, thus creating an amenity for uses linked with the Tunnel, So
before things got under way, a seawall 1,700m long was built with massive bund walls to
stop the sea from breaking through. Two rows of sheet piles spaced 10m apart were
driven into the sea bed to form the sides of the wall. The gap was filled with concrete to a
height of 12m. It took three years and 210,000 cubic metres of concrete to complete the
bund. Self-compacting concrete was transported nearly 1,000m to reach the discharge
point below water level. A cohesive concrete suitable for pumping with a high
cementitious content was necessary. To keep hydration temperatures under control 40%
replacement pfa was incorporated in the mix. To complete this part of the project, a
revetment and wave wall apron 9m wide was built along the-length of seawall bund
consuming a further 10,000 cubic metres of concrete.
In laying the rail tracks, a pair of independent precast concrete support blocks were
placed under the track at 600mm intervals over the entire length of the two running
tunnels. Concrete in all its forms, and in all its engineering properties has been used to
maximise efficiency of construction. The concrete both shields and protects the myriad of
structures required to safeguard the investment in this massive enterprise.
route would disrupt city traffic and because there was simply no room left near the
surface to build additional pipelines.
The feasible option was to construct a new ring main from Coppers Mill in East
London and form the southern link to Ashford, in the west, via Battersea and Surbiton.
The northern link would run via Holland Park, Hammersmith and Kempton before
connecting to Ashford. The encircling tunnel will be connected to a number of water
treatment works in the catchment area through the existing network. The ring main will
be used to shunt water from one network to another, to balance the distribution network
and rationalise total demand into the next century. Water from the ring main will be
pumped directly into the distribution network through 12 strategically located shafts.
Each one will be 12m in diameter and will be capable of handling flows sufficient to
sustain a population of 200,000.
The project is valued at 260 million and will utilise 280,000 tonnes of concrete when
complete. 80% of the concrete is in the form of unreinforced precast concrete wedge
block liningfirst pioneered in London in the late 50sand segmental tunnel linings in
the distribution shafts.
out water. The tunnels were ballasted in such a way that it had a net buoyancy of 100
tonnes when partly flooded and attached to the two large buoyancy pontoons. The units
were towed into position by a team of tugs, then the pontoons were flooded and slowly
but steadily the massive concrete sections were lowered into place adjacent to the
previous unit.
Once in place on the canal bed foundation, the tunnel unit is drawn hard against the
adjacent section using hydraulic rams. The gasket of the tunnel units form a watertight
seal under hydrostatic pressure, when the new tunnel unit is dewatered. The bulkhead of
the connecting units is broken out and the roadway is then constructed for the next 90m
section.
Immersed tunnels have been used to bridge canals over roadways, to build roadways
across airports, rail and metro tunnels under rivers and waterways. Some are 24m wide
units, some are 50m wide and contain an eight-lane carriageway for mixed transport use.
Probably the longest tunnel section immersed in one piece was the 268m long segment
for the Hem Rail Tunnel under Noordzee Canal near Amsterdam.
The Future
The BCA is keen to encourage the formation of a new group to research and promote the
development and growth of underground construction of all types in the United
Kingdom. Any interested parties should contact the Author.
ACKNOWLEDGEMENTS
The assistance of David Bennett in researching the background to this paper is gratefully
acknowledged.
Appropriate concrete technology 260
REFERENCES
Leader Paper
INTRODUCTION
The future is always hard to predict. Most of the laws of a social, environmental and
technological change are based on the principles which are expressed by the equations of
non-linear dynamics, or by so called Chaos theory [13].
New types and raw materials and radically new manufacturing methods 263
It is apparent from the mathematics of this theory, that a very minor change in the
initial state of a non-linear dynamic system can lead to the enormous changes in a state of
this system, even over a relatively short period of time. For example, the factors
influencing weather patterns are the components of a classical non-linear dynamic
system, and, as we know, the reliable weather forecasts seldom exceed three or four days.
There are nonetheless several factors influencing the future of humanity which
undeniably exist, which are detectable and even measurable with some degree of
accuracy [4,5]. These include:
1. The continuous growth of the worlds population, which, from a current estimate of
just under six billion, is increasing by an average of 245,000 people daily (there are an
average of 386,000 birth and 141,000 death every day in the world). The estimates of
the earths population by the year 2050 range from eight to twelve billion [6].
2. The limited area of arable land. The total land area is 135,973,730 square km, and thus
the average current population density is 44 persons per sq. km. The important
problem for sustainable land development, which is related to the future food supply,
is to some extent the land surface can be more completely utilised, while supporting an
increasingly higher population. It is estimated that at present about one-third of the
earths total land area is capable of being cropped, but the main limiting factor remains
the lack of water. In the early 1970-s only about one-fifth was so used, of which
approximately ten per-cent was irrigated. It should be noted, that in many parts of the
world continuous irrigation leads to the significant increase in soil salinity, and hence
to a potential loss of arable land.
3. The limited reserves of fossil fuels. Total reserves of fossil fuels in the world are
currently estimated at approximately 11.3 teratonnes (i.e. 11.3 1012 tonnes) of coal
equivalent, of which (at the current consumption rate) coal reserves of 8.5 teratonnes
should last approximately 220 years, and the oil reserves of 0.5 teratonnes could last
about 30 years. Possible discoveries of new reserves may extend this estimate by
another 10, or 20 years at the most.
There are additional reserves of gas of about 0.3 teratonnes and of the fuels
extracted from oil shales and tar sands, which may be as high as 2.0 teratonnes of
coal equivalent. These may replace oil, when it runs out. However, if the demand
on energy from fossil fuels will increase proportionally to the growth of
population, the reserves of all fossil fuels will last significantly less than 200
years,a very likely scenario, judging by the current trends, and,
4. The inevitable increase in the environmental pollution, resulting from an uncontrolled
population growth. Specific to cement industries, the main effects on water pollution
are: high probability in the increase of pH, and the potential danger to aquatic life.
However, in some cases an increase in pH of water may actually be beneficial, as it
may neutralise the harmful effects of acid rain. Other factors include discharge of
warm water, resulting in the reduction of dissolved oxygen in rivers and lakes and in
the abnormally high rate of plant growth or eutrophication (e.g. blue-green algae).
The most likely effects on air pollution are: increase in greenhouse gases,
particularly in the carbon dioxide, as well as increase in the concentrations of
oxides of sulphur, nitrogen, etc., resulting in an acid rain, photochemical smog,
etc..
Appropriate concrete technology 264
As the growth of these factors approaches their limiting value, the catastrophical failure
of the system which relies in its performance on these factors, becomes inevitable.
Thus, the system must adapt to the change, in order to survive.
Manufacturing of hydraulic cements is influenced by all four of the above factors. The
growth of population represents an increased demand on building and construction
materials, and thus on the output of cement; there is already a fragile balance in many
parts of the world between the land allocated to forests, farming and quarrying; the use of
fossil fuels in cement industries accelerates the depletion of these non-renewable energy
resources and contributes to the build up of greenhouse gases in the atmosphere.
Therefore cement manufacturing industry, as any other, must implement a method of
sustainable development, in order to achieve at least some degree of control of the
factors, mentioned above.
Sustainability is the ability to maintain a desired condition over time [7,8].
Sustainable development is a tool for achieving sustainability, not a desired goal.
New cements, manufactured by the new technologies should also be evaluated for the
way in which the end products made from these cements interact with the environment.
Let us consider some of the above technological changes one by one.
Aluminous cement, of course, is known to have durability problems in warm and moist
environment.
There are other types of iron-rich cements, such as Ferrari cement, in which the raw
meal is selected so as to replace most of aluminium oxides with those of iron. This
cement has lower energy consumption to fire, by comparison with Portland, and the
clinker, although it is harder than that of Portland cement, requires less energy to grind.
Other iron-rich cements are made by replacing, for example, most of silica oxide with
iron. This cement is produced from raw material of Portland cement type to which some
bauxite is added. The properties are similar to a rapid-hardening Portland cement.
Several experimental cements, one with high proportion of iron ore in the raw meal
and another produced by firing limestone with pyrite ash are currently being evaluated
for the long term behaviour.
Another interesting new development is the replacement of argillaceous component of
the raw meal with basalt. Basalts contain higher percentage of iron and aluminium oxides
than clays and shales. The firing temperature can be reduced to 1300C, and the phase
composition can vary over a considerable range, but generally the content of tricalcium
aluminate and tetracalcium alumino-ferrite is higher than in Portland cement at the
expense of either tricalcium or dicalcium silicates.
Extensive laboratory evaluation so far, indicates good strength characteristics, and
reasonable dimensional stability of basalt cements. These results must be confirmed by
the long term experimental and field data, including durability tests.
of clinker formation will depend mainly on the rates of diffusion and of heat exchange,
when these processes are slow.
In order to speed up the process of diffusion and of the mass transfer, it is necessary to
increase the flow of furnace gases past the raw meal particles, and to decrease the particle
sizes and/or to increase their porosity.
Kiln designers should be considering the combined effects of the rates of firing,
temperature, soaking times, kiln atmosphere, etc., as well as the cooling systems. Increase
in the rotary kiln efficiency requires modifications to the atmosphere and temperature, or
the implementation of completely different firing systems, such as fluidised beds, plasma
torches and electromagnetic beams.
These new types of kilns may require non-traditional raw feed, and produce clinkers
which differ from Portland both in phase compositions and in properties.
Some of the new cements, which were mentioned above, can be manufactured more
cost-effectively in these new types of kilns [911,23].
Portland cement in concrete, which also contains high range water reducing admixtures
(surfactants).
A completely different development of new binders has resulted from the application
of the fundamental principles of mechano-chemistry. Mechano-chemical activation of
seemingly inert siliceous materials resulted in the development of silica-water-
suspension-binders or SWSB.
It was shown, that concretes with SWSB, containing alumino-silicates, after hardening
and curing at temperatures of between 40C and 120C, achieved compressive strength of
about 60 MPa. Silica-water-suspension-binders are manufactured in specially designed
high intensity planetary mills.
It should be noted that slag-alkaline cement and SWSB based concretes have excellent
durability, comparable to that of naturally occurring igneous rocks [18 22,26].
Rejuvenated Cement
Currently, large amounts of materials from demolished buildings and structures are
recycled. This is particularly true for concrete.
It has been shown, that a considerable proportion of cement grains in concrete, which
was in service in buildings and structures for very long periods of time, remain only
partially hydrated. Coarser grains, with the diameter of greater than 30 microns, and
particularly those in excess of 50 microns in diameter, usually contain a core of
unhydrated cement in the middle. Lime forms abut 20% of the hydration products in
Portland cement.
Thus, fine fraction of recycled concrete can be rejuvenated, using mechano-chemical
activation, to become a hydraulic cement in its own right (Patent pending) [14, 22].
CONCLUSIONS
The increasing world wide awareness of the Global problems of overpopulation, energy
crisis, and environment pollution facing the current, but particularly future generations,
will lead to the establishment of principles of sustainability.
To achieve sustainability in cement manufacturing, we must first of all enforce energy
conservation, i.e. replacement of non-renewable fossil fuels with renewable fuels or with
wastes.
New types of cements, and new methods of clinker making requiring less energy than
the conventional manufacturing methods should be developed. This should be done in
conjunction with the study of the effects these new technologies will have on the
environment. The study of environmental effects should not be confined to just the
dangers of air and water pollution during the clinker making process. It should extend to
evaluation of the interaction of materials and products made from these cements with the
environment.
The products made with new cements should be energy and cost efficient,
environmentally friendly, durable and, in some cases, capable of hazardous wastes
encapsulation [27].
Appropriate concrete technology 272
There are several new technologies and many new cement types which have been or
are being developed, and which could satisfy most of the above requirements. We must
put these to good industrial use as soon as practically possible.
REFERENCES
Keywords: Sulfoaluminate cement, low energy cement, high early strength, heat of
hydration, porosity.
Dr Ch.Malami is Chemical Engeneer and head of the analytical laboratory of the
Research Center. Her main research intersts are synthesis of new types of cements and
study of their performance, especially in aggressive environments.
Dr Th.Philippou is Chemical Engineer. He is manager of the Research Center. Main
interests in cement manufacturing, concrete durability, building materials.
P.Tsakiridis is Chemichal Engineer. He worked in Hellenic Cement Research Center as
graduate student.
Hydraulic behaviour of non-expansive sulfoaluminate cement 275
INTRODUCTION
The interest on the calcium sulfoaluminate (C4A3) based cements originated from the
expansive effect, which can be obtained from the hydration of the C4A3 phase and the
subsequent ettringite formation15. The synthesis of expansive cements is based on this
property. Expansive cement Type K contains suitable proportions of C4A3, C and lime
blended with portland cement.
Investigations of the hydration of sulfoaluminate cements have shown that appropriate
proportioning of C2S, C4AF C4A3 and C can produce cements with properties of rapid
hardening, high early strength as well as exceptionally high strengths at later ages, with
low or without any expansion68.
On the other hand the study of the formation of a clinker with the above mineralogical
compounds showed that it can be synthesized using usual raw materials in combination
with industrial by-products9,10, it can be sintered at temperatures about 150200C lower
than Portland cement clinker3,6,8,11, and it needs 50% lower grinding energy12; these are
features that characterize the sulfoaluminate cements as low energy or energy saving
cements.
For the above reasons, increasing attention has been paid to this type of cements, not
only as expansive agents, but also for uses similar to Portland cement, especially where
properties of rapid hardening and high early strength are needed.
In the present work a sulphoaluminate cement (SAC) was prepared in the context of a
project aiming at the preparation of energy-saving cements having hydraulic behaviour
similar to Portland cement (OPC). Based on the specific literature3,7,8,1216 a non-
expansive clinker was synthesized17. The cement prepared was compared to OPC
regarding its physical and mechanical characteristics.
EXPERIMENTAL DETAILS
The details of the preparation and composition of the sulfoaluminate clinker are described
in a previous work17. The clinker was ground to cement in a laboratory Bond-mill
without any addition of gypsum, requiring 2700 revolutions, in order to attain a specific
surface (Blaine) 3770 cm2/g. The same mill grinds an OPC clinker to cement of the same
finess by 4000 revolutions.
The potential composition of the SAC and the OPC, which was used as reference, are
given in Table 1.
Appropriate concrete technology 276
The dimensional stability of the prepared SAC was tested according to EN and ASTM
standard methods and was found similar to OPC17.
The setting time and the mechanical strength were measured according to EN
Standard methods18. For the determination of the heat of hydration the AFNOR method
was applied19.
The development of the pore structure was studied by Hg intrusion (Carlo Erba
Porosimeter 4000) in mortar specimens similar to those prepared for the mechanical
strength measurements.
Porosity. The total pore volume of the SAC mortar specimens was significantly lower
than that of OPC especially at the early hydration ages (Table 2). After 90 days of
hydration the total pore volume of OPC specimens became similar to this of SAC. The
pore size distributions of the SAC specimens (Fig. 2) showed that already after 2 days of
hydration the most of the pores had diameters lower than 1000 A, while the OPC
specimens had this size as mean pore diameter at the later ages (Fig. 3). After 28 and 90
days, hydration of the most OPC compounds resulted in the formation of a narrow
distribution of pore diameters, while the corresponding curves of SAC formed a shoulder
on the lower diameter side. It is considered that the latter could have a narrower form, if
the hydration of SAC was more complete. All the above observations indicate that the
studied SAC could exhibit a good behaviour in an aggressive environment.
CONCLUSIONS
REFERENCES
1. H.LAFUMA, Expansive Cements, Proc. 3rd Intern.Symp.Chem. Cem., London 1952, Cement
and Concrete Association, London 1954, p.581592.
2. L.MOHAN, Advances in Some Special and Newer Cements, in Progress in Cement and
Concrete, S.N Ghosh(Edr.), ABI, New Delhi 1991, Vol.I, Part I, p.253313.
3. T.NAKAMURA, G.SUDOH, S.AKAIWA, Mineralogical Composition of Expansive Cement
Clinker Rich in SiO2 and its Expansibility, Proc. 5th Intern.Symp. Cem.Chem., Tokyo 1968,
Vol.IV, p.351365.
4. W.KURDOWSKI, Expansive Cements, Proc. 7th Intern.Cong. Chem.Cem., Paris 1980, Vol.I,
p.V-2/111.
5. W.KURDOWSKI, C.M.GEORGE, F.P SORRENTINO, Special Cements, Proc. 8th Intern.Cong.
Chem.Cem., Rio de Janeiro 1986, Vol.I, p.292318.
6. G.SUDOH, T.OHTA, H.HARADA, High Strength Cement in the CaOAl2O3SiO2SO3
System and its Application, Proc. 7th Int. Cong.Cem.Chem., Paris 1980, Vol.III, p.V/152157.
7. P.K.MEHTA, Investigations on Energy-Saving Cements, World Cement Technology 1980,
Vol.11(4), p.166177.
8. DENG JUN-AN, GE WEN-MIN, SU MU-ZEHN, LI XIU-YING, Sulfoaluminate Cement
Series, Proc. 7th Int.Con.Cem.Chem., Paris 1980, Vol.IV, p.V/381386.
9. WEI JIN-ZHAO WANG YUE-E, Burning of C3SC4A3 Fly-ash Cement and its Characteristic
Features, Silicates Industriels, 1983/45, 9397.
Appropriate concrete technology 280
Keywords: High alumina cement (HAC), ground granulated blastfurnace slag (ggbs),
phase equilibria, strtlingite.
Dr Keith Quillin works in the Inorganic Materials Division of the Building Research
Establishment at Garston. His main research interests lie in the hydration chemistry of
cements.
Appropriate concrete technology 282
INTRODUCTION
Concretes made from high alumina cement (HAC) have been found to lose strength with
time, particularly in hot and humid environments, through a process known as
conversion. In the conversion process the initial crystalline products of hydration, the
metastable calcium aluminate hydrates* CAH10 and C2AH8, are replaced by the stable
calcium aluminate hydrate C3AH6 (hydrogarnet) and AH3[1]. The conversion reactions
occur extremely slowly at 5C, but become more rapid as the temperature is increased
and C3AH6 formation is almost immediate above 50C. C3AH6 has a higher density (2530
kg m3) than either CAH10 (1730 kg m3) or C2AH8 (1950 kg m3) and so conversion is
accompanied by a decrease in the solid volume and therefore an increase in porosity that
can lead to a loss in strength[1,2].
The Building Research Establishment has been investigating cements in which HAC is
blended with latently hydraulic materials such as ground granulated blastfurnace slag
(ggbs) or pozzolanic materials such as pulverised fuel ash (pfa) and metakaolin[39].
Most of the work carried out so far has concentrated on mixtures of HAC with ggbs
which BRE has patented and registered under the trade name BRECEM[3]. Concretes
made by using BRECEM do not exhibit the loss of compressive strength that occurs in
HAC concretes when kept at 38C under wet conditions over a prolonged period of time.
BRECEM concretes also perform well on storing in sulphate solutions and in aggressive
marine and acid water environments[8,9].
The absence of strength loss in these blended HAC cements is a consequence of the
modified chemistry on hydration caused by the high silica contents of the additions.
When blended HAC cements are hydrated at ambient temperatures the metastable
calcium aluminate hydrates form initially due to the hydration of the HAC component.
However, as the pozzolanic or latently hydraulic material hydrates the concentration of
silica in solution increases. A new hydrate, C2ASH8, known as strtlingite or gehlenite
hydrate, then forms causing the calcium aluminate hydrates to redissolve substantially.
As a result the proportion of the mix made up of the calcium aluminate hydrates will be
much less than in HAC and so the effects of the conversion reaction on the strength of the
concrete will be greatly reduced.
Majumdar and Singh [3] have studied the paste hydration of HAC with a number of
additions other than ggbs. Those studied were gasifier slag, metakaolin, silica fume,
Italian pozzolana, a high lime-containing pfa and a low lime-containing pfa. Bentsen et
al[10] and Bayoux et al[11] have studied phase formation in pastes of calcium aluminate
cements with silica fume. Collepardi et al[12] have studied the hydration of HAC with
silica fume and fly ash. Several of these materials have the potential to stabilise HAC
under hot and humid conditions. Gasifier slag, high lime pfa, metakaolin and silica fume
appeared[3] to be the most successful after 1 year based on the compressive strength.
C2ASH8 formed as a persistent phase (up to 1 year) in nearly all the mixes studied at both
20C and 40C. However, it was present[3] only
in trace quantities in a 60 HAC: 40 pfa (low lime-containing) mix at 40C after 180 days
and was absent after 1 year.
Phase equilibria in the CaO-Al2O3-SiO2-H2O system in relation to blended HAC cements 283
The compressive strength of concretes made using HAC/ggbs mixes at 60C has been
reported [13] to be lower than expected. These results are consistent with those of a
recent study carried out at BRE[14] in which the temperature, compressive strength and
phase composition of a 1m3 cube of BRECEM concrete cured under adiabatic conditions
were monitored. The temperature of the cube was found to rise to a maximum of 58C
and the compressive strengths of cores taken from the cube were low. XRD showed that
large amounts of C3AH6 were present. The results suggest that the HAC component
hydrated quickly and, as a consequence of the temperature rise on hydration, the
metastable hydrates quickly converted to C3AH6. C2ASH8 was not present initially,
suggesting that it is either unstable or that it forms slowly at elevated temperatures.
C2ASH8 was, however, found in large quantities at later ages when the temperature had
fallen [8].
The stability of C2ASH8 is clearly a key factor in the long term performance of
concretes made using blended HAC cements. If it is unstable it will eventually be
replaced by another hydrate or hydrates. The main aim of the work reported here was to
determine the stability of C2ASH8, the hydrates that form with it and the range of
compositions and temperatures over which it forms. This has been done primarily by
studying phase equilibria in the CaOAl2O3SiO2H2O system at temperatures between
5C and 60C. Such a study is valuable as the principal hydrates that form on hydrating
HAC both on its own and in mixtures with ggbs and pozzolanas also form within the
chemically simplified CaOAl2O3SiO2H2O system.
The solid phase compositions of some blended HAC cements have also been studied
using accelerated ageing techniques. These techniques allow the hydrates that would
normally be formed only after long periods of hydration to be determined in a relatively
short period.
EXPERIMENTAL
C3AH6 when silica is available) was determined from the d-spacing corresponding to the
420 line (using quartz as an internal standard) [13]. Aqueous solutions were analysed as
soon as possible after filtration for calcium and aluminate ions using atomic absorption
spectroscopy, and for silicate ions using a molybdenum blue photometric method. The
results of the solution analyses have been used in a modelling study of parts of the CaO
Al2O3SiO2H2O system [15].
RESULTS
consistent with the compositions determined at the lower temperatures. AH3 (gibbsite) is
formed in substantial quantities and is likely to be a stable phase.
CSH gel was present in some mixes. There was some evidence of the formation of
crystalline calcium silicate hydrates at 60C; either CSH(1) or a tobermorite-like
material. However, it was not possible to make a definite assignment from the XRD
peaks. The phase assemblage C2ASH8/C3AS0.3H5.4/AH3 appears to be stable at 60C. The
assemblage C3AS0.3H5.4/C3AS0.8H4.4/CSH may also be stable at 60C. It was detected
after 4 months but after 16 months C3AS0.8H4.4 had been replaced by C2ASH8. Both these
assemblages were also found to be stable at 38C.
C2ASH8 was not formed in the single mix run at 90C. A siliceous hydrogarnet with a
composition of approximately C3AS0.3H5.4 together with a zeolite phase with an XRD
pattern similar to those of thomsonite (NaCa2Al5SiO20.6H2O) and scolecite
(CaAl2Si3O10.3H2O) were identified by XRD. A calcium silicate hydrate phase with an
XRD pattern similar to that of hillebrandite (C2SH) may also have been present.
Table 1. Hydrates forming in HAC/ggbs mixes.
T/C HAC:ggbs Age/d AH3 CAH10 C2ASH8 hydrogarnet CSH Others
5 60:40 232 C4AcH11
20 60:40 232
20 50:50 230
50 60:40 232
50 40:60 234
90 60:40 276
50 50:50 223 ?
DISCUSSION
The development of C2ASH8 in pure oxide mixes at 60C has been found to be very
slow. Its formation in BRECEM concretes has also been slow at elevated temperatures.
At its present state of development, therefore, BRECEM may not be suitable for use in all
applications for mass concrete, although this is an area in which HAC is also unsuitable.
BRECEM could be used in mass concrete applications where slower strength
development is allowed, but may not be suitable if rapid strength development was a
priority[11]. However, promising results have been obtained for precast BRECEM
concretes that were sprayed after demoulding to control the temperature rise on
hydration. Strength development was good and C2ASH8 was detected in substantial
quantities at early ages[8]. It may be possible to accelerate its formation at this
temperature by using a source of silica that is more reactive than ggbs. If so it may be
possible to improve the compressive strength development of mass BRECEM concretes.
Damidot and Glasser[16] have carried out modelling studies on the CaOAl2O3SiO2
H2O system at 25C and their results were consistent with those reported here.
Damidot[17] has also extended this study to show that C2ASH8 is more stable in the
presence of sulphate and carbonate ions than is C3AH6. This is consistent with the
observed increased sulphate resistance observed for BRECEM concretes.
Figure 1 shows the stable and metastable phase assemblages found in the study of the
CaOAl2O3SiO2H2O system at temperatures up to 38C, projected on to the CaO
Al2O3SiO2 surface. Phases that were found to be compatible in the experimental work
on the CaOAl2O3SiO2H2O system are linked using solid lines. The figure also shows
the idealised compositions of HAC, together with those of potential additions. The
compositions of blended cements will lie between those of HAC and the particular
addition. Blended HAC cements cannot be accurately modelled by the CaOAl2O3SiO2
system as they contain other oxides in significant quantities. Other hydrates such as
calcium carboaluminate hydrates and hydrotalcite may also form. Figure 1 is therefore a
simplification. Despite this, Figure 1 should allow the main equilibrium products formed
on hydrating an HAC blend to be predicted. These will be the 3 hydrates forming the
triangle in which the overall mix composition falls. The relative amounts will be
inversely proportional to the distance between the hydrate and mix compositions. As
HAC hydrates more rapidly than the additions the hydrated phases present will initially
be the calcium aluminate hydrates and AH3. C2ASH8 will form as the addition hydrates.
The extent to which the addition hydrates will be dependent on the availability of
water and the reactivity of the silica-containing addition. For an addition to be viable it
must increase the availability of silica during the hydration reaction and must result in the
formation of substantial amounts of C2ASH8 in order to limit the effects of the conversion
reaction as discussed above. The amount of C2ASH8 forming will be greater in blends
with overall compositions that are closest to that of C2ASH8 in Figure 1. Therefore blends
of HAC with ggbs, gasifier slag or pfa (USA) should all contain large amounts of
C2ASH8, depending on the relative proportions of the two constituents. HAC blended
with metakaolin and pfa (UK) would be expected to contain less C2ASH8 as the
compositions of mixes lie further away from the C2ASH8 composition although it does
not automatically follow that the compressive strengths of concretes made using these
blends would be low. Collepardi et al[12] have shown that the addition of low lime-
containing pfa to HAC is not advantageous in reducing the effects of the conversion
reaction. However, studies on concrete cubes made using HAC and metakaolin blends
have shown promising results[3].
The phases formed[3] on hydrating 50:50 blends of HAC with the additions listed in
the introduction were CAH10 or a siliceous hydrogarnet, C2ASH8 and, in some cases,
AH3. The presence of CAH10 and hydrogarnet in the same mix would indicate that the
conversion process was still occurring leading to the formation of the stable assemblage
C2ASH8/hydrogarnet/AH3. However, from Figure 1 it is apparent that the compositions of
50:50 blends of HAC with silica fume, Italian pozzolana, metakaolin and both types of
pfa may fall outside the composition triangle defined by the stable phase assemblage
C2ASH8/hydrogarnet/AH3 (although variations in the compositions of the constituents
and the presence of other oxides such as Fe2O3 and MgO will affect the exact position of
the compositions). They lie in a triangle defined by C2ASH8, AH3 and a zeolite-type
phase such as gismondine (CAS2H4). The calcium aluminate hydrates that form in these
blends[3] may therefore eventually disappear, although this process may be extremely
slow at ambient temperatures. Studies of the hydration of HAC/metakaolin blends
described above showed that hydrogarnet (or CAH10 at lower temperatures) was present
after several months, together with C2ASH8 and AH3. Zeolites have been found to
form[18,19] in blended Portland cements at temperatures below 90C.
Phase equilibria in the CaO-Al2O3-SiO2-H2O system in relation to blended HAC cements 289
The compositions of the 50 HAC: 50 ggbs and 40 HAC: 60 ggbs BRECEM mixes are
within the triangle defined by the phases C2ASH8, CSH and a hydrogarnet with the
approximate composition C3AS0.9H4.2. However, because of the relatively slow hydration
of the ggbs component the phases detected in BRECEM concretes after 5 years are
C2ASH8, AH3 and a hydrogarnet with an approximate composition of C3AS0.3H5.4. As
C2ASH8 is the predominant phase in BRECEM mixes any changes in the minor phases, if
they occur at all, would probably not be significant. Such changes would not be expected
to occur in the 60 HAC: 40 ggbs mix for compositional reasons. The accelerated ageing
study of HAC/ggbs blends described in section 3 showed that CSH formed in 40
HAC:60 ggbs blends at 50C.
CONCLUSIONS
C2ASH8, the principal phase formed on hydrating blended HAC cements, is a persistent
phase in the CaOAl2O3SiO2H2O system at temperatures up to at least 60C. This
result is consistent with durability studies on BRECEM concretes in which the
compressive strength is maintained for at least 5 years under wet conditions at 38C.
Other blended HAC cements have also given promising results in durability studies after
1 year but their long term behaviour needs to be confirmed. The increased stability of
BRECEM cements on exposure to aggressive environments can also be assigned to the
stability of C2ASH8.
The low compressive strength of BRECEM concretes cured under adiabatic conditions
is probably associated with the slow formation of C2ASH8 at temperatures of about 60C.
C2ASH8 is, however, likely to be stable at this temperature. It may be possible to
accelerate its formation at 60C if a suitably reactive source of silica can be identified.
C2ASH8 is probably unstable at 90C.
Comparisons of the compositions of blended HAC cements with phase compatibility
diagrams for the CaOAl2O3SiO2H2O system show that HAC blended with ggbs,
gasifier slag or high lime-containing pfa should produce the most C2ASH8 on hydration.
Hydrogarnet (or one of the metastable calcium aluminate hydrates) and AH3 would also
be present. HAC blended with metakaolin or low lime-containing pfa would be expected
to contain less C2ASH8 on compositional grounds, and it is possible that a zeolite-like
phase may eventually be formed, especially at high addition levels. However, accelerated
ageing studies on HAC/metakaolin blends have not shown the presence of any new
phases.
BRECEM may eventually be used as a specialist repair material for existing concrete
structures or in structural concrete elements such as precast piles, prestressed beams, tank
sections and wall/floor slabs. However, its viability for use will depend on the costs of
basic materials[8] and its acceptance via standards and codes of practice.
Appropriate concrete technology 290
REFERENCES
1. H.G.Midgley& A.Midgley, The conversion of high alumina cement, Mag. Concr. Res., 27, 59
(1975).
2. R.J.Collins& W Gutt, Research on long term properties of HAC concrete, Mag. Conc. Res., 40,
195 (1988).
3. A.J.Majumdar & B.Singh, Properties of some blended high-alumina cements. Cem. Conc. Res.,
22, 1101 (1992).
4. A.J.Majumdar & B.Singh, UK patent GB 2211 182 B (1991).
5. A.J.Majumdar, B.Singh & R.N.Edmonds, Hydration of mixtures of Cement Fondu aluminous
cement and granulated blastfurnace slag, Cem. Conc. Res., 20, 197 (1990).
6. A.J.Majumdar, B.Singh & R.N.Edmonds, Hydration of calcium aluminates in the presence of
granulated blast furnace slag. Proc. Intern. Symp. on Calcium Aluminate Cements, London,
1990, pp 259271.
7. A.J.Majumdar & B.Singh, The hydration of CA2 and its mixtures containing slag. Cem. Conc.
Res., 22, 1019 (1992).
8. G.J.Osborne & B Singh, The durability of concretes made with BRECEM cement comprised of
blends of HAC and ggbs. Proc. 5th CANMET/ACI Int. Conf. on fly ash, silica fume, slag and
natural pozzolans in concrete, Milwaukee. Vol. 2, SP153 (1995).
9. G.J.Osborne, Performance of HAC/blastfurnace slag concretes in aggressive environments Proc.
CONSEC 95 International Conference on Concrete under Severe ConditionsEnvironment
and Loading. Vol. 2, p1302, Sapporo (1995).
10. S.Bentsen, A.Seltveit & B.Sandberg, Effect of microsilica on conversion of HAC, in Calcium
Aluminate Cements, ed. R.J.Mangabhai, p294 (1990).
11. S.Marcdargent, M testud, J.P.Bayoux & A.Mathieu, 9th ICCC, Vol. 4, p651 (1992).
12. M.Collepardi, S.Monosi & P.Piccioli, The influence of pozzolanic materials on the mechanical
stability of aluminous cement, Cem. Conc. Res., 25, 961 (1995).
13. S.Hirose & Y.Yamazaki, Hydration of high alumina cement mixed with blastfurnace slag.
Gypsum and Lime, 233, 216 (1991).
14. G.J.Osborne, BRECEM: a rapid hardening cement based on high alumina cement, Proc. Instn.
Civ. Engrs. Structs & Bldgs, 104, 93 (1994).
15. K.C.Quillin & A.J.Majumdar. Phase equilibria in the CaOAl2O3SiO2H2O system at 5, 20
and 38C. Adv. Cem. Res., 6(22), 47 (1994).
16. D.Damidot & F.P.Glasser, Investigation of the CaOAl2O3SiO2H2O system at 25C by
thermodynamic calculations. Cem. Conc. Res., 25, 22 (1995).
17. D.Damidot, Improvement of calcium aluminate cement resistance to sulphate and carbonate
attacks by silica addition: Thermodynamic approach. Proc. Int. Coll. on Corosion of Cement
Paste, Krakow, 1994. Ed. W. Kurdowski (1994).
18. K.Quillin, S.L.Duerden & A.J.Majumdar. Accelerated ageing of cements for use in radioactive
waste disposal, DoE/HMIP/RR/93.023 (1993).
19. P.S.De Silva & F.P.Glasser, Phase relations in the system CaOAl2O3SiO2H2O relevant to
metakaolin-Ca(OH)2 hydration. Cem. Conc. Res, 23, 627 (1993).
HIGH PERFORMANCE CONCRETE
CONTAINING MODIFIED RICE HUSK
ASH
B Chatveera
Thammasat University
P Nimityongskul
Asian Institute of Technology
Thailand
Keywords: High performance concrete, mechanical properties, modified rice husk ash,
superplasticizer.
Dr Burachat Chatveera is a lecturer in the Department of Civil Engineering at
Thammasat University, Rangsit Campus, Pathum Thani 12121, Thailand. He received his
D.Eng. from Asian Institute of Technology, Thailand.
Dr Pichai Nimityongskul is an associate professor in the School of Civil Engineering at
Asian Institute of Technology, Pathum Thani 12120, Thailand. He received his D.Eng.
from Asian Institute of Technology, Thailand.
INTRODUCTION
In the developing countries, considerable efforts have been directed towards the
utilization of indigenous and waste materials. Not surprisingly there has been interest in
the use of rice husk as a building material and limited success has been achieved with the
Appropriate concrete technology 292
use of the husk in particle board and cement-husk building block [1]. A more promising
area of interest however has been the use of the rice husk ash (RHA) either as pozzolan
[2] or in the manufacture of cementitious materials [3].
The present study is aimed at investigating the mechanical properties of high
performance concrete in which the matrix contains modified rice husk ash. The main
parameters are the type and percentage of MRHA. The percentages of MRHA used as
cement replacement vary from 10 to 60% and the percentage of MRHA used as cement
addition is 10%. The dosage of superplasticizer is kept constant at 2.4% of cementitious
material by weight. The aggregate-cement and sand-aggregate ratios by weight are kept
constant and equal to 3.25 and 0.40, respectively. The slump is maintained at 20 2 cm.
The study is undertaken with the primary aims of determining the followings:-
(i) The effects of different types and percentage replacements of MRHA on the
compressive strength at different ages and 28-day modulus of rupture of high
performance concrete and slump retention over a duration of 90 minutes.
(ii) Optimum mix of high performance concrete containing MRHA based on the
mechanical properties obtained.
EXPERIMENTAL DETAILS
MRHA Type 1
RHA, water and a dispersing agent were mixed together by the mixer. The RHA and a
dispersing agent were initially fed into the mixer followed by a portion of the required
water. The mixing duration was 1 minute. The RHA: water ratio was 1:1. The dosage of a
dispersing agent was 0.2% of RHA-water mixture by weight.
High performance concrete containing modified rice husk ash 293
MRHA Type 2
RHA, water, a dispersing agent and MIGHTY 150 were mixed together by the same
mixer. The RHA and a dispersing agent were initially fed into the mixer. The required
water and MIGHTY 150 were subsequently added into the mixer. The mixing duration
was 2 minutes. The RHA: water ratio was 1:1. The dosages of a dispersing agent and
MIGHTY 150 were 0.2% of RHA-water-MIGHTY 150 mixture by weight and 3% of
RHA by weight, respectively.
MRHA Type 3
RHA, water, a dispersing agent and MIGHTY 150 were mixed together by the same
mixer. The mixing procedure was similar to that of MRHA type 2. The RHA: water ratio
was 1:1. The dosages of a dispersing agent and MIGHTY 150 were 0.2% of RHA-water-
MIGHTY 150 mixture by weight and 3% of RHA-water mixture by weight, respectively.
Properties of Pozzolans
It can be seen that the three different pozzolans designated as RHA, MM-FA and HK-FA
composed of different amounts of oxides. RHA before modification contained very high
amount of SiO2, and little amount of Fe2O3, CaO, MgO and SO3. Because of the high
content of SiO2, which plays an important role on pozzolanic properties; and its high
water requirement, it was evident that RHA met the requirements of pozzolan class N as
shown in Table 1. When comparing with MM-FA and HK-FA, it was found that the SO3
content of MM-FA were much higher. MM-FA had higher SiO2 and Al2O3 contents but
the Fe2O3 and CaO contents were higher. This may be due to the different kinds of the
raw materials and the combustion processes by which pozzolan is formed. It was also
observed that RHA possessed lower specific gravity than fly ash. The reason is that RHA
has very high content of silica.
Appropriate concrete technology 294
The fineness of RHA passing no. 325 sieve and pozzolanic activity index had larger
values than fly ash. This means that the particle size of RHA is smaller than fly ash.
Because RHA particles are not spherical, RHA cement requires more water to produce
the same consistency as that of fly ash cement.
i) Strength
The results on compressive strengths which were summarized and shown in Figure 1
indicated that among the three types of MRHA investigated, MRHA type 3 was the most
effective pozzolan and the optimum percentage replacement for all types of MRHA was
30%. It can be observed that the compressive strengths at ages of 3 and 7 days of MRHA
concrete were lower than those of the control concrete. This is due to the fact that MRHA
reacted slowly with calcium hydroxide liberated from the reaction between the C3S and
C2S compounds and water to form calcium silicate hydrates. Thus, the early compressive
strength of MRHA concrete were found to be lower than that of the control concrete. This
phenomenon is similar to fly ash concrete. However, at ages of 28 and 56 days, MRHA
concrete having MRHA replacement up to 30% showed compressive strength higher than
the control concrete. At MRHA replacement 40%, the compressive strength at all ages
was found to be less the same as those of the control concrete. However, if the percentage
replacement of MRHA exceeded 50%, the compressive strength at all ages was
considerably reduced. It was also indicated that the highest compressive strength was
obtained when using 10% MRHA as cement addition. Regarding the
Table 1 Chemical and physical properties of
pozzolans as compared with those of ASTM
requirements
CHEMICAL PROPERTIES ASTM C61891 TYPE OF
CLASS POZZOLAN
N F C RHA MM- HK-
FA FA
SIO2+AL2O3+FE2O3 (MIN, %) 70 70 50 92.28 59.97 80.91
SO3 (MAX, %) 4 5 5 0.18 1.64 0.34
MGO (MAX, %) 5 5 5 0.18 3.52 1.01
NA2O (MAX, %) 1.5 1.5 1.5 0.03 1.12 0.20
LOI (MAX, %) 10 12 6 3.67 0.27 2.59
PHYSICAL PROPERTIES
FINENESS; AMOUNT RETAINED WHEN 34 34 34 0.36 17.35 6.08
High performance concrete containing modified rice husk ash 295
testing results on the modulus of rupture at 28 days shown in Figure 2, the trend was
observed to be more or less the same as that of compressive strength.
ii) Workability
Regarding the properties of fresh concrete containing MRHA, the relationships between
slump and elapsed time of fresh concrete containing different percentages of MRHA
types 1, 2, and 3 are shown in Figure 3. The results indicated that for MRHA types 1 and
2, the slumps of concretes containing 50 and 60% MRHA replacements were sharply
High performance concrete containing modified rice husk ash 297
reduced within a period of 15 minutes after mixing whereas for fresh concretes
containing 10 to 30% MRHA replacements, the workabilities of concretes were
maintained over a period of 90 minutes. For MRHA type 3, the slump retention of the
fresh concrete was better than the other two types of MRHA when 50% of MRHA
replacement was used. For all types of MRHA, the percentage replacements below 40%
resulted in significantly improved the slump retention performance of fresh concrete. It
can be noted that for MRHA replacements of 10 and 20%, the slump losses were found to
be only 15 to 20% over the duration of 90 minutes. It is evident that the replacements of
cement by MRHA (10 and 20%) significantly improved the slump retention
characteristics of fresh concretes as compared to the control concrete. This may be due to
the effect of dispersing agent which was used in the modification of rice husk ash. An
important point which should be emphasized is that high amount of MRHA replacement
resulted in a poor slump retention characteristic of fresh concrete.
CONCLUSIONS
High performance concrete can be developed by replacing cement with modified rice
husk ash which was obtained by premixing rice husk ash to water, a dispersing agent and
superplasticizer in slurry form. The highest compressive strength was obtained when 30%
of MRHA type 3 was used to replace ordinary Portland cement. The strength and slump
retention of concrete were considerably reduced when the percentage of MRHA
replacement exceeded 50% by weight.
REFERENCES
1. URS RESEARCH CO. FOR THE U.S. DEPARTMENT OF AGRICULTURE. Rice Hull
UtilizationFinal Report. Contract No. 12141009571(74), September 1970.
2. COLUMNA, V B. The Effect of Rice Hull Ash in Cement and Concrete Mixes. M.Eng. Thesis
No. 678, Asian Institute of Technology, Bangkok, Thailand, 1974.
3. MEHTA, P K. Rice Hull Ash Cement High-Quality, Acid-Resisting. ACI Journal, Vol. 72,
No. 5, 1975, pp 235236.
4. ISLAM, M S. Grinding Methods and Its Effect on Reactivity of Rice Husk Ash. M.Eng. Thesis
No. ST-817, Asian Institute of Technology, Bangkok, Thailand, 1981.
5. WE, A B. Production of RHA and Its Applications in Mortar and Concrete. M. Eng. Thesis No.
ST-8120, Asian Institute of Technology, Bangkok, Thailand, 1981.
6. KOKUBU, M. Principal Paper Fly Ash and Fly Ash Cement. The Fifth International Symposium
on the Chemistry of Cement, 1968, pp 75113.
7. DAVIS, R E, CARLSON, R W, KELLY, J W AND DAVIS, H E. Properties of Cement and
Concrete Containing Fly Ash. ACI Proceedings, Vol. 33, 1937, pp 577612.
SLAG ALKALINE POLYMER CEMENT
CONCRETES
P V Krivenko
V A Raksha
L V Raksha
SRIBM
Ukraine
Keywords: Slag Alkaline Cement and Concrete, Latex, Emulsion Petrolatum, Porosity,
Logarithmic Decrement of Damping, Medium Radius of Pores.
Professor Pavel V Krivenko is Director of Scientific Research Institute of Binders and
Materials, Kiev University, Ukraine. He specialises in the field of the special binders and
concretes. Professor Krivenko has many publications the world over.
Dr Vladimir A Raksha is an Assistant Professor Kiev University. He specialises in
concretes with increased damping properties.
Miss Ludmila V Raksha is a PhD. Researcher of the Kiev University. Her main research
interest include the special concretes with increased damping properties.
INTRODUCTION
Slag alkaline cement (SAC) and concretes on their base possess high physical-
mechanical and some special properties [1,2]. Further improvement of properties of these
materials may be achieved by their modification with additives of organic and mineral
origin [3,4,5].However, low strength of the Portland cement matrix restricts the use of
Slag alkaline polymer cement concretes 299
these materials. This paper reviews the principle of latex and petrolatum modification of
slag alkaline cements and concretes and discusses their typical and special properties.
EXPERIMENTAL DETAILS
Materials and Methods. Slag alkaline cements and concretes on their base have been
chosen as basic materials for investigation. Slag alkaline cemens (SAC) are composed of
ground blast furnace slag with Mb=1.2. Quartz sand with a gradation factor 1.7 and
crushed granite with particle size 510 mm were used as aggregates.
The additives of petrolatum emulsion and latex in a form of stabilized dispersion
(50%solids) were used for modification. The addition of latex into the slag alkaline
cements was done in two ways: Itogether with alkaline solution; IIwhile grinding the
slag. The additive of petrolatum emulsion was added when grinding the slag. The testing
procedure for determining. The main physico-mechanical properties didnt differ the
from conventional procedurer appied to the traditional cementitious materials.
Experimental Part. Discussion of Results. It is known, that the modification of portland
cement materials with the latex dispersions is hindered by the unstabiliby in the
dispersion medium of hardened binder [5,6]. These factors, certainly, take place under the
modification of slag alkaline cements. Hence it is, necessarily to take into account the
high alkality of the medium (pH=1213) of these cements. The modification of SAC
can be attained by the use of stabilized latex, mixed for 23 minutes in a high-speed
mechanical mixer just before application, as shown by our research.
The influence of quantity and method of adding latex, on the alkaline solution to slag
ratio, setting time and strength characteristics of these cements were determined. The
results are presented in Table 1.
The addition into cements composition of the latex in quantity to 2.5% (dry state)
does not depend on the methods of application and has no influence on the strength of
these cements. A further increase in content of latex (to 10%) results in a strength decline
as compared with that of control. The application of latex dispersions allows to extend
slightly the setting time. This is, evidently, a result of the formation of polymer films on
slag grains, which prevent its hydration. It is demonstrated also the decrease of alkaline
solution to slag ratio from 0.28 to 0.24 accounting to a plastisizing effect of latex and
explained by the presence of surface active substances (SAS) the within compositions of
their stabilizers. The kinetic change of strength of modified slag alkaline cements
hardened during different conditions for one year was revealed. The studys results are
given in Figures 1 and 2.
As is follows from the obtained data the character of strength change of latex-modified
cements is identical to the change of strength of the control. At equal quantities of latex
and similar curing conditions the strength characteristics of the latex-modified cements
treated by the first method are higher than those of cements treated by second method.
The flexural strength of the latex-modified slag alkaline cements is over 1015% higher
than that of the control. This tendency iprevails for one year of hardening.
Appropriate concrete technology 300
The shrinkage deformation of the latex-modified slag alkaline mortars is higher than
that of control composition (1.21.3 mm/m and 0.9 mm/m, respectively) (Figure 3,
curves 1, 2 and 3). The introduction into concrete of coarse aggregate with fractions 510
mm or reinforcing with coarse basalt fibres (d=100200 mcm) is considerably decreased
the shrinkage deformation (Figure 3, curves 4, 5). It this case, the most considerable
shrinkage deformation are observed during 3050 days, which stabilized at 60 days.
The damping properties of the slag alkaline concretes depend first of all on the similar
characteristics of the slag alkaline cements. Table 2 is presented the investigation results
of LDD of the modified slag alkaline concretes. A decline in the absolute value of LDD
of the slag alkaline concrete as compared with that of the slag alkaline cements is
attributed to a saturation of the bulk of the material with quartz sand characteristic of low
LDD values (LDD of quartz sand is 0.005). The damping properties (evaluated by the
LDD) of the latex-modified slag alkaline concretes are found to be at a level of the
similar characteristics for polymer concretes.
Appropriate concrete technology 304
Analyzing the obtained data, presented in Table 3, suggests to conclude that the crack
resristance of these concretes is greater, but the coefficient of softening is lower, when
increasing the latex content to 2.5%. However, the water resistance of latex-modified slag
alkaline concretes is considerably higher, it is allows to use them in water hydraulic
structures.
The testing procedure for repeated shock strength was done in the latex-modified slag
alkaline mortars (latex content found as 1% (in dry state) of slag mass) on.impact.testing
machine with a falling load of 3 kg. The specimen were prerared in the form of cylinders
of height and diameter of 25 mm. Control materials endured 15 impacts while the
modified120 impacts. This accounts for the increase in impact resistance of the
samples by 8 times. Such a rise in impact resistance is explained by the peculiar structure
of the modified concrete, thus stipulated by the presence of an elastic polymer layer in the
contact zone of the binder and aggregate. The presence of such an elastic layer facilitates
for the reduction in the internal tension of the concrete and increase of its crack
resistance.
Slag alkaline polymer cement concretes 305
CONCLUSIONS
1. The slag alkaline cements and concretes can be modified with water solution polymer
additives: latex and emulsion petrolatum.
2. The slag alkaline polymer-cement concretes possess higher flexural strength, repeated
shock strength, high damping properties and crack resistance.
3. The slag alkaline concretes by their physical-mechanical and special properties can be
used in special fields of constraction (seismic, etc.) and non-civil engineering
(vibration-damping foundations, contractions for high-way engineering, containers,
foundations of precision equipment).
REFERENCES
1. KRIVENKO, PV. Special Slag Alkaline Cements. Budivelnik Publish, Kiev, 1992, pp 197.
2. KRIVENKO, PV, KONSTANTINOVSKY, B Ya AND RAKSH, VA. Slag Alkaline Binders
and Concretes for Precision Equipment. Journal Cement, Nos 1112, pp 3439.
Appropriate concrete technology 306
INTRODUCTION
The most widely used building material in industrialized countries is based on Portland
cement as a binder. One type of Portland cement is produced in Tanzania in the more
developed coastal zone, but production is inadequate to facilitate extensive rural housing
development of higher quality. Moreover, cement is too expensive for most villagers,
since a bag of cement would consume more than half the monthly income for the average
farmer in a smaller village. Tanzania has a poorly developed infrastructure. Transport of
the cement to the more remote areas in the country is expensive, time consuming and not
reliable, because a considerable part of the dirt roads is inaccessible in the rainy periods.
As an example, the price of the cement doubles when transported from Dar es Salaam
over 750 km to Mwanza, the third largest town in the country.
Attention should therefore focus on the development of appropriate technology
concepts on village level, thereby making use of local raw materials and other indigenous
resources. It is known for quite some time that a highly reactive pozzolanic material can
be obtained from controlled burning of rice husks (a FAO review mentions two German
patents dealing with the use of RHA in concrete that were registerd in 1924). Since those
days, a considerable amount of effort has been invested in this field and applications have
been realized in South America as well as in Asia. Reference should be given here to the
work of Mehta [1], based on fundamental insight into the chemistry involved in this
materials technology. The use of RHA-cement has been promoted in the South-East
Asian region by the ESCAP Regional Center for Technology Transfer through the
organization of a series of workshops in 1978 and later years. A full session was devoted
to RHA during the CANMET/ACI Conference in Milwaukee this year, showing the
world-wide interest in this field [2].
Since rice is one of the major crops in Tanzania, the present study concentrated on
production and application of rice husk ash in remote villages in the Mwanza region.
Table 1 presents 1992/93 data on rice production in the relevant districts. This would
provide for about 50 thousand tonnes per year of RHA. Maintaining cultural and
traditional inherit age as to the building technology, a major application of a higher
quality binding agent would be the production of a more durable and denser floor
material. Another application would be stabilizing the mud used
Table 1. Rice production in districts around
Mwanza
Mwanza Region Rice Production [thousand tonnes]
Sengerema District 250
Kwimba District 200
Ukerewe District 230
Mwanza District 170
Geita District 100
Mangu District 50
Rha-cement as a replacement for portland cement in rural tanzanian villages 309
in the wattle and daub type of housing, leading to enhanced durability. A popular
application is for plastering walls, with the idea of reducing erosion. But the present
experiences by a local Non Government Organization (NGO), the Mwanza Rural
Housing Program (MRHP), reveal some inverse proportionality between quality of
binder and durability. This is obvious, because the traditional wall construction (mud
blocks or wattle and daub) lead to relatively large deformations which cannot be
accommodated by the stiffer plaster layers, so they debond as a consequence and fall off.
Fig. 1 shows a site of field testing by MRHP near Missungwi, close to Mwanza. Only in a
more advanced situation, the (local) production could be perceived of building blocks
made of masonry-quality of mortars, such as the ones developed in this research.
The objectives of this study were
to analyse the Tanzanian husks for their usefulness in producing RHA;
to investigate the optimum production conditions for RHA;
to determine the mechanical properties of RHA-based mortars;
to design the field production in a Tanzanian rural area.
The rice husks were received from the Arusha region in Northern Tanzania (ie the
Ngarenarch rice mills). The raw material ranges in particle sizes from less than 0.038 mm
Appropriate concrete technology 310
to 2 mm. The chemical analysis revealed that the material contained 21.2% of silica
(further, 26.2% cellulose, 21.2% hemi-cellulose, 11.6% lignin and
19.8% pigments, pectine and protein). A combined study using back scattered electron
(BEI) and X-ray images of the husk showed the silica to be distributed mostly under the
husks outer surface, giving it an effective hardness between 5.5 and 6.5 on Mohrs scale
and leading to a considerable abrasion of the grinding equipment (Fig. 2). This confirms
the general concept of a soluble form of silica transported through the plant, concentrated
at the outside surface of straw and husk through evaporation, whereupon it polymerizes
into an opaline cellulose-silica membrane [3]. Details of the analysis are given in
references [4,5].
The general composition of RHA produced from these husks was determined as an
average of a series of analyses, employing an inductively coupled plasma 200 emission
spectrometer (ICP-200). Table 2 shows the chemical analysis in comparison with similar
data obtained on a Vietnamese rice husk type [4,5]. Based on a DTA analysis, a series of
ashes was produced in a laboratory oven under
different temperature-time regimes, whereupon the carbon content (the loss on ignition,
LOI) was determined. Some results are presented in Table 3; note that ash percentages
are by weight of the husk. Another sample of husks was subjected during 10 hours to
600C and a last one for 15 hours to 900C, the latter ash being as white as lime. All
ashes were ground for 160 minutes in an agate mortar. Thereupon, the particle size
distribution (psd) was determined by Malvern 2600 equipment. Details are provided in
references [4,5]. Differences between curves were not dramatic, with 50% of the particles
under 5 to 10 m. Examples of psds of ashes representing the 350C67 h and
600C10 h treatments are given in Fig. 3. The fineness of the ashes (ie. BET) was
additionally determined by nitrogen adsorption (with Autosorb-6B gas Sorption
Analyser). The data averaged over two analyses, which involve preheating at 150 and
360C, respectively, are presented in Table 4.
An ash (from Tanzanian rice husks) produced in a kiln (also used in the present study)
under conditions supposedly representative for production in the field (temperatures
varying around 700C) and intended to be used for the production of high strength
Appropriate concrete technology 312
concrete in Vietnam contained 23% of unburnt carbon. The ash was grinded in a 2 litres
laboratory ball mill to various degrees of fineness. BET data (obtained by nitrogen
adsorption) are presented for comparison in Table 5. The ash particles have a large
internal porosity before grinding, as revealed by secondary electron images of the ash.
The time of grinding is between brackets. Initially, the surface area increases with
grinding time. But upon collapse of the porous structure, the surface area drops. The
addition of a naphtalene type of superplasticizer during grinding (indicated by + in Table
5) proved to be very efficient. The best ash in Tables 3 and 4 (500C, 20 hours burning
time) is probably
in the same state of disintegration of internal porosity. This state is to be prefered because
of a reduced water demand.
The ashes were also investigated by SEM, TEM and X-ray diffraction. Up to 500C,
the secondary electron images of the ash revealed particles as spherical or globular in
shape and with a porous structure. Partially crystallized ash was found at 600C, whereas
at 900C this feature was dominant. This is confirmed by X-ray diffraction patterns. The
SEM pictures revealed the globular structure to increase in size with rising combustion
temperature from 510 m at the lowest temperature to 1050 m. Individual particle and
pore sizes are up to 1 m. Agglomerates seem to become more compacted at higher
combustion temperatures. Moreover, at 600C fine porous crystalline grains, smaller than
Rha-cement as a replacement for portland cement in rural tanzanian villages 313
1 m, are displayed, possibly revealing the transformation between the amorphous and
the crystalline state. This process was completed at 900C, where a crystalline structure
was observed with crystallites in the range of 1 m or larger. Some representative
samples of SEI are presented in Fig. 4.
Stringent regulations in The Netherlands as to air pollution and safety governed the
design of a kiln, intended for the production of ash under conditions representative
Table 5. BET specific surface area of RHA;
between brackets is the duration of grinding [h];
only the last ash is grinded in combination with a
super-plasticizer. Note that the husks are of
Vietnamese origin.
RHA-quality SBET SBET
in m2/g
RHA(0) 123 2
RHA(14) 137 11
RHA(18) 151 11
+
RHA(18) 58 6
and grinding; i.e. 50% of the particles was smaller than 20 to 10m going from the ashes
burnt under 350 C to those under 600 and 900C. The specific surface area varied
between 0.24 and 0.31 m2/g. The largest value (0.31 m2/g) was found for the 500C, 2
hours grinding treatment, the smallest for the 900C, 2 hours grinding treatment (0.22
m2/g).
The moderate differences in ash characteristics made it possible to select a relatively
high temperature regime for the kiln burning process. The temperature at the bottom of
the combustion chamber was controlled at a level of about 600C. Average temperatures
in the combustion chamber were probably around 700C. Unburnt parts were removed
from the ash by sieving. The ash was finally ground for 130 minutes, considered to be a
reasonable effort under rural conditions, based on the type of applications.
achieve a certain level of workability, which was measured afterwards by the slump and
flow test. The water to binder ratio was derived from the compositional data. As a result
of this practical set up (adapting rural conditions), the slump and flow rate as well as the
water to binder ratio fluctuated among similar mixes of about 6 litres. Standard prisms
4040160 mm) were made following laboratory procedures; i.e
Rha-cement as a replacement for portland cement in rural tanzanian villages 317
moulds were filled in two separate layers, followed by 10 seconds compaction on the
vibration table. The second layer exceeded the capacity of the mould. After compaction
the excess of material was removed. Specimens were stored before testing in a climatized
room (20C, 99% RH). Demoulding was carried out very carefully after 48 hours.
Initial and final setting time were determined using the Vicat apparatus. Typical data
of the two main RHA-blended systems are presented in Fig. 6 (open bar: final setting
time; dotted bar: initial setting time). It reveals the well-known phenomenon of the RHA
pozzolana delaying both setting times, in fact well above the level indicated in ASTM C
595 (for PC blended by a pozzolana). Only PC blended for a smaller part with RHA
would fulfil this standard. Flow values (ASTM C 109) and slump of the various mixes
slightly varied, i.e. slump between 2 and 5 mm. However, mixes had a considerable
different water to binder ratio of 0.5 for the PC mortar and between 0.8 and 1.0 for the
blended mixes.
compressive strength was recorded on pieces of the broken beams introducing the load
via 4040 mm platens (ASTM C 349). A total of 173 prisms was used for this field
investigation. RHA content in the binder varied in all mixes between 0.5 and 0.7. This
was complemented by either PC or by lime. In the PC-RHA-lime mixes the PC content
was 0.1. Clay contents of 0.1 and 0.3 where applied in RHA-lime mixes. For further
details, see reference [4]. Fig. 7 shows the development of bending strength within a 1
month period for the three material systems. The development of compressive strength
over a period of 1 year is shown for the best performing mixes in Table 6.
Table 6. Development of compressive strength of
mortars based on various binders. Note that
R=RHA, l=lime, P=PC, W=Water and B=Binder.
Codes are the same as in Fig. 7.
Code Mixture W/B [g/g] Slump [mm] Compressive strength [MPa]
Composition
4 0.5(R)0.5(1) 0.92 5.0 1.30 1.72 2.55 2.29 2.48
9 0.7(R)0.2(l)0.1(P) 0.90 2.0 1.17 1.53 1.83 1.63 1.55
11 0.6(R)0.3(1)0.1(P) 0.90 2.0 1.38 2.11 2.29 2.47 2.31
12 0.5(R)0.4(1)0.1(P) 0.90 2.5 1.37 2.46 3.10 3.47 3.27
14 0.65(R)0.35(P) 0.90 0.5 2.32 5.85 8.59 7.81 9.08
15 0.5(R)0.5(P) 0.875 3.5 2.29 6.50 10.38 12.60 13.60
The compressive strength of RHA-lime-PC mixes was found to fulfil the requirement of
ASTM C 91 for masonry work even for a PC content under 50%. The RHA-lime mixes
performed less well, except in case of mix 4 in which the RHA content was lowest (i.e
50%). For the primary purposes of densification of a floors surface and of stabilization
of the mud for walling this is sufficient.
The conditions in open heaped-up burning of the rice husk material are for most of the
material probably similar to or even better than in the specially designed kiln used in the
experiments [9]. The present investigations did not reveal a significant portion of the
silica to be transformed into the crystalline phase for temperatures up to 700C. Thus,
open heaped-up burning would provide a cheaper solution for the primary applications of
RHA in Tanzanian villages.
RHA contains only small amounts of aluminium oxide. In order to improve strength
and to increase setting times, activated clay could be added. Cabrera et al [10] showed
that mixes containing activated red tropical soils from the Southern Indian Shield in Sri
Lanka to be superior over other mixes in which OPC was blended by PFA (containing
more Al2O3 than the RHA). The non-activated clay used in a part of the present
Rha-cement as a replacement for portland cement in rural tanzanian villages 319
experiments gave rise to a strength reduction. Data are therefore not included in this
paper.
Two brick-type of ovens are projected in Kalebezo village, Sengerama District, Mwanza
Province, to also start the more controlled production of RHA (and clay) aiming for more
advanced purposes, like the local production of bricks, tiles, etc. In some cases small
additions of PC will be considered.
REFERENCES
1. MEHTA, P K. The chemistry and technology of cement made from rice husk ash. Proceedings
UNIDO/ESCAP/RCTT Workshop on Rice Husk Ash Cement, Peshawar, Pakistan, Regional
Center for Technology Transfer, Bangalor, India, 1979, pp 113122.
2. Proceedings Fifth CANMET/ACI International Conference on Fly Ash, Slag, Silica Fume and
Natural Pozzolans in Concrete, Nat. Resources Canada, Ottawa, 1995.
3. HOUSTON, D F. Rice Chemistry and Technology, American Association of Cereal Chemistry,
Monograph Series, Vol. IV, 1972.
4. SABUNI, E L. Physico-Chemical Investigations of Tanzanian Rice Husks and of Rice Husk Ash
for Use in Rural Tanzania, Report 03.21.1.32.21, Fac. Civ. Engr., Delft Univ. Techn., Jan. 1995.
5. SABUNI, E L. Research into the Potentialities of Rice Husk Ash Cement for Application in
Rural Tanzania, Report 03.21.1.32.05, Fac. Civ. Engr., Delft Univ. Techn., Jan. 1995 (with 27
references).
6. STROEVEN, P, DALHUISEN, D H, QUY, N T, BUI, D D AND DONG, A V. Toward ga-
graded HPC in Vietnam using crushed rock, very fine sand, and RHA blended cement.
Proceedings Second CANMET/ACI Int. Symposium: Advances in Concrete Technology, Ed. V
M Malhotra, SP 154, Natural Resources Canada, Ottawa, 1995, pp 263278.
7. BUI, D D. Gap-Graded Concrete With Very Fine Sand Using Portland RHA Blended Cement,
Report 03.21.1.32.04, Fac. Civ. Engr., Delft Univ. Techn., August 1994 (with 42 references).
8. STROEVEN, P, DALHUISEN, D H, QUY, N T, BUI, D D AND DONG, A V. Toward Gap-
Graded HPC in Vietnam Using Crushed Rock, Very Fine Sand, and RHA Blended Cement,
Advances in Concrete Technology, SP-15414, Ed. V M Malhotra, Amer. Concr. Inst, 1995, pp
263278.
9. SHIMIZU, G AND JORILLO Jr, P. Study on the use of rough and unground ash from an open
heaped-up burned rice husk as a partial cement substitute. Vegetable Plants and their Fibres as
Building Materials, Ed. H S Sobral, Chapman and Hall, London, 1990, pp 321333.
10. CABRERA, J G AND NWAUBANI, S O. Strength and chloride permeability of concrete
containing red tropical soils, Mag. Concr. Res, 45, 164, 1993, pp 169178.
EFFECT OF THE BINARY
ADMIXTURES OF FLY ASH AND
SHALE ASH ON THE FLUIDITY AND
COMPRESSIVE STRENGTH OF
CEMENT CONCRETE
N-Q Feng
Q-F Zhuang
D-H Wang
Tsinghua University
China
Keywords: Binary Admixtures, Portland Cement(PC), Fly Ash (FA), Shale Ash (SA),
High Performance Concrete(HPC), Fluidity.
Professor Feng Naiqian is Director of the Building Materials, Tsinghua University,
Beijing, China. He specializes in the use of zeolite as a kind of admixture. In recent years,
he has done much research on high performance concrete with the use of zeolite, fly ash,
shale ash, silica fume and slag. He also specializes in fracture mechanics of concrete. He
has published widely and serves on many Technical committees.
Dr Zhuang Qingfeng is a doctor student of Prof. Feng Naiqian. He is working on the
high performance concrete and its fracture properties.
Appropriate concrete technology 322
Dr Wang dehuai is a doctor student of Prof. Feng Naiqian. He is working on the mix
proportions of high performance concrete.
RAW MATERIALS
1. Fly ash: the first grade fly ash produced by Huangpu Power Station was used. Its
physical properties and chemical compositions are listed in Table 1 and Table 2.
2. Shale ash: from Maoming, Guangdong Province. Its chemical compositions are
listed in Table 3.
3. Coarse aggregate: crushed gravel, particle size range 5~20 mm, bulk density 1340
kg/m3, specific gravity 2.65, silt content 0.6%, flaky and elongated content 19.7%.
4. Fine aggregate: river sand, fineness modulus 3.0, bulk density 1540 kg/m3, specific
gravity 2.6. absorption 0.5%.
5. Naphthalene superplasticizer: Brand NF.
6. Cement: #525 Portland cement.
EXPERIMENTAL DESIGNS
1. With a constant 550 kg/m3 of binder (350 kg/m3 of cement and 200 kg/m3 of binary
Table 1 Physical properties of fly ash
SIEVE LOSS ONb WATER MOISTURE BULK
RESIDUE IGNITION REQUIREMENT CONTENT DENSITY
(size of 45 m) RATIO* (kg/m3)
12% 5% 95% 1% 548~563
* Assume the water requirement of mortar (cement: sand=1:2.5, by weight) with pure cement to be
W1, and that of another mortar (cement: fly ash : sand=0.7:0.3:2.5, by weight) which has the same
fluidity with the former to be W2. Then the water requirement ratio of fly ash is defined as
(W2/W1)100%.
admixtures), and water/binder ratio of 0.35, the varieties of compressive strength and
fluidity with different FA/SA ratios were investigated.
2. With a constant 560 kg/m3 of binder (336 kg/m3 of cement and 224 kg/m3 of binary
admixtures), the varieties of compressive strength were investigated with different
water/binder ratios (0.32, 0.34, 0.35, 0.38), so did the effect of FA/SA ratio on the
compressive strength with a constant water/binder ratio of 0.32.
3. With a constant 570 kg/m3 of binder (314 kg/m3 of cement and 256 kg/m3 of binary
admixtures), the varieties of compressive strength were investigated with different
water/binder ratios (0.31, 0.32, 0.37), so did the effect of FA/SA ratio on the compressive
strength with a constant water/binder ratio of 0.31.
Appropriate concrete technology 324
4. With a constant 580 kg/m3 of binder (290 kg/m3 of cement and 290 kg/m3 of binary
admixtures), the varieties of compressive strength were investigated with different
water/binder ratios (0.30, 0.31, 0.32, 0.34, 0.40)
The correspondent concrete mix proportions are listed in Table 4.
Table 5 The experimental results
SERIES No. SLUMP COMPRESSIVE STRENGTH (MPa)
(mm)
3d 7d 28d
1 1 165 33.9 45.0 55.6
2 150 45.8 55.7 66.5
3 190 49,1 58.4 64.3
4 105 54.8 60.6 64.3
2 5 175 29.1 49.2 60.3
6 190 36.8 56.8 63.0
7 190 24.9 53.5 58.9
8 180 32.6 47.5 55.1
9 180 35.6 49.7 51.3
3 10 205 31.0 39.3 61.6
11 215 36.1 48.2 64.0
12 175 39.1 52.9 65.2
13 180 38.0 49.9 62.4
14 95 33.2 44.3 53.6
4 15 205 36.1 48.1 62.4
16 180 40.4 54.0 63.2
17 185 39.4 55.7 60.2
18 180 36.6 50.7 57.6
19 195 32.9 42.6 56.0
PREPARATION OF SPECIMENS
According to each mix proportion, 0.015 m3 of fresh concrete was mixed in a forced
mixer for 3 minutes. After the measurement of slump value, it was molded with a
standard vibration, and the mold was removed after 24 hours. All specimens were cured
under the standard conditions for the test of compressive strength.
Effect of the binary admixtures of fly ash 325
3. Keeping the binder and mineral admixtures contents the same, the strengths of
concretes with different water/binder ratios and the relationships between the concrete
strengths and FA, SA proportions on the same water/binder ratio are shown by Series 3 in
Table 5 and Figure 3. It can be seen that: the larger the binder/water ratio, the higher
the concrete strength. With the same water/binder ratio (0.31), the effect of FA
admixture(256 kg/m3) only is less distinctive than that of binary admixtures
(FA:SA=4:1~3:1), which is also consistent with the discussion of part 1.
4. Keeping the contents of binder and mineral admixtures fixed as 580 kg/m3 and 290
kg/m3, respectively, the strengths with different water/binder ratio are shown by Series 4
in Table 5 and Figure 4.
Effect of the binary admixtures of fly ash 327
It is a general tendency that concrete strengths drop with the increment of water/binder
ratio. But the proportion of FA and SA in concrete can affect its strength considerably.
For example, the strengths at 3, 7, and 28 days of concrete with only FA admixture and
W/C of 0.30 (No. 15) are less than those of concrete with FA-SA binary admixtures and
W/C of 0.31 (No. 16). Thus, it is important to use binary admixtures and use them at
proper proportions in concrete.
Furthermore, even with the same water/binder ratio, the proportion of cement is very
important. For example, the cement contents of No.3 and No.8 are 350 kg/m3 and 336
kg/m3, the strengths at 3, 7, and 28 days of the former are 50%, 23%, and 17% higher that
those of the later, respectively, but the water/binder ratios and the proportions between
FA and SA are the same. Therefore, in order to gain higher early strength, the cement
content cannot be too low.
CONCLUSIONS
1. High performance concrete with slump of 1820 cm and compressive strength of more
than 60Mpa can be made with a cement content of 350 kg/m3, a mineral admixtures
content of 200 kg/m3, a water/binder ratio of 0.35, and superplasticer of 0.8% by weight
of binder.
2. The strength of concrete with FA and SA binary admixtures (FA: SA=4:1~3:1) at 3,
7, and 28 days are higher than that with FA admixture only.
3. In order to produce high performance concrete with compressive strength of more
than 60 MPa, the cement content should be no less than 350 kg/m3.
A STUDY OF CARBONATION IN
BINDING SYSTEM WITH CEMENT
AND HYDRAULIC ADMIXTURES
I Robu
Civil Engineering Institute
Romania
Keywords: Carbonation, Fly ash, Silica fume, Ordinary Portland cement, Curing,
Corrosion.
Dr Ion Robu is a Senior lecturer at the Civil Engineering InstituteBucharest, chair of
the Chemistry of Building Materials Department. His research concerns pozzolanic
binders, possible utilisations of local materials and industrial wastes and all aspects
related to concrete durability.
INTRODUCTION
The utilisation of industrial wastes such as: slag, fly ash, silica fume, sludge,
phosphogypsum is nowadays an economic and ecological necessity throughout the world.
In Romania, a particular attention has been paid to the utilisation of fly ash and silica
A study of carbonation in binding system with cement and hydraulic admixtures 329
fume in hardening structures that make use of their pozzolanic ability in binding systems
with or without cement [1]. The use of these two wastes in cement based hardening
structures (pastes, mortars, concretes) requires improvements in composition and in
curing, both associated with the capacity of carbonation of these structures. Therefore,
this paper deals with establishing the carbonation depth for different proportions of
hydraulic admixtures as cement substitute in connection with other composition and
curing factors that influence the carbonation capacity. The results of the study are limited
at ages of 23 years, but they will be completed with new data in time. Some very
interesting documentary syntheses and critical appreciations have been done in Romania
in the field of concrete carbonation and reinforcement corrosion [2,3], that should be
completed with the longterm research of the Japanese scientists [4,5].
EXPERIMENTAL DETAILS
Materials
Ordinary Portland cement I 35 (type BS 12:1978), silico-aluminous fly ash and silica
fume were used in these tests. Quartz sand in 0.082 mm size for mortars and natural
aggregates up to 31 mm size for concretes were also used.
Mix Proportions
The proportions of hydraulic admixture (substituting cement) in pastes and mortars were
of 0, 10, 20, 30, and 50%. The water/cement (W/C) or water/cement+ admixture
(W/C+A) ratio was fluctuating, but the consistency was kept constant. To stand out the
influence of W/C or W/C+A ratio, various types of pastes were used: normal consistency
pastes and more fluid ones (+20% water). For mortars, two consistencies were also used:
C1=3 cm and C2=8 cm. Only fly ash in the following dosages was used in the preparation
of concrete specimens:
100 kg/m3 (10 FA1) and 180 kg/m3 (10 FA 2) for C10, workability S1: slump=10
mm.
100 kg/m3 for C15 and C20 (15 FA 1 and 20 FA 1), workability S2: slump =55 mm.
A variable quantity of mixing water ensured the same workability for control test
concretes and for fly ash admixture concretes. For C15 a more fluid concrete was
prepared (slump=95 mm, noted F).
Curing Environments
The specimens were preserved for different periods of time (5, 14, 21 days and one year)
in moist air (95% RH), followed by curing in ordinary conditions of temperature and
humidity (23C, 5060% RH). The concrete specimens were cured indoors and outdoors
after moist air curing.
Appropriate concrete technology 330
The test specimens for the carbonation study were prisms of 4040160 mm for pastes
and mortars and cubes with 100 mm side for concretes. For the reinforcement corrosion
study, cubes with 100 mm side and cylindrical specimens 150 mm in diameter and 150
mm high were used (this study is not subject of the present paper). In order to appreciate
the concrete degree of carbonation, several methods are quoted in literature: the
colorimetric test with alcoholic solution phenolphthalein 1%, thermal analysis, X-ray
diffraction, pH-method, radioactive CO2 marking and the use of manganese hydroxide
[6]. We have mainly used the colorimetric measuring the average carbonation depth (Dc)
in the discoloured (neutralised) zone. The carbonation depth (higher in the upper part)
was established in transversal sections obtained by splitting the specimens by the means
of special designed devices. We also used the thermal analysis and X-ray diffraction on
pastes and mortars with significant carbonation.
Pastes Carbonation
The results of carbonation of cement pastes with or without hydraulic admixtures are
presented graphically in Figures 1a, b, c, d for normal consistency pastes. The influence
of the hydraulic admixture nature is shown in Figure la (silica fume pastesSF) and in
Figure 1b (fly ash pastesFA). Silica fume reduces more than fly ash the intergranular
solution alkalinity and contributes to more compact structures that oppose to the diffusion
of the carbonic gas. When using silica fume, it is necessary to homogenise it well with
cement, in dry conditions, and to mix longer the paste after adding the water (because of
the emphasised agglomeration tendency). Figures 1c and 1d reveal the influence of initial
curing period in moist air for paste of cement with fly ash admixture; it can be seen that
the carbonation depth diminishes with the increase of the duration, due to the decrease in
micro-cracking trend. For fluid pastes, the average carbonation depth was higher than the
one obtained for normal consistency pastes; in this case, the capillary porosity is higher
and the diffusion of the carbonic gas is easier. The data obtained show clearly the
influence of composition and curing factors on carbonation.
Concretes with characteristic resistance of 10, 15 and 20 N/mm2 were used, for which the
carbonation is emphasised. The average carbonation depth for these concretes is shown in
Figures 3, 4 and 5. One can see the significant influence on carbonation of the following
factors: concrete resistance class, proportion of admixture (fly ash), fresh concrete
workability and initial preserving period in moist air. For concrete cured indoors, the rate
of carbonation was much higher than that of concrete cured outdoors because the
concentration of CO2 was higher and there was no additional supply of water. In
Romania, studies on concretes with silica fume admixture up to 20% were also done.
show alterations of diffraction lines between the carbonated zone and non-carbonated
one. Latest data regarding 3-year carbonation indicate that the moderation trend continues
in time. The evolution of carbonation
CONCLUSIONS
1. Natural carbonation occurs at all cement based hardening structures with or without
hydraulic admixtures. The average carbonation depth reaches maximum values for air-
curing (with no initial hardening period in moist air) and minimum values after long
Appropriate concrete technology 334
preserving in moist air. An important role has the W/C or W/C+A ratio which influences
the porosity and the carbonic gas diffusion.
2. At the hardening structures with hydraulic admixtures, the carbonation depth rises
along with the proportion of cement substitute admixture; up to 1020% of hydraulic
admixture, the values of the carbonation depth do not differ too much of those of the
control tests.
3. The colorimetric test used to reflect the carbonation proved to be convenient and
fast. Thermal analysis and X-ray diffraction provide differentiations between the
carbonated layer and the non-carbonated one, but they are difficult to exploit in practice.
4. An initial curing period of 714 days in moist air would eliminate the risk of steel
depassivation through carbonation for reinforcements with minimum 20 cm cover. When
the carbonation reaches the steel corrosion becomes possible.
REFERENCES
1. Robu, I. Binding compositions with fly ash, Ph. D. Thesis, Civil Engineering Institute, 1986.
2. Steopoe, A. Carbonation of cement, mortars and concretes, Rev. Materiale de Constructii, 2(4),
1972, pp 179189.
3. Bob, C. Theories on service life of reinforced concrete buildings, The Symposium Behaviour in
situ of buildings, INCERC, Bucuresti, 1988.
4. Tsukayama, R., Abe, H. and Nagataki, S. Long-term experiments on the neutralisation of
concrete mixed with fly ash and the corrosion of reinforcement, 7-th International Congress on
Chemistry of Cement, Paris, 1980.
5. Nagataki, S., Ohga, H., Kim E.K. Effect of curing conditions on the carbonation of concrete with
fly ash and the corrosion of reinforcement in long-term tests, International Symposium, Madrid,
1986.
6. Neville, A.M. Properties of concrete, Ed. Tehnica, Bucuresti, 1980.
APPLICATION OF THE CEMENT
HYDRATION EQUATION TO
CONCRETE
K K Sideris
M S Konsta
C G Karayannis
Democritus University of Thrace
Greece
INTRODUCTION
Since strength at age of 28 days is regarded as a criterion of desirable quality and design,
the estimation of the strength of concrete at later ages is not usually accomplished,
because long time studies, up to the end of the hydration, are needed. Such studies are not
available in the literature. The solution in this problem is given by the cement hydration
equation [14]. Using this equation it is possible to predict the compressive strength of
concrete after the age of 28 days, without performing compressive tests.
The Cement Hydration Equation introduced by Prof. K.Sideris [14] is expressed by
K=Kb*tP, where K represents a hydration criterion, K and b are constants, t is the
hydration time and p is the hydration coefficient. As hydration criteria can be considered
the heat of hydration, the non-vaporable water, the Ca(OH)2 elimination, the density, the
total porosity, the water sorption or the compressive strength (of cement paste, mortar or
concrete).
The Cement Hydration Equation was applied to the compressive strength of concrete
which was determined for a hydration time up to 90 days [5,6,7]. An example was given
for the values of the compressive strength of concrete according to F.M.Lea [8] as it is
shown in Figure 1.
In the present work the Cement Hydration Equation is applied in order to predict the
compressive strengths of two concrete mixtures with different cement content, up to the
age of 5475 days. The hydration number p is determined from the mortar compressive
strength according to DIN 1164 for hydration times 3, 5, 7, 10, 14, 18 22 and 28 days,
according to [1], for more than 3 measured values of the compressive strength.
The first and second phases of the hydration [14], characterized by a hydration criterion
K in the co-ordinate system with K on the ordinate and (1/t)p on the abcissa, follow a
straight line path. The quantitative relationship is given by the following equation:
K=Kb*tp (1) Equation (1) corresponds to the exponential function of the general form
y=ab*tp in which:
K: is the hydration criterion (heat of hydration, non-vaporable water,
porosity, density, sorption, Ca(OH)2 elimination, compressive strengths of paste, mortar and
concrete, etc.)
K: is a constant, which represents the intersection of the straight line of equation (1) with the
ordinate
b: is the slope of the straight line of equation (1).
t: is the hydration time expressed in days (t>>0).
p: is the hydration number, depending only on the chemical composition of the cement type
used.
The cement hydration equation (1) has been applied to a number of criteria indicated
below:
Appropriate concrete technology 338
Equation K=Kb tP applies to the hydration criteria shown in table 1, only if the type of
the cement used remains the same (Portland cement, pozzolanic cement, p.f.a. cement
and slag-rich blast furnace cement) [14]. It is important to notice that the value of the
hydration number p remains the same as long as the chemical composition of the cement
type used also remains the same.
The value of the hydration number p does not depend on:
1. the hydration criterion used 3. the fineness of the cement
2. the curing temperature [760] 4. the ratio w/c
5. the aggregates grading curve and the aggregates quality when the hydration criterion of the
compressive strength of mortar or concrete is examined.
The value of the hydration number p can be estimated with the use of any hydration
criterion shown in table 1. The accuracy of the determination of the value of p depends
on the accuracy of the measured values of the hydration criterion used.
In this case three measured values (K1, K2, K3) of the hydration criterion used are
needed. These values must be determined with high accuracy in order to obtain a reliable
value of the hydration number p. Even a small deviation of one or more of the measured
values leads to unreliable values of p. Such accuracy of the measured values K1, K2 and
K3 may be obtained only using the hydration criterion of the heat of hydration. The use
of the formula shown above is not recommended when the hydration criterion of the
compressive strength of cement paste, mortar or concrete is used. In this case the second
method mentioned below is suggested.
Method 2: This method can be used with any of the hydration criteria shown in table 1. It
is specially recommended though for the evaluation of the compressive strength, since
the measured values in this case are not of high accuracy. In this case the use of 68
measured values is recommended. The value of p is estimated with higher accuracy if the
criterion of the compressive strength of mortar is used, according to DIN 1164 or
ASTM/C-349.
In order to evaluate the value of the hydration number p, the compressive strength of a
mortar mixture (M.M. 1:3:0.5) was experimentally determined for the ages of 1, 2, 3, 5,
7, 10, 14, 18, 22 and 28 days. Specimens of dimensions 4416 cm were prepared
according to DIN 1164.
Pozzolan cement was used, up to 35% content in pozzolan per weight, with a fineness of
3900 cm2/g Blaine and compressive strength equal to 45.2 N/mm2, as specified in DIN
1164. Limestone aggregates with a maximum
Table 3 Determination of equations of compressive
strength of concrete.
Cement content 275 Kg/m3
Age (days) 3 7 14 28 50 90 180 365 730 5475
Measured values 11.9 18.4 24.0 28.7 31.5 35.3 38.5 41.8 44.4
2
of fW (N/mm )
Linear regression fw=67.61188, b=66.69468, r=0.9997398, s=0.2653123
for the pairs Compressive strength equation:
fwMeas(1/t)0.16
fw=67.6118866.694688*(1/t)0.16 (N/mm2), t3
Application of the cement hydration equation to concrete 341
fWRegression 11.7 18.7 23.9 28.5 31.9 35.2 38.6 41.7 44.4 50.8
3
Cement content 300 Kg/m
Age (days) 3 7 14 28 50 90 180 365 730 5475
Measured values 16.8 23.8 28.3 32.9 37.1 39.6 42.7 46.4 48.6
2
of fw (N/mm )
Linear regression fw=71.4227, b=65.2165, r=0.9995724, s=0.3325698
for the pairs Compressive strength equation:
fwMeas(1/t)0.16
fw=71.422765.2165*(1/t)0.16, (N/mm2), t3
fWRegression 16.7 23.6 28.7 33.2 36.5 39.7 43.0 46.1 48.7 54.9
particle size of 32 mm were used. For the aggregate grading curve (B32) 8 fractions
(32/16, 16/8, 8/4, 4/2, 2/1, 1/0.5, 0.5/0.25, 0.25/0) were used as specified in DIN 1025.
The cement content of concrete was 275 Kg per m3 for the first mixture and 300 Kg per
m3 for the second mixture. The water to cement ratio was equal to 0.60 for each case. The
final compressive strength was calculated as the mean strength of the experimentally
measured values of three specimens. Results are shown in table 3.
Based on the values of table 2 and the hydration number p=0.16 a linear regression
analysis of the compressive strength values fw in respect to the corresponding hydration
time, expressed by (1/t)0.16, was performed for the cases of 3, 5, 7, 14, 28, 50, 90, 180,
365 and 730 days.
Using the results of this linear regression, the equation of the compressive strength for
each of the concrete mixtures was defined. Linear regression results as well as the
compressive strength equations are presented in Table 3. The variation of the
compressive strength versus time is presented in figure 4.
CONCLUSIONS
The equations which correlate the compressive strength of concrete mixtures with time,
can be formulated by the application of the cement hydration equation. Using these
equations, the compressive strength of concrete can be reliably estimated at any age, up
to the end of the hydration period, which is considered to be in 15 years (5475 days) [1
4], without performing laboratory tests for long time periods. The formulation of these
equations can easily be obtained using values of the concrete compressive strength,
measured at the early ages of the hydration [1, pp 344, chapter 7].
Application of the cement hydration equation to concrete 343
REFERENCES
INTRODUCTION
Due to the escalating cost of affordable housing in Zimbabwe in recent years, it has
become necessary to seek ways of reducing the cost of housing. Building materials
constitute a major input as they sometimes account for as much as 75% of the cost of
low-income houses.
Large quantities of industrial wastes are being generated in the steel and chemical
industries. The disposal of these wastes poses problems, it is cheaper to find alternative
use for the wastes. The ZISCO-STEEL company in Zimbabwe produces granulated blast-
furnace slag at the rate of 500 kg per tonne of hot metal. This gives about 340,000.00
tonnes of slag annually. Hence the slag mountains adjacent to the steel works.
The present work is part of a program, at the University of Zimbabwe, designed to
study and to promote the use of industrial wastes in construction. The use of GGBS as a
partial cement replacement material in the production of fibre reinforced concrete Roman
II type of tiles have been studied and presented. The methods of testing the tiles have also
been studied and further suggestions are presented.
Zimtiles and Parry Instamac are among the largest tile producers in Zimbabwe. The
tiles used in this work were produced at the Parry Instamac works in Harare. The main
objective of this investigation is to provide information for the use GGBS in the
production of roofing tiles locally.
Materials
The materials used in producing the tiles are Portland cement, sand, water and fibre. The
basic mix is 1:3 cement: sand with the addition of 1% fibre. The most commonly used
fibres are sisal, jute, henequen or coconut coir. The fibres should be flexible and not
brittle and should not contain oily substances or sugar. The sisal fibre used as
reinforcement in the tiles originate from sisal plants which are grown extensively in
Zimbabwe. Sisal is reported to be one of the strongest natural fibres available. A report
by Aziz (4) show that sisal fibres have an average diameter of 0,3 mm and a tensile
strength of between 280 and 560 MPa. The fibre length is in the range 50 mm to 150 mm.
The fibres were obtained by scrapping the sisal leaves, drying and chopping them ready
for use. Gram (5) has shown that the durability problems in the use of sisal fibre
reinforcement in tiles are due to alkali attack of the fibres by the pore water present in the
cement matrix. This can be minimized if the alkalinity of the pore water is reduced by the
addition of pozzolanic materials such as GGBS.
The GGBS used in the investigations originated from ZISCO-STEEL plants in
Redcliff and were marketed by FOSROC. The chemical composition is shown in Table 1.
Appropriate concrete technology 346
Component FeO K Pb Mo P B Zn
% 0.5 0.22 0.02 0.02 0.012 0.011 0.003
The Parry Associates Roman II tile is sketched in Figure 1 and is also shown on a
completed roof structure in Figure 2. The tiles have a large overlap, deep valley and close
fit to ensure that the roofs are leak proof without the need for underfelt or plastic
sheeting. They are normally available in three thicknesses; 6 mm, 8 mm and 10 mm.
Production
The main equipment for making the tiles is a vibrating screeding machine powered by a
multivibe vibrator on which a shaped flat screed is produced, and the moulds which give
the screed its 3-dimensional tile form as shown in Figure la. The basic dimensions are
shown in Figure 1b. The production technique is described in the Parry Instamac manual
(1).
Series of tiles were made by replacing part of the cement in the control by GGBS in
varying proportions. Samples containing the following percentages of GGBS in the
binder were made: 0%, 10%, 20%, 30%, 40%, 50%, 60% and 80%.
A low cost curing technique was used which involves the use of a sealed tank covered
with polythene sheet and exposed to direct sunlight. Figure 3 shows a cross section of the
tank. The water level in the tank was just below the surface of the gravel such that the
tiles were not in direct contact with static water. Static water tends to leave map traces on
the tiles.
The heat inside the sealed tank causes vapour to rise thereby providing moisture for
curing. A high temperature and humidity condition therefore prevails within the tank. The
probe of a digital thermometer was inserted in the tank before it was sealed.
An average temperature of 28C was obtained in the curing tank over the seven-day
curing period (Sept. 20Sept. 26, 1994) compared with the average ambient temperature
of 22C. The registered day temperatures were up to 60% higher than the ambient
temperature. The difference between the night temperatures was much lower Figures 4
and 5 show the temperature readings for 1 day and 7 days respectively.
Appropriate concrete technology 348
The humidities both inside and outside the curing tank were measured with wet and dry
bulb hygrometers and a hygrometric table used to compute the relative humidities. The
average relative humidity in the tank was 90% compared with an average of 38%
humidity for the surrounding area.
The tiles were demoulded after 24 hours and then placed in the curing tank for 7 days.
Previous tests have shown that curing for longer periods did not provide significant
additional benefit.
After removal from the curing tank, the tiles were left to cure in air for a further 7 days
before they were ready for use. The same standard procedure was followed for making all
the samples.
Properties of high temperature-humidity cured OPC-GGBS fibre concrete roofing tiles 349
EXPERIMENTAL
Water Permeability
Water permeability tests were carried out to the procedur described by Johansen (6). A
similar method is also used at the Parry Instamac production yard for testing tiles. Four
tiles were selected from each sample for the test. Two small weirs were formed on each
tile and water was allowed to stand in a pool for a period of 24 hours. A tile was
considered to have failed the test if after the 24-hour period there was free water on the
underside. The results of the water permeability tests are shown in Table 2.
Table 2. Results of the permeability tests
SAMPLE GGBS REPLACEMENT, % TYPE OF ADMIXTURE RESULT
C-0 0 No admixture Pass
C-1 10 No admixture Pass
C-2 20 No admixture Pass
C-3 30 No admixture Pass
C-4 40 No admixture Pass
C-5 50 No admixture Pass
C-6 60 No admixture Pass
C-7 70 No admixture Pass
C-8 80 No admixture Fail
20 0 Conplast P211 Pass
21 10 Conplast P211 Pass
22 20 Conplast P211 Pass
23 30 Conplast P211 Pass
24 40 Conplast P211 Pass
25 50 Conplast P211 Pass
26 60 Conplast P211 Pass
27 70 Conplast P211 Fail
28 80 Conplast P211 Fail
40 0 Conplast P430 Pass
41 10 Conplast P430 Pass
42 20 Conplast P430 Pass
43 30 Conplast P430 Fail
44 40 Conplast P430 Fail
Appropriate concrete technology 350
Strength Tests
Tests for the strength in bending were carried out using three methods:
i) The method described in BS 550 part 2:1971;
ii) The method described in the ITW/Parry manual;
iii) A modification of the ITW/Parry method.
The BS method recommends that six tiles be selected for testing from each sample. The
tiles were immersed in water for 24 hours and tested soon after removal from water. The
tile to be tested was supported on two bearers. The distance between the bearers being
two thirds of the length of the tile. The load is applied centrally through a third bearer of
diameter 38mm. A sketch of the arrangement is shown in Figure 6. The loading was
applied at the rate of 800 N/minute.
The ITW/Parry method recommends that the load be applied evenly across the center
of the tile as sketched in Figure 7. The wooden platen of 25mm width is designed to
follow the profile of the top surface of the tile as shown in Figure 7.
A third testing arrangement was devised in which the supports and the bearer for load
application were designed to follow the surface profile of the tile supported. The
arrangement is sketched in Figure 8.
Properties of high temperature-humidity cured OPC-GGBS fibre concrete roofing tiles 351
A comparative result shown in Figure 9 shows that a much lower strength was obtained
from the BS method in comparison with the other two arrangements. This is due to the
nature of the loading and the shape of the tile. In the BS method the platen rests on the
tile at two points where the loads are applied as point loads. The tiles failed in direct
tension or through the development of local stresses at corners rather than in bending.
The BS method is designed to suit press extruded tiles. These tiles are comparatively flat
in cross section and have the same overall height at several points as shown in Figure 9.
Properties of high temperature-humidity cured OPC-GGBS fibre concrete roofing tiles 353
The rest of the results were obtained from tests to the BS method. Figure 10 shows
average test results at 28 days and 3 months for the tiles.
It is generally observed that the best strength results were obtained for GGBS content of
20% for samples cured in the tank under the high temperature and humidity conditions.
For samples cured at room temperature the strength generally decreased with increase in
GGBS.
The BS transverse strength requires a minimum breaking load equal to 3.2 times the
effective width (200mm for the Roman II tiles). This requirement is easily met in samples
Appropriate concrete technology 354
containing up to 30% GGBS at 28 days. The other samples could generally fulfil the BS
requirement at older ages.
Another difference between the test methods is the fracture mechanism. In the BS
testing the tile cracks and fails along the length of the tile while the ITW/Parry method
fails across the width as shown in the sketch in Figures 6 and 7. The two methods,
therefore, give cracks that are at right angles to each other.
When the tiles were being demoulded at 24 hours, some breakages were observed
particularly for tiles containing high percentages of GGBS. It became obvious that more
curing is required for such tiles. Samples with GGBS content of 30% or less did not
experience handling breakages.
The permeability test showed a general tendency of increase in permeability with
GGBS content. The use of admixture did not improve the permeability of the tiles.
The investigations have shown that the strength of the roofing tiles can be improved
by replacing a limited amount of the cement with GGBS if the curing technique described
above is used. It has also been demonstrated that the testing arrangement described in the
BS 473 (2) does not give a true bending strength of the tiles. To obtain the strength in
bending of the Roman II tiles, it is necessary to apply the load through a platen that
matches the surface profile.
The use of water reducing admixtures is not beneficial to the tiles as regarding the
durability and water permeability.
More work is currently being done to smooth out the strength curve between 0% and
30% GGBS content and also to investigate the influence of the fibre reinforcement on the
increase in strength.
ACKNOWLEDGEMENTS
The author wishes to thank Parry Instamac (Pvt.) Ltd. Zimbabwe and the director John
Duxbury for making the tiles and for information received.
REFERENCES
1. ITW Fibre concrete or vibracrete tile technology; Users manual: JPM Parry and Associates.
2. BS 473, 550: Part 2:1971 Specification for concrete roofing tiles and fittings.
3. Ground granulated blast-furnace slag as a cementitious constituent in concrete: reported by ACI
committee 226: ACI materials Journal, July-August, 1987.
4. AZIZ A., PARAMASIVAM P and LEE S.L.(1975); Concrete reinforced with natural fibres;
Concrete Technology and Design Vol.2, Surrey University Press pp. 319.
5. GRAM H.E. (1988), Natural fibre concrete roofing. Concrete Technology and design vol. 5,
Blackie and son ltd. London pp. 256285.
6. JOHANSSON B. Concrete roofing tiles. Building Issues, Lund University Centre for Habitat
Studies. Vol 4, No 2, 1995, pp 415.
MICROPORE STRUCTURE
VARIANCES DURING HYDRATION OF
CEMENT PASTES
Ch Ftikos
A Georgiades
National Technical University of Athens
J Marines
Heracles General Cement Co
Greece
Keywords: Ordinary Portland cement (OPC), Blended cements, Pozzolans, Fly ashes,
Hydration of cement, Aggressive media. Pore sizes Distribution.
Professor Christos P Ftikos is Phd Chemical Engineer, Director of the Lab. of Inorganic
Materials Technology, Chemical Engineering Dept, Technical University of Athens,
Greece. He specialises in Cement and Concrete Technology, Pozzolans and Durability of
their products, as well as in Electroceramics. Professor Ftikos has published widely and
serves oh many Technical Committees.
Dr Andreas A Georgiades is Phd Chemical Engineer, Lecturer in the Lab. of Inorganic
Materials Technology, Chemical Engineering Dept, Technical University of Athens,
Appropriate concrete technology 356
INTRODUCTION
Both practice and research have already demonstrated that porosity and pore size
distribution mainly affect some physical and mechanical properties of hardened cement
pastes. Various formulas have been proposed concerning relationships between
compressive strength and porosity [1]. Several studies are also referred to the durability
and permeability of concrete related to its pore structure [2]. Curing conditions as well as
composition of the pastes in particular mineral admixtures influence total porosity and
pore size distribution [3].
It is also reported that cement pastes containing pozzolans have greater porosity than
those made from Ordinary Portland Cement [4]. B.K.Marsh et al [5] suggested that
making cement mortars with seawater instead of fresh water intails a greater pore volume
in pore diameters less than 100 A and lower volume in the diameter range of 10010000
.
The purpose of this study has been to study the pore size distribution of several cement
pastes during their curing with different baths such as fresh water, sea water and solutions
containing Cl and Mg2+.
EXPERIMENTAL
Pastes were prepared from these cements in an appropriate mixer, using a W/C ratio of
0.25, 0.27, 0.30 and 0.33 for the cements C1, C2, C3 and C4 respectively, as determined
by standard consistency test (ASTM C-187). The pastes were moulded in bars 160mm
x20mm x20mm and cured for 24 hours at 202C and R.H=90%. After that the pastes
were cured for 7, 28, 90, 180 and 360 days in the following curing baths, at 202C:
a. Fresh water.
b. Aq Solution of 2000 ppm Cl ions, coming from dissolution of NaCl.
c. Aq Solution of 2000 ppm Mg2+ ions, coming from dissolution of Mg(NO3)2.
d. Aq Solution of 2000 ppm Cl ions, 3000ppm SO42 ions and 2000 ppm Mg2+ ions,
coming from dissolution of NaCl, Na2SO4 and Mg(NO3)2 respectively and named
Seawater.
At predetermined ages samples of the prepared pastes were taken from each bath in order
to examine their porosity. The hydration was interrupted by crushing and washing them
with acetone, diethylether and vacuum drying at room temperature for 24 hours in a free
CO2 atmosphere.
Where:
r=pore radius (E); =surface tension of mercury (=484 dynes/cm); =mercury contact
angle (=141) and p=absolute pressure exerted (kg/cm2).
Our interest is focused to the pores of radii up to 2500 . It was found that, for pore
radii greater than 2500 the porosity measurements were remarkably affected on the
particles size of the sample.
Figure 1 and Figure 2 present the average pore volume distribution for the cements and
curing baths used at the predetermined ages of hydration. In these figures four
characteristic intervals are observed, referred to the ranges 45100 , 100550 , 550
1100 and 11002500 , where the variance dVp/Vp is remarkably varied.
It is clearly observed also that the time of hydration influences significally the
structure of porosity in all the cement pastes and curing baths used in the studied range of
pore radii (452500 ).
From these figures, it is observed that the time of hydration of the cement pastes
affects on the structure of porosity in the range of pore radii 45 and 2500 . The
relationship between time of hydration and pore volume is shown in Fig. 3, Fig. 4, Fig. 5
and Fig. 6. The strength of the relationship found between time of hydration and pore
volume has been evaluated statistically by means of regression analysis. These
relationships can be expressed by the following equations:
a. Water Bath
Vpe=k/(t+m) (Figure 3Curves a and b)
the coefficients e, k and m range among the following values (curves a and b respectively):
e=3.934.65, k=3.181057.82x105 and m=1.160.96
b. Sea-water Bath
Vpe=k/(t+m) (Figure 4Curves a and b)
the coefficients e, k and m range among the following values (curves a and b respectively):
e=3.384.07, k=37.1710522.96x105 and m=0.220.39
c. Mg Bath
Vpe=k/(t+m) (Figure 5Curves a and b)
the coefficients e, k and m range among the following values (curves a and b respectively):
e=4.493.90, k=1.2910582.9105 and m=1.060.22
d.Cl Bath
Vpe=k/(t+m) (Figure 6Curves a and b)
the coefficients e, k and m range among the following values (curves a and b respectively):
e=3.193.77, k=55.410574.1105 and m=1.300.05
Where,
t is the time of hydration, [days]; Vp is the pore volume, [cm3/g]
Figure 6. Cl bath
CONCLUSIONS
Evaluation of results presented in this paper leads to the conclusion that pore structure,
especially referring to the range of pore radii between 45 and 2500 A, is a function of the
hydration time. Furthermore, relationships based on statistical analysis of the test results,
are given for the estimation of this relation between time of hydration and pore volume
distribution.
Appropriate concrete technology 362
REFERENCES
INTRODUCTION
The technology of building with earth was developed by trial and error from the earliest
beginnings of humankind. Specific methods vary with geographic areas and cultural
tradition, but are basically the same. Most of these methods make use of local soil, are
labor intensive, and are logical and effective [2].
Unfortunately, adobe structures are severely affected by nature due to deficiency of
earthen fabric. There are many individual factors that contribute to the decay of earthen
Appropriate concrete technology 364
architecture, but by far the most destructive is slow decomposition over time due to
weathering.
The use of natural unstabilized adobe plasters to preserve earthen architecture against
destructive action of water has produced unsuccessful results. Also the traditional mortars
incorporating binders such as cement, lime and gypsum have different physical and
mechanical properties than earth and are not harmonius with the original structure. Also
many of the chemical amendment methods had limited successes with various
disadvantages [3].
The objective of this research was to develop durable compositions incorporating
pozzolanic materials for the protection of earthen material against rain. Recent research
has been inspired by a historical binder used before invention of cement, throughout the
centuries in Anatolia by different civilizations. This water resistant binder was named
Horasan and composed of volcanic or wood ash, fired earth powder, lime and water. In
some references egg-white is also included in this composition. However, there is no
evidence concerning the exact ratios of the ingredients.
Materials
In this study, pozzolanic materials such as fly ash and brick powder were incorporated to
the local earth to improve the durability of plain soil. Hydrated lime was also added to the
mixtures to activate pozzolanic reactions.
The pieces of irregular or damaged fired bricks were collected, crushed and reduced to
fine powder size. Hydrated lime procured in commercial paper sacks.
The other least known ingredient, fly ash, is a by-product of the coal combustion
process at power plants. To prevent air pollution caused by thermal plants,
electrostatically precipitated fly ash accumulates daily throughout the world in enormous
quantities. It is the first rating waste in size, for most of the industrial countries. This
quantity of waste causes serious environmental, technical and economic problems that
need to be solved. Fortunately, fly ash can be effectively utilized since it will combine
with lime to produce a cementitous material in presence of water [1].
In the present study, fly ash samples from Soma B thermal plant were used in the
experiments. This plant consumes 1.75 million tons of low calorie lignite (2400 kcal/g)
every year, with a 41 % C-type ash output. The chemical and physical properties of the
ash are shown in Table 1.
Local soil is used in the fabrication of adobe bricks, adobe walls and in the plaster
compositions.
EXPERIMENTAL DETAILS
Various amounts of fly ash and lime were added to brick powder start pozzolanic
reactions and to increase the mechanical, physical and workability proporties of the
compositions.
Table 1. Chemical Analysis & Physical properties
of Soma B fly ash
Improving the properties of adobe by pozzolanic materials 365
Dry and wet mixing procedures of the mortars were accomplished by means of a
homogenizer and a Hobart mixer to ensure uniformity. Fabrication of the specimens was
accomplished by table vibration in steel forms. The specimens were then stored under
laboratory conditions until testing. Standard 50 mm cube specimens were tested to
determine their 7th and 28th days compressive strength values. The tests were repeated
on duplicate specimens soaked in water for four hours.
Plaster mortars and their respective mechanical properties are shown in Table 2. As
can be observed from Table 2, the mechanical properties of the duplicate specimens did
not decrease meaningfully ever after four hours of soaking in water. The compressive
strength values are more than required values for the job. (The generally required
compressive strength for adobe bricks is about 1 N/mm2.)
The second phase of the research project, was developing pozzolanic plasters for
adobe walls.
Table 2. Compressive Strengths of Various
Compositions
Mix Soil Lime Fly Ash DRY COMP. STR. WET COMP. STR.
No (%) (%) (%) (N/mm2) (N/mm2)
7 DAYS 28 DAYS 7 DAYS 28 DAYS
1 3 12 4.2 8.3 3.9 8.2
2 6 9 3.7 7.7 3.5 7.5
3 7.5 7.5 3.6 7.2 3.4 7.3
4 9 6 2.9 5.8 2.5 5.4
Appropriate concrete technology 366
A composition of 3% lime, 12 % fly ash and 85 % brick powder was taken as a base
mixture for plastering amendments. By adding various percentages of soil to this
reference mixture, mortar specimens were prepared and tested with the same procedure
described above.
Plaster mortars and their respective mechanical properties are also shown in Table 2.
The compositions mentioned above were implemented in the plastering of small scale
walls were subjected to simulated rain to study their resistance to erosion.
The test walls were subjected to 2 hours of continous simulated rain during five day
time periods. An adobe wall specimen plastered with plain soil was also tested for
comparison. Two specimens were tested for every pozzolanic mixture.
After completion of simulated rain cycles, damage suffered by the specimens was
visually rated as light, moderate or severe. The test results are shown in Table 3.
As can be observed from Table 3, pozzolanic plasters were much durable than plain
soil plasters. For example, Mixture No. 1 visually showed almost no damage during
exposure to rain simulation of 10 cycles, where as plain soil plaster almost disintegrated
after 5 cycles.
Table 3. Results of Simulated Rain on Test Walls
Plaster No. Number of Cycles Damage
P (Plain Soil) 5 Severe
1 10 No damage
6 10 Light
7 10 Moderate
8 10 Moderate
From the preceeding test and information gathered, the following general observations
and conclusions can be stated:
a The generally required compressive strength for adobe bricks is about 1 N/mm2. Even
the lowest compressive strength of the test series is higher (1.4 N/mm2) than this
value.
b The main hazardous effect on adobe bricks or earthen architecture is water. This
deficiency may be overcomed by the use of pozzolanic mixtures in brick
Improving the properties of adobe by pozzolanic materials 367
manufacturing or in plastering jobs. The tests performed on small scale specimens and
rain simulation tests proved the durability of the compositions against water effect.
c The color of the mixtures can be arranged by incorporating local soil to the mixtures. It
should not be forgotten that, the high percentages of soil have negative effects on
durability and mechanical properties of plasters.
REFERENCES
1. BARADAN, B. 1990. A new restoration material for adobe structures. 6th International
Conference on the Conservation of Earthen Architecture, New Mexico, Preprints pp. 149152.
2. P.G.MC. ENRY, Jr. 1986. The role of earth block manufacturing. Middle East and Regional
Conference on Earthen and Low-Strength Masonry Buildings in Seismic Areas Conference
Proceedings, Ankara, pp. 257263.
3. TAYLOR, R.M. 1990. An Evaluation of the New Mexico State Monuments Adobe Test Walls at
Fort Selden. 6th International Conference on the Conservation of Earthen Architecture, New
Mexico, Preprints pp. 383389.
Theme 5
NON-FERRROUS
REINFORCEMENT
Chairmen Mr W E Brewer
Brewer & Associates
USA
Dr S B Desai
Department of the Environment
United Kingdom
Professor J Morris
University of Witwatersrand
South Africa
Leader Paper
INTRODUCTION
Since 1970s s extensive research has been done on fibre reinforced concretes (FRC). A
lot of promising innovations have been achieved in laboratories and in the first
applications. However, the practical applications have been quite limited. The practical
use has remained so limited partly because of the uncompetitive price of the FRC
materials and structures, and partly because of the unsatisfactory properties of FRC, e.g.
regarding its durability. However, we can be satisfied that there is great potential for
successful applications of FRC in many different areas. Besides the further development
of the material, a casewise systematic performance-based analysis of the requirements
and the optimised structural innovation interactively with the specification of material
parameters and with the choice of the FRC material can be the key to wider and more
useful practical applications of FRC.
Designing the structures and their materials according to the performance
requirements of different applications can be called the tayloring of the material.
Concrete as such already possesses great versatility and potential for modification
through the mix design and through the structural design and detailing. When adding
fibre reinforcement, the alternative modification possibilities become vast. The challenge
for both the materials engineers and for the structural engineers is the optimal utilisation
of these vast alternatives. In research, the different requirements must be clearly realised
and classified. After that the researches must produce for the designers calculation
models in order to enable the controlled and optimal tayloring of the structural properties
during the design stage.
The basic requirements of the material properties in structures vary greatly. In large
civil engineering structures like bridges and dams strength and durability are of the
highest priority. Buildings typically have quite different requirements in different parts of
the building. Strength, fire safety and durability are the dominating requirements in the
building frames in order to guarantee the long service life and safety of the building.
Visual appearance, thermal and moisture performance, durability, toughness and
moderate strength are the main categories of requirements for facades. The infill
structures must possess quite different properties; they could be called the soft parts of
the building. Controlled sound insulation, healthiness, fire safety, moderate strength and
easy fastening of furniture and other equipment are typical performance criteria for the
infill structures.
For economical production and for flexibility in alterations during use, several parts of
the building like the infill must have good workability properties during and after
construction. The good workability includes the workability of the concrete mass during
mixing, transport, handling, casting and compacting, but also easy workability of the
hardened structures through sawing, drilling, nailing, screwing and bolting. The
workability properties are especially important for the materials and structures which are
aimed at renovation purposes.
In order to be able to meet the challenge of a controlled and optimised structural
design for performance requirements, the designer must have the methodological
capability and a deep knowledge of the properties of different FRC materials. Also the
researchers can benefit from the performance- based systematics in innovative and
effective tailoring of the materials and in the innovative developments of the structures
Tailoring the properties of concrete structures with appropriate non-ferrous reinforcements 371
and their design methods. Besides the traditional mechanical, physical and chemical
aspects, the health-related properties and environmental impacts must also be taken into
account.
Fibre reinforcements
Fibre reinforcement can be introduced into concrete in different forms, as shown in
Figure 1 /1/. Thus, the FRC composite material can vary owing the reinforcement from
totally dispersed fibres to very sophisticated layered, directed or shaped fibres.
Additionally, different bundled and woven fibres are used.
Applying the fibre reinforced concrete, different structural compositions can again be
manufactured, as shown in Figure 2 /1/.
Appropriate concrete technology 372
Synthetic fibres mean all types of fibres which are produced in industrial chemical
processes. Most often these materials are different types of polymers, carbon or glass.
Amorfous steel is also presented here, although it belongs to ferrous materials. A list of
common commercially available synthetic fibres and their properties are presented in
Table 1 /3/. The properties presented in Table 1 are only estimates, because the properties
may vary somewhat depending on the manufacturer.
Tailoring the properties of concrete structures with appropriate non-ferrous reinforcements 373
Some properties of natural fibres are presented for wood fibres in Table 2 /4/ and for the
fibres from other plants in Table 3 /2/.
Table 2. Properties of wood fibres.
Fibre Density Diameter Length Tensile Modulus of
(g/cm3) (m) (mm) strength elasticity (GPa)
(MPa)
Cellulose
Wood, straw (dry) 1.5 1540 0.54 300500 810
Appropriate concrete technology 374
The leading principle for the modelling of FRC is the general law of mixtures /3/
(1)
(2)
(3)
(4)
(5)
The mechanical meaning of eq. 5 is that the fibres can take all the force which is
moving from the matrix after the tensile rupture of the matrix. When designing the
composite material using this equation it is important to use the upper limit value of fctu
and the lower limit value of ffu, because the design otherwise leads to an unsafe condition
and to the progressive rupture of the material at the cracks.
The critical length of the fibre, when the slipping begins, is
where df is the diameter of the fibres or corresponding factor (=4(fibre cross sectional
area/ fibre cross sectional perimeter)
f bond strength of the fibre in the matrix
After cracking, the tensile strength of the composite material is dictated by the fibres
and can be calculated with the equation
(6)
There exist different theoretical and experimental values for the efficiency factors of
aligned fibres, 2-dimensional fibres and 3-dimensional fibres /3/,/5/,/6/.
Especially for individual types of composites, but also for some general uses, there
exist also experimentally determined calculation models of the performance properties.
These models often contain also physical, chemical and sometimes even biological
properties /5/,/7/,/11/.
With the aid of the basic equations, different elastic and plastic calculation models can
be used for the calculation of the bending and shear capacities of the FRC structures.
The characterisation of the FRC materials includes the properties of the ingredients and
the properties of the interfaces /8/. The interface phenomena occur both at the interfaces
between the cement paste and aggregate and between the concrete matrix and fibres. In
the reserach and development of FRC materials the interface phenomena are the real key
factors. A range of modern equipment and methods are now available for such interface
research. There are many standardised testing methods for FRC materials. The
application of multiple testing methods especially regarding durability properties is
essential for reliable results /10/. Besides these special methods and calculation models
also the methods and models of ordinary concrete can be used. A selection of calculation
models on durability performance is presented in the design guide for concrete structures
/11/.
A central characteristic factor is the toughness index, e. g. regarding the ASTM
methods C 101892 and C 45790 /12/. It is important to realise that the toughness often
decreases remarkably with time because of the corrosion of fibres and because of the
Tailoring the properties of concrete structures with appropriate non-ferrous reinforcements 379
increase of tensile strength of matrix concrete and the strengthening of the bond between
the concrete and fibres /12/.
Integrating the application of the performance properties is made during the structural
design stage.
The controlled service life and durability of the structures is of great importance when
applying complicated FRC materials. For this purpose the corresponding methodology
generally used for concrete structures can be applied /11/. The flow chart of the service
life design is presented in Figure 7 /11/.
The integrated design includes multiple criteria decision-making, where the properties
with different measures must be compared. This leads to quite wide discussions on the
basic values when deciding the criteria for multiple optimisation.
Appropriate concrete technology 380
Design procedure
The design of FRC structures follows the common design procedure, but some additional
procedures are added in order to take account of the high number of alternative materials
and their combinations in mix design. Special attention is payed also to the service life
and to the cost effective optimisation of the material and structural properties for the
entire service life of structures.
The design procedure includes the following phases, from which in each case the
actual phases are chosen for the active use:
1. Identification of the mechanical, physical, chemical and biological loadings affecting
the actual structures in their working environment during the service life.
2. Specification of the actual performance requirements and the limits and characteric
values of environmental impact for the structures.
3. Selection of alternative constituents of the FRC, which possess the basic properties
fulfilling the specified performance requirements.
4. Identification of the intended manufacturing processes (factory/site manufacture) and
the detailed manufacturing methods.
5. Sketching alternative structural compositions and forms using the selected materials,
with the objective of meeting the specified performance properties, environmental
impact limits and manufacturing requirements.
6. Specification of alternative mix designs for the specified performance requirements
utilising selected calculation models of material properties corresponding to the
identified loadings.
7. Application of the mix designs the sketched structural compositions, thus designing
alternative structural solutions.
8. Comparison of the alternative structures regarding the specified performance
requirements and taking into account the environmental impact of the alternative
structures during the entire service life.
9. Selection of the structural solution and the materials for the final design.
10. Final design and detailing of the structure.
Durability design
The design procedure at the durability design stage follows the following phases /11/:
1. Specification of the target service life and the design service life
2. Analysis of environmental effects the structure
3. Identification of durability factors and degradation mechanisms
4. Selection of a durability calculation model for each degradation mechanism
5. Calculation of durability parameters using available calculation models
6. Possible updating of the calculations of the ordinary mechanical design
7. Transfer of the durability parameters the final design
The entire procedure is quite complicated. For that reason it can be simplified in practical
cases using the expertise and earlier experiences of the designer. The methodology can be
followed at the product development stage of new prefabricated units or of new building
types for specific uses.
Tailoring the properties of concrete structures with appropriate non-ferrous reinforcements 381
Design methods
At the mechanical design of non-ferrous reinforcements there are practical design rules
and calculation methods for fibre reinforced concretes /5/,/6/,/7/,/17/,/18/,/19/, for non-
metallic tendons /21/,/22/,/23/,/24/,/25/, for glued surface reinforcements /26/ and for
fabric reinforcements /26/,/27/,/28/.
In durability design the general methodology and methods for the design of concrete
structures can be applied /9/,/10/,/11/. The detailed degradation models needed for
specific types of concrete structures with non-ferrous reinforcements are numerous /5/,
/8/,/29/,/30/,/31/,/32/,/33/,/34/,/35/,/36/. This area is certainly the weakest with respect to
existing quantitative calculation models. Therefore qualitative and approximative
knowledge and testing of products continue to be used in most cases.
The analysis and optimisation of the environmental impact of structures can be
calculated by applying the general methods and information available on the
environmental profiles of submaterials /37/.
CONCLUDING REMARKS
Fibre reinforced concrete technology possesses an extremely high potential for advanced
applications in both developed parts and developing parts of the world. In old
industrialised countries the manufacturing can be automated in factories and mechanised
on site works. In developing countries the manufacture often includes more manual
works, which reduces the costs in those conditions.
There is a vast range of different fibre materials, either synthetic or natural materials.
Local materials are often the cheepest and therefore preferred especially in the
developing countries. The current trend towards ecological building also tends to favour
the use of local natural materials even in developed counties. One promising possibility is
Appropriate concrete technology 382
also the application of hybride fibres and the combination of soft aggregate material
together with fibres.
There has been and still are some problems regarding the long-term durability of non-
ferrous FRC materials and products. The long-term durability of synthetic polymeric
fibres is not always known for several decades, which is usually the service life
requirement. The combined action of several severe loadings like temperature and
moisture changes combined with weathering, frost attack, chemical weakening and
mechanical loadings has caused the progressive growth of the cracks, leading to serious
damages e.g. to facades. Traditionally, quite small fibre contents have been used in order
to improve the economy of FRC. This is very dangerous in severe conditions, because the
small fibre amount can not prevent the long-term progressive damage at the cracks. In
order to avoid these kinds of difficulties it is important to apply efficient multiple testing
procedures supported by theoretical calculations, comparisons and optimisations.
Additionally, new materials and structural innovations and inventions are still needed in
order to overcome the existing problems.
In the case of more massive structures like beams, columns and slabs the fibre
reinforcement can be combined with bar reinforcement. In this case the role of the fibres
is to take the secondary stresses, while the bar reinforcement takes the highest primary
stresses and forces. The fibres can e.g. replace the stirrups, thus enabling the automated
manufacture of the structures. Often the fibres can be concentrated in to the most stressed
parts of the structures, thus helping to diminish the amount of fibres and the costs.
The nearly unlimited variations of FRC materials make it possible to tailor the
materials and structural design for each purpose in order to meet the multiple
requirements of the users and of ecological and economical aspects. This necessitates
innovative product and production development as well as a design procedure in practical
applications.
REFERENCES
31. KOSA, K., NAAMAN, A.E. & HANSEN, W. Durability of fiber reinforced mortar. ACI
Materials Journal 88(1991)3, pp. 310319.
32. MAJUMDAR, A.J. & LAWS, V. Class fibre reinforced cement. Oxford 1991. BSP
Professional Books. 197 p.
33. MAJUMDAR, A.J. & WALTON, P.L. Durability of fibre cement composites. 2nd Int.
CANMET/ACI Conf. Durability of Concrete. Montreal, 49 August 1991. Detroit 1991,
American Concrete Institute, ACI SP-126. Vol. II. Pp. 754771.
34. Measurement of properties of fiber reinforced concrete. Reported by ACI Committee 544. ACI
Materials Journal 85(1988)6, p. 583593.
35. MORGAN, D.R. Freeze thaw durability of steel and polypropylene reinforced shotcretes: a
review. 2nd Int. CANMET/ACI Conf. Durability of Concrete. Montreal, 49 August 1991.
Detroit 1991, American Concrete Institute, ACI SP-126. Vol. II. Pp. 901918.
36. PIRIE, B.J., GLASSER, F.P., SCHMITT-HENCO, C. & AKERS, S.A.S. Durability studies and
characterization of the matrix and fibre-cement interface of asbestos-free fibre-cement products.
Cement & Concrete Composites 12(1990)4, p. 233244.
37. RILEM Workshop: Environmental aspects of building materials and structures. Manuscript of
Symposium Report. To be published in 1995. Technical Research Centre of Finland, VTT
Building Technology.
DEVELOPMENT OF COVER METERS
FOR STAINLESS REINFORCEMENT
TWO SUCCESSFUL APPROACHES
J C Alldred
Protovale Oxford Ltd
UK
INTRODUCTION
APPROACH A (MODEST)
If a conducting target is present within the field of influence of the search coil, the
switch-off induces eddy currents to flow in the target, which initially are sufficient to
regenerate the collapsed magnetic field. As there is no source of energy to maintain these
eddy currents, they will decay away with a time-constant dependant upon the dimensions
of the target (radius of the eddy current path) and the conductivity of the metal. These
exponentially-decaying eddy currents generate a similarly-decaying secondary magnetic
field which propagates back to the search coils and induces a small voltage which also
decays exponentially. Example coil voltage waveforms are shown in Figure 1. The
receiver circuitry of the detector must ignore the first few microseconds of coil voltage,
which contain the large back-emf transient and also a rapidly-decaying signal component
from ionic currents owing to any moisture in the medium in which the target is
embedded.
For the detection of both mild and high-tensile steel, this delay time is typically in
the region of 28s; but the signals from stainless steel only persist for a time somewhat
less than this (apart from very large diameter bars), so the instrument gives little or no
response to small and medium sized stainless rebars. By reducing the delay time to
between 16 and 18s, an adequate signal can be received from most types of stainless
steel wall-ties [3] and all sizes of stainless steel reinforcement. In practice, various other
pulse widths and durations also need to be changed, so all pulses are derived from a
single master clock which can be switched between two frequencies.
the signals from smaller bars and attenuating the signals from larger bars, so that the
output signal to display is a function of cover but independent of bar size. This can be
readily implemented by switching the value of just one feedback resistor in the receiver
amplifier. However, to calibrate for eight different bar sizes (6, 8, 10, 12, 16, 20, 25 and
32 mm) for both high-tensile and stainless-steel bars, a total of sixteen such resistors are
required. To remove the need for complex front-panel switches, and to avoid the need to
route critical signals between the circuit-board and front panel, the resistor selection is
actually performed by two 8-way solid-state multiplexers on the circuit board; a 3-bit bus
selects bar diameter, and a fourth wire selects multiplexer and clock frequency and hence
steel type.
A photograph of the completed instrument, provisionally named SuperStab, is
shown in Figure 2.
APPROACH B (THOROUGH)
Although the simpler model just described proved to be eminently suitable for locating
and measuring stainless steel rebars, and is believed to be the first and only instrument in
the world able to do so, it was decided to treat it as merely a test-bed for researching the
necessary technology, and to develop a full-specification stainless-steel cover meter to
achieve an even more exacting target. This new development can be thought of as an
enhancement of the recently-introduced Protovale CM9 CoverMaster [6]; but in truth, the
CM9 is an intermediate stage (for high-tensile steel only) in the development of the
Asdic cover meter for both high-tensile and stainless steels.
If two measurements of v are taken at times t and t+t, the time-constant can be derived
from the ratio of the two voltages
=t/[loge (vt/vt+t)]
The time-constant is known to increase monotonically with bar diameter, but an exact
relationship can not be derived and can only be measured experimentally. In practice, it is
not necessary to solve for , merely to tabulate vt+t/vt versus diameter. In the solution
adopted, both the initial delay and the incremental delay t were each nominally 10s.
The method of sampling the waveform is depicted in Figure 3. The ratio of the signals
after the two delays was calculated by an 8-bit microprocessor, and the appropriate
diameter found from a look-up table.
Furthermore, whereas the signals received from high-tensile bars are strongly
dependent upon the angular orientation of the search probe with respect to the bar, the
orientation effect from stainless-steel bars is much less marked in amplitude and even
less noticeable in decay time; as a result, transverse bars of the same size also produce
negligible errors in diameter-determination.
If the interfering nearby bars are of a different diameter to the bar being measured, the
deduced diameter will lie somewhere between the diameters of the measured and
interfering bars, weighted towards whichever bar is yielding the strongest signal. This
will mean that if, for example, the interfering bar is two sizes larger (or smaller) than the
bar being measured, and is contributing a signal whose amplitude is less than one-quarter
of that of the bar being measured, the deduced bar bar diameter will be in error by less
than half a bar size; this is sufficient to invalidate the estimation of steel grade, but the
deduced diameter will nevertheless be correct.
A further practical difficulty when attempting to deduce the diameter of high-tensile
bars by measurement of signal amplitudes, is the need to position the search probe
accurately over and aligned with the bar under test; because this new decay-time analysis
does not rely on absolute amplitudes, diameter measurements remain accurate even when
made with the head displaced somewhat from the line of the bar, either laterally or
angularly.
As far as the author is aware, this new bar-sizing method is the only one to exhibit
such a tolerance of positioning errors and such a high degree of immunity to the
proximity of nearby bars of similar size.
steel, the master clock is 96kHz and the main and auxiliary sample delays are nominally
10 and 20s respectively. The outputs of the two integrators are continuously-present dc
levels; the analogue-to-digital converter digitises these alternately at a rate which is
independent of the above timings. In this way, there is no duplication of signal-
processing circuitry; only one additional integrated circuit package is required to
demultiplex the different sampling pulses.
Epoxy-Coated Steel
The epoxy coating should have no influence on the measurement whatsoever; therefore
follow the guidelines for high-tensile steel in the section above.
Stainless Steel
For location and cover measurement and diameter determination, it will be necessary to
use Solution B as it is the only method known to fulfil all requirements.
For location and cover measurement of bars of known diameter, Solution A is also
suitable.
If either of these instruments is set to high-tensile mode, they will (possibly grossly)
over-estimate cover, and may fail to even locate the smallest sizes of stainless bar. For
Appropriate concrete technology 394
location only, the Imp Stainless-steel Wall-tie Locator is also suitable. Instruments that
have not been designed specifically for stainless steel will not be suitable.
REFERENCES
INTRODUCTION
Fusion bonded epoxy coated reinforcement (FBECR) was developed in the USA in the
1960s in response to widespread problems of reinforcement corrosion in concrete bridge
decks. The epoxy coating is impermeable and an electrical insulator, and resists corrosion
of reinforcement by providing a physical barrier to the oxygen and moisture necessary for
the corrosion reaction, and by interrupting electrical continuity of the corrosion cell. The
USA was the first country to conduct a major study into the effectiveness of FBECR and
the first to introduce a national standard for the material1. By 1987, epoxy-coated bars
accounted for around 5% of total reinforcement consumption in the USA. In the UK, a
Appropriate concrete technology 396
dedicated plant for production of FBECR was set up in 1987. FBECR is now also
produced in several other European countries, in Japan and in the Middle East. However,
although the product has now been available for some time, it has yet to be widely
assimilated into Codes of Practice for design.
The surface texture of an epoxy coating is much smoother than the normal millscale
surface of a hot rolled, or black reinforcing bar, and affects the interaction, or bond,
between reinforcement and concrete. For design purposes, bond is considered to be a
factor in both serviceability and ultimate strength. At the serviceability limit state, crack
widths and deflection are influenced. At the ultimate limit state, strength of laps and
anchorages depends on bond. In addition, bond characteristics of flexural reinforcement
will influence rotation capacity of plastic hinges. Extreme loss of bond is also known to
influence flexural and shear capacity of beams. However, changes of this degree lie
beyond the change in bond that will arise from epoxy coating, and is not considered
further here.
Aim
This paper reviews data on structural performance of epoxy coated reinforcement, with
the aim of deriving recommendations for design practice to enable structures reinforced
with FBECR to achieve equivalent performance to that of structures reinforced with
millscale surface ribbed bars.
Figure 14. The dependence of bond reduction on slip suggests that the search for a unique
bond reduction factor for coated bar will be fruitless, and that
Confinement
Bond action of ribbed bars generates bursting stresses that tend to split the surrounding
concrete. It is established that bond strength increases with confinement to the bars, either
from increased cover, transverse reinforcement, or lateral pressure. Depending on the
Code of Practice in use, some or all of these factors may be taken into account in
detailing of laps and anchorages. Choi et al demonstrated that the absolute increase in
bond strength with increasing cover is similar for both coated and black bars, and hence
that the difference, when expressed as a ratio, reduces with increasing cover5. Treece and
Jirsa arrived at a similar conclusion from tests on lapped joints6.
Casting position
Compaction of concrete tends to be poorer close to the top of a pour, and settlement of
fluid concrete may lead to formation of voids below bars cast near the top of a pour.
Bond strength is dependent on the quality and compaction of the surrounding concrete,
Appropriate concrete technology 398
and bars cast in locations near the top of a pour tend to be weaker in bond than those cast
near the bottom. In an extensive study employing beam end test specimens, Hadje-
Ghaffri et al examined the interaction of slump, casting position and compaction7. They
found bond strength of coated bars to be affected less by poor casting position than
black bars, and suggested a reduced top bar factor. Tests conducted at BCA8 also showed
that apoor casting position affected epoxy coated bars less than black bars.
Rib Geometry
In tests on specimens representing lapped joints Cairns and Abdullah showed that the
strongest epoxy coated bars could develop nearly the same bond strength as the weakest
black bars9, and that the weaker rib patterns, with lower relative rib areas, generally
exhibited smaller bond strength reductions. Choi et al5 also found a strong dependence of
bond strength reduction on rib pattern, but reported that bars with a higher relative rib
area were less affected by coating. The difference may be associated with the provision
of confining reinforcement, or with a correlation between relative rib area and bar
diameter in Chois study. The opposing conclusions of these two studies have still to be
accounted for.
Fire resistance
Fusion bonded epoxies are thermosetting polymers that cannot be turned to a molten state
by application of heat, although they may pyrolyse at higher temperatures. The material
does soften around the glass transition temperature of approximately 110C, however.
Pullout tests conducted on coated and black reinforcement cast into a reinforced concrete
slab14 under fire conditions showed that all test bars were able to attain their yield
strength. Although it is evident that the coating material softened at temperatures above
110C and that bond behaviour was affected, it should not be concluded that fire
resistance is lessened by use of epoxy coated bars. It is also necessary to bear in mind
Structural design using epoxy coated reinforcement 399
reservations concerning the relevance of pullout type tests to practical situations set out
above.
Summary
It is generally agreed that bond strength and stiffness of coated bars is lower than that of
black bars. The strength reduction is generally estimated at between 0% and 25%,
although reductions of around 40% have been reported. An assessment of bond based on
performance of straight bars at ambient temperatures under monotonic short term loading
under conditions of minimum confinement consistent with Code provisions should give
results that will be conservative for other circumstances.
Coated bars have now been produced for around 30 years, but it is only relatively
recently that Codes of Practice have introduced guidance on design and detailing of
concrete structures reinforced with epoxy coated bars. Reduction factors for bond
strength at anchorages and lapped joints are now included in a few national Codes of
Practice15,16, although the accuracy of the factors introduced is debated. At the
serviceability limit state, an epoxy coating is either taken to have no effect on
serviceability behaviour17, or any effect is ignored. Other Standards documents, such as
EC218 and the CEB-FIP Model Code19, recognise the influence of bond on serviceability
performance, but do not provide specific values for use when detailing with FBECR.
Many studies on bond of FBECR have been concerned primarily with derivation of a
numerical factor to quantify the bond strength reduction, crack width increase, or other
change in structural performance attributable to coating. Such factors can be obtained
through direct comparison of measurements of performance of nominally identical
specimens reinforced with black bar and with FBECR. This approach has two major
disadvantages:
a) the relationship between structural performance of FBECR and black bars is dependent
on the design of the test specimen. Extrapolation to other circumstances may be
misleading.
b) the conditions under which FBECR and black bars are used in practice may differ, and
could offset or accentuate differences attributable to bond.
It is thus necessary to examine the theoretical and practical background to the simplified
rules presented in Design Standards and Codes to derive recommendations for design.
Appropriate concrete technology 400
For practical design, crack widths and deflections are controlled through simple deemed
to satisfy rules, and not by direct calculation. This is at least in part because
serviceability is intrinsically less critical than ultimate the ultimate limit states, and the
major calculation effort required for a detailed estimate is rarely justified. Crack widths
are controlled by limiting transverse spacing between longitudinal bars, and deflections
controlled by limiting the span/depth ratio. Bond characteristics do not feature in such
rules. These deemed to satisfy rules are not considered in this paper, which instead
concentrates on adaptation of the rules provided for more detailed assessment of
performance at the serviceability limit state. However, the approach evolved could
subsequently be used in parametric studies for the development of simplified rules.
In a cracked cross section tension forces are carried by the steel only. However,
concrete does possess some tensile strength, and between transverse cracks may carry
part of the tension force. Bond characteristics influence the rate at which load is
transferred between bar and concrete, and influences structural performance in two ways:
a) Between transverse cracks, reinforcement strains are reduced below the value that
would be calculated if the section were fully cracked, Figure 2. The reduction in
average bar strain is termed tension stiffening, and is dependant on bond.
b) The distance between successive transverse cracks is influenced by the rate at which
tension in the longitudinal bars is transferred to the surrounding concrete. The stiffer
the load/slip relationship, the more rapid the transfer of force, and the shorter the
distance required to increase tensile stress in the concrete to the level at which a crack
will form, Figure 2.
Deflection
Deflection of a beam may be expressed by Equation 1, which shows that a reduction in
tension stiffening will result in increased deflection 6. The magnitude of ts depends not
only on bond characteristics, but also on the proportion of reinforcement in the section,
the cross sectional shape, the load level at which the comparison is made, the grade of
concrete used, and on whether short or long term behaviour is considered. The
relationship between deflections of beams reinforced with coated and black bars therefore
cannot be expressed by a unique coefficient.
Equation
1
Several investigators report little change in deflection where coated bars are used.
Closer inspection reveals that most of the specimens were relatively highly reinforced,
and would be insensitive to loss of tension stiffening. The presence of a lapped joint may
also have masked loss of tension stiffening in a number of studies. In a series of tests on
seven pairs of nominally identical beams of various cross sections, Cairns measured an
average increase in deflection of around 20% where coated bars were used. Deflections
for individual pairs differed by between 5% and 35%20. Taken as an average, the
increased deflections corresponded to a reduction in bond of around 50% for coated bars.
The
Crack Widths
For practical purposes, mean crack width wm may be estimated as mean strain in
reinforcement times the distance Srm between cracks, Equation 2. A variety of of
expressions have been proposed for estimating crack spacing and are generally similar in
form, Equation 3. Coefficient k1 takes account of bond properties, and is greater for bars
with less stiff load-slip behaviour. The increase in k1 is widely assumed to be inversely
Appropriate concrete technology 402
related to bond. A change in bond characteristics may thus have a twofold influence on
crack widths, as crack spacing will be greater and tension stiffening less when bond
stiffness is reduced.
Equation
2
Equation
3
noted on the graph. Predictions are generally in good agreement with observations.
However, some tuning of the approach may be required for inclusion in existing Codes.
For example, EC2 modifies Equation 3 by substituting 50mm in place of the variable
cover, and a better fit to test data is obtained if k1 is increased by 40% instead of 100%.
Codes of Practice make the assumption that bond stress is uniform throughout the length
of a lapped joint. This leads to the erroneous (but convenient) assumption that lap
strength is directly proportional to lap length. Statistical analyses of test data for black
bars provide ample evidence that this is not so, and that the increase in strength is less
than proportional to the increase in lap length. Bond tests are generally carried out with
lap lengths short enough to ensure the occurrence of bond failure before lapped bars start
to yield. However, Code values for design bond strength must be based on the lap length
required to develop characteristic strength of the reinforcement. Figure 4 illustrates the
variation in strength of lapped joints of black and epoxy coated bars measured in three
studies8,21,22. Slopes of paired plots are similar, but show that the reduction in average
bond strength with increasing lap length is less marked for epoxy coated bars. Results
from shorter laps will therefore overestimate how much full strength laps are weakened
by coating.
CONCLUSIONS
It has not been possible to fully explore all aspects of the bond performance of FBECR in
a paper of this nature. The following conclusions, of relevance to Codes of Practice, are
nonetheless warranted.
1 Bond stiffness and strength is reduced by an epoxy coating.
2 Crack control performance of FBECR is equivalent to that of plain round bars, and
beams reinforced with FBECR exhibit greater deflections than similar beams
reinforced with black bars. The difference in structural performance is dependant on
the properties of the member and on the loading as well as on bond.
Appropriate concrete technology 406
3 Ultimate strength of laps and anchorages is reduced by an epoxy coating. For practical
design purposes, this reduction is offset by an increase in bond strength where higher
cover is used for durability in adverse exposures. A comparison with provisions of
EC2 suggests that present rules for design of lapped joints of black bars provide an
adequate margin of safety against failure of lapped joints of coated bars. Validity of
this conclusion is at present restricted to bars of up to size 25 in beams or slabs
constructed of concretes not exceeding Grade 40, and detailed for chloride
contaminated exposures.
Further Work
1 There is no data on crack control properties of FBECR in members loaded in direct
tension.
2 The reduction in tension stiffening associated with FBECR has not been adequately
quantified.
3 It would be of value to measure strength of lapped joints and of anchorages of black
and coated bars under respective least favourable conditions of use to verify that
Code rules provide a consistent margin of safety.
4 More data is required on the effect of coating on plastic rotation capacity.
5 Further work to determine the influence of rib geometry on the reduction in bond
strength attributable to coating is desirable.
REFERENCES
1. ASTM A775/A 775M-86. Standard Specification for Epoxy Coated Reinforcing Steel Bars.
American Society for Testing & Materials. Philadelphia. 1986.
2. British Standard BS7295:1990. Fusion bonded epoxy coated carbon steel bars for the
reinforcement of concrete. British Standards Institution, London, 1990. 7pp.
3. Cairns, J., & Abdullah, R. An evaluation of bond pullout tests and their relevance to structural
performance. The Structural Engineer. Vol 73, No. 11. June 1995. pp179185.
4. Cairns, J., & Abdullah, R. Fundamental tests on the effect of an epoxy coating on bond strength.
American Concrete Institute Materials Journal. Vol. 91, No. 4. July/August 1994. pp 331338.
5. Choi, O.C. et al. Bond of epoxy coated reinforcement: rib parameters. Proceedings of the
American Concrete Institute, Materials Journal, Vol.88, No.2, Mar-Apr 1991. pp.207217.
6. Treece, R.A. & Jirsa, J.O. Bond strength of epoxy-coated reinforcing bars. PMFSEL Report No.
871, University of Texas, Austin, Jan 1987. 85pp
7. Hadje-Ghaffari H, Choi O C, Darwin D & McCabe, S L, Bond of epoxy coated reinforcement:
cover, casting position slump and consolidation. ACI Structural Journal, Vol 91, No. 1, Jan
1994. pp5968.
8. Clarke, J L. Tests on fusion-bonded epoxy-coated steel. BCA Project Report RN1.025.01.
Slough, UK. Nov 1990. 23pp.
9. Cairns, J., & Abdullah, R. Influence of rib geometry on strength of epoxy coated reinforcement
splices. American Concrete Institute Structural Journal. Vol 92, No. 1. Jan-Feb 1995. pp2327.
10. Clifton, J.R. & Mathey, R.G. Bond and Creep Characteristics of Coated Reinforcing Bars in
Concrete. American Concrete Institute Journal. Vol 80, No.4. July/August 1983. pp288293.
11. Johnston, D.W. & Zia, P. Bond fatigue of epoxy coated reinforcing bars. Materials &
Structures: Research & Testing. No. 97. Jan/Feb 1984. pp3034.
Structural design using epoxy coated reinforcement 407
12. Cleary D B and Ramirez, J A. Epoxy coated reinforcement under repeated loading. ACI
Structural Journal, Vol 90, No. 4, July-Aug 1993. pp451458.
13. Hamad, B S, Jirsa, J O & dAbreu d Paolo N. Effect of epoxy coating on bond and anchorage of
reinforcement in concrete structures. Report 11811F. University of Texas at Austin. Dec 1990.
14. Lin, T.D. et al. Pullout tests of epoxy coated bars at high temperatures. ACI Materials Journal.
Nov-Dec 1988. pp544550.
15. ACI Committee 318. Revisions to Building Code Requirements for Reinforced Concrete. Vol.
85, No. 6. Nov/Dec 1988. pp645674.
16. Japan Society of Civil Engineers. Recommendation for Design & Construction of Concrete
Structures Using Epoxy-Coated Reinforcing Steel Bars. JSCE. Tokyo. 1986
17. British Standard Code of Practice, BS8007:1987. Standard specification for design of concrete
structures for retaining aqueous liquids. British Standards Institution. London 1987.
18. Eurocode 2, Design of Concrete Structures DD ENV 19921-1:1992.British Standards
Institution, London.
19. Comite Euro-International du Beton & Federation Internationale de la Precontrainte. CEB-FIP
Model Code 1990 for Concrete Structures. CEB-FIP. Lausanne. 1992.
20. Cairns, J. Performance of epoxy coated reinforcement at the serviceability limit state.
Proceedings of the Institution of Civil Engineers, Structures & Buildings. Vol 104. Feb 1994.
pp6173.
21. Cairns, J., & Abdullah, R. Ultimate strength of lapped joints of epoxy coated reinforcement.
Proceedings of the Institution of Civil Engineers, Structures & Buildings. Vol 104. Nov. 1994.
pp391400.
22. Makhlouf H M, Al-Tamimi, A, Akbari, A H. Bond Carrying capacity of epoxy coated
reinforcement. Proc. Conf. On Corrosion and Corrosion Protection of Steel in Concrete.
Sheffield Academic Press. Sheffield. 1994. pp12181230.
23. Orangun C.O., Jirsa J.O., & Breen J.E. A re-evaluation of test data on development length and
splices. American Concrete Institute Journal. Vol. 74, No. 3. March 1977. pp1 14122.
24. Park, R. and Paulay, R. Reinforced Concrete Structures. Wiley. New York. 1975.
25. Abrishami, H H, Cook, W D, and Mitchell, D. Influence of epoxy coated reinforcement on
response of normal and high strength concrete beams. ACI Structural Journal, Mar/Apr 1995.
SUPERCOVER CONCRETE: A NEW
METHOD FOR PREVENTING
REINFORCEMENT CORROSION IN
CONCRETE STRUCTURES USING
GFRP REBARS
C Arya
G Pirathapan
South Bank University
UK
INTRODUCTION
Vast sums of money are spent annually in the U.K. and elsewhere on the repair and
maintenance of concrete structures [1]. A large proportion of this expenditure is to deal
with the deterioration caused by reinforcement corrosion. This corrosion is attributable to
carbonation of the concrete and/or the presence of chlorides in solution around the
reinforcement.
Many ways have been suggested for preventing this problem including treating the
surface of the concrete with silanes or siloxanes [2], replacing a proportion of the cement
with ground granulated blast furnace slag [3] or pulverised fuel ash [4], and using
stainless steel or coated reinforcement. A cheaper alternative is simply to increase the
depth of cover to the reinforcement.
Although easy to implement, deep covers will give rise to cracks with large widths at
the concrete surface. These may be undesirable for reasons of durability but are often
unacceptable because of appearance. However if rods made of non-metallic materials
such as glass fibre reinforced plastic (GFRP) were to be introduced in the cover zone, it
should be possible to control surface cracking to within acceptable limits. Using this
approach, the steel reinforcement would primarily provide the tensile strength and will
not corrode for the life of the structure due to the deep concrete cover. The non-metallic
reinforcement situated near to the concrete surface would primarily control surface crack
widths and will also not corrode.
If such a system could be developed, it would offer the following advantages:
(1) low cost and cheap to implement solution
(2) primarily relies on existing materials and technology
(3) applicable to a wide range of structures.
The purpose of this paper is to provide further details of this method, termed Supercover
Concrete, and to describe some of the laboratory investigations currently being
undertaken to assess its performance.
SUPERCOVER CONCRETE
The main feature of this method is the reinforced cover. Two questions that need to be
addressed are:
1) what thickness of concrete cover to the steel reinforcement should be provided and
2) is GFRP effective in controlling cracking in the short and long term?
With regards to the first question, it is well established that the penetration of chlorides
into hardened concrete roughly follows a square root time function [5]. Moreover, field
experience shows that many concrete highway structures in the U.K. show signs of
deterioration within twenty years of construction [6]. Thus, if the cover to the
reinforcement was increased from 40mm to, say, 100mm i.e. an increase of 60mm above
the nominal cover specified in Eurocode 2 [7], the time taken for chlorides to reach the
level of the reinforcement will increase by a factor of (100/40)2 i.e. approximately six
Appropriate concrete technology 410
fold. Therefore providing nominal covers of 100mm to the steel reinforcement should
eliminate the possibility of corrosion occurring during the design life of most structures.
The second question is more difficult to answer. Short term flexural tests on concrete
beams reinforced with equal amounts of steel or GFRP show that the beams reinforced
with GFRP experience greater deflections and have larger surface crack widths compared
with steel reinforced beams [8]. This is due to the fact that the stiffness of GFRP rebars is
about 25% that of steel and that the bond strength of GFRP with concrete is also
significantly lower being approximately 65 % that of the steel bars to concrete strength
[9]. The long term performance is also difficult to judge since there is evidence that
certain glass fibre composites deteriorate in alkaline environments typical of those found
in concrete [10, 11]. In addition, most resins are reported to have poor resistance to creep
[12]. Both factors could therefore increase the surface crack widths with time.
Since it was impossible to predict the behaviour of the GFRP rebars in this design it
was considered necessary to carry out laboratory investigations. The following section
describes the work which is currently under way to assess both the short term and long
term behaviour of beams made with this design.
EXPERIMENTAL PROCEDURE
The beams were allowed to cure in their moulds for two days. Thereafter they were
removed from the mould and wrapped in damp hessian and polythene, and allowed to
cure for a further 26 days.
Prior to testing, the beams were painted white in order to ease monitoring of crack
development during loading. The beams were tested in bending using a four point loading
arrangement as shown in Fig. 2. The effective span of the beam was 2.9m. The load was
applied at a rate of 0.03 kN/sec, generally in increments of 12.5 kN or 25 kN up to a
maximum of 210.5 kN in order to produce a design moment of 100 kNm in the 1m
central pure bending region of the beam.
At each load increment, the crack widths were measured using a portable microscope.
The position and extent of individual cracks was also recorded. In addition, the mid-span
defections of the beams were measured by means of a linear displacement transducer.
Appropriate concrete technology 412
The beams were stressed in four point bending using steel rollers and yoke frames as
shown in Fig. 3 to induce cracking in the GFRP reinforced covers. The loading at each
yoke was increased to 16kN to produce a design moment of 4kNm in the 900mm pure
bending regions in the beams. Any relaxation in the loading due to creep effects was
corrected by monitoring the load cells attached to the steel yokes. The beams were stored
in a room at a temperature and relative humidity of, respectively, 214C and 3510%.
The mid-span deflections of the beams were measured using dial gauges and the crack
widths were monitored using a portable microscope.
Fig. 5 shows the predicted and experimental crack widths as a function of design moment
for Beams 14. The predicted surface crack widths, w, were calculated using equation 1,
taken from BS8110 [14]:
A new method of preventing reinforcement corrosion in concrete structures using GFRP rebars 415
(1)
where acr is the distance from the point considered to the surface of the nearest
longitudinal bar; the average strain at the level where cracking is being considered;
cmin is the minimum cover to the tension steel/GFRP rebar; h is the overall depth and x is
the depth of the neutral axis.
As can be seen from Fig. 5a, the predicted crack widths for Beam 1 always exceed the
experimental values. Generally, there is good agreement between the experimental and
predicted crack widths. With Beams 24, it was found that if the presence of the GFRP
rebars was ignored in calculating the depth of neutral axis and stress in the steel, good
agreement existed between the predicted and experimental crack widths (Figs. 5b5d).
This suggests that the GFRP rebars do not significantly affect the moment capacity of the
section, which might otherwise result in an over-reinforced failure. The GFRP rebars are
largely ineffective structurally probably because they are not tied to the shear
reinforcement.
(2) Deflections
The maximum mid-span defections for Beams 14 were, respectively, 4.7mm, 5.3mm,
5mm and 4.0mm. The predicted deflection for Beam 1 was 5.9mm and was calculated
using equations 2 and 3, taken from BS8110 [14]:
Appropriate concrete technology 416
(2)
where L is the effective span of the member; K is a constant that depends on the shape of
the bending moment diagram; 1/rb is the curvature at mid-span and is given by:
(3)
in which fc is the design service stress in the concrete; Ec is the short term modulus of the
concrete; fs is the estimated design service stress in tension reinforcement; d is the
effective depth of the section; x is the depth of the neutral axis and Es is the modulus of
elasticity of the reinforcement.
The maximum permissable deflection is 11.6mm and therefore all the beams comply
with the deflection requirements in BS8110. Fig. 6 compares the deflections occurring in
Beams 14. Generally, it can be seen that the sections containing GFRP rebars showed
similar deflection characteristics to Beam 1 with no GFRP and therefore the existing
approach for estimating deflection in BS 8110 can be used, if the presence of the GFRP is
ignored.
(2) Deflections
Fig 8 shows the change in mid-span deflection with time for beams L1 and L2. After five
months of testing the actual deflections in beams L1 and L2 were, respectively, 1.99mm
and 2.32mm which are both significantly smaller than the predicted value of 6.3 mm.
CONCLUSIONS
This paper has described a novel method called Supercover Concrete for preventing
reinforcement corrosion in concrete structures using GFRP rebars. The method involves
using conventional steel reinforcement together with concrete covers in excess of
100mm, with a limited amount of GFRP rebars in the cover zone. Short term loading
tests on single span, simply supported reinforced concrete beams made using this method
show:
Appropriate concrete technology 418
1. Introducing GFRP rebars in the concrete cover does not significantly affect the
structural performance of the member.
2. The GFRP rebars are effective in controlling surface cracking and can be used to
reduce surface crack widths to values currently recommended in BS8110 and
Eurocode 2.
3. Deflections under working loads are not significantly affected as a result of introducing
the GFRP rebars in the cover zone.
4. Existing procedures given in BS8110 can be used to assess structural strength, design
crack width and deflections.
It is too early at this stage to draw any conclusions regarding the long term performance
of this design method. However, it should be remembered that any deterioration of the
GFRP will not compromise the strength of the structure.
ACKNOWLEDGEMENTS
The authors thank International Grating Inc., Huston, Texas, for donating the GFRP
rebars used in this work.
REFERENCES
12. EHSANI, M.R. (1993) Glass-fiber reinforcing bars. Alternative materials for the reinforcement
and prestressing of concrete, Ed. J.L.Clarke, E & FN Spon, London, pp3454.
13. INTERNATIONAL GRATINGS INC. Kodiak: Fiberglass-reinforced plastic rebar data sheet,
1992, Huston, Texas.
14. BRITISH STANDARDS INSTITUTION. BS8110: Structural use of concrete. Part 1: Code of
Practice for Design and Construction; Part 2: Code of Practice for Special Circumstances. BSI,
London, 1985.
CRACKING OF CARBON FIBRE
REINFORCED MORTARS
L Kucharska
Technical University of Wroclaw
A M Brandt
Polish Academy of Sciences
Poland
Key words: Cement based composites, Pitch carbon fibres, Silica fume, Acoustic
emission, Flexural toughness, Relative toughness Index It.
Professor Leokadia Kucharska is Head of Building Materials Department at the
Technical University, Wrocaw, Poland. Her background is chemistry and she is
interested in relations between composition, rheology, structure and properties of brittle
engineering materials. Recently, she has been involved in research in High Performance
Concretes and in cement based fibre reinforced composites. Dr. L. Kucharska has
published numerous papers in the field of ceramics and of cement composites.
Professor Andrzej M.Brandt is Head of the Strain Fields Department of the Institute of
Fundamental Technological Research (IFTR), Polish Academy of Sciences, Warsaw,
Poland. His main research interests are in cement based composites, their mechanical
behaviour and performance. Dr. A.M.Brandt is the author of several papers and books on
optimization and mechanics of structures and of materials. He is a consulting member of
ACI 544 Committee (Fiber Reinforced Concrete).
Cracking of carbon fibre reinforced mortars 421
INTRODUCTION
Carbon fibres exhibit high efficiency as micro-reinforcement due to their strength and
dispersion in cement based matrices. As reinforcement they can lead to high flexural and
tensile strength, impact resistance and toughness of composite materials. Taking into
account also their durablity in various environmental conditions and relatively low cost,
in many structures they may be considered as optimum reinforcement, e.g. cladding
panels in high rise Ark-Mori building in Tokyo and in Al Shaheed Monument in
Damascus, roofing sheets, repairs of concrete structures, foundations for machinery,
bridge beams and decks, permanent forms, lightning arrestors, transmission poles, wave
absorbers, etc.
One of the first test results on pitch carbon fibres used as dispersed reinforcement for
brittle cement based matrices was published in the 1980s. OHAMA et al. [1] tested
specimens with various kinds of the fibres and obtained excellent mechanical properties
for these composites. According to AKIHAMA et al. [2] the improvement was related to
two phenomena: pull-out and fracture of the fibres. NISHIOKA et al. [3] also observed
these two different processes and suggested that fibres below 600 MPa tensile strength
should not be used as reinforcement. Mechanisms of control of microcracks by carbon
fibres were considered by BRANDT and GLINICKI [4], BAYASI [5]. Later,
KUCHARSKA [6], KUCHARSKA and BRANDT [7] have shown that with appropriate
mixture proportions and sufficient fibre content the development of multiple cracking and
the enhancement of the mechanical properties was possible.
Several other problems were also considered:
uniform dispersion of the fibres and reduction of fibre volume to decrease the cost,
QUIJUN ZHENG and CHUNG [8], PU-WOEI CHEN and CHUNG [9],
influence of the length of fibres on their efficiency, OHAMA et al. [1], PARK et al.
[10],
influence of the kind and content of aggregate, SOROUSHIAN et al. [11].
In this paper, tests on cement mortars reinforced with short pitch-based carbon fibres
with and without silica fume reported previously by KUCHARSKA and BRANDT [6],
[7], [12], are considered again. Cracking in tested specimens and load-deflection curves
related to acoustic emission activity are analysed in detail. Enhancement of flexural
toughness of this kind of composites by the phenomenon of multiple cracks as shown in
KUCHARSKA [6] is confirmed.
EXPERIMENTAL
The components of the mixes and their proportions have been described in detail in [6]
and [7]. The basic mix components were:
Ordinary Portland cement (OPC) of Polish origin from Rejowiec,
superplasticizer (SP) with melamine formaldehyde sulfonate as an active ingredient,
1.1% per mass of cement mass,
low modulus pitch-based carbon fibres (CF) C-1035 produced by Kureha Carbon
Industry Co. Ltd., length 3 mm, dia.14.5 m, f=720 MPa,
silica fume (SF) from aziska in Poland, and
Appropriate concrete technology 422
The fibres were dispersed using the wet method, [9], and the mixing time was extended
up to 15 minutes. An ordinary mortar mixer was used. The plate specimens, cast in flat
moulds 140x160 mm and 15 mm deep, were demoulded after 24 hours and were cured in
saturated lime water until two days before testing.
Test were carried out after 28 and 56 days. Specimens 15x20x160 mm were sawn out
of the plates and subjected to bending. Third-point loading was applied using an Instron
1251 machine with a closed-loop servo control system enabling stable tests to be
performed. The tests were carried out under deflection control at rate of 0.05 mm/min
until the applied load reached zero or the displacement reached 1 mm, whichever was the
first.
The displacement transducer (LVDT) was attached to the specimen at the supports by
a special yoke to measure load-point deflections with respect to support points. This
method eliminates errors related to rigid movements of the specimen and to other
possible parasitic displacements. Load-deflection curves served for calculation of total
absorbed energy Wtot as the area under the curve up to the end of the test.
Acoustic emissions were recorded using a transducer with a nominal resonance
frequency of 200 kHz. The receiving head was situated on the specimen between a
support and loading points. In the reported tests the AE equipment recorded all signals
above certain level of discrimination, however the magnitude and amplitude of an
individual AE event could not be measured. Therefore, after these results it was
impossible to execute a deep analysis and to detect the source of the event: crack in the
matrix or interface, or fracture of a fibre. Only the number of single events, the rate of
increase and the total number at different stages have been recorded. For further analysis
two stages of loading were selected: peak load and maximum deflection at the end of the
test.
A PC based acquisition system was used within the ASYST environment and
sampling rate of 1 Hz was applied.
Cracking of carbon fibre reinforced mortars 423
In Figs. 1 and 2 load-deflection curves and the magnitude of the AE are shown for
selected characteristic specimens with different Vf and with or without silica fume (SF).
The number of AE events at two stages as mean values of series of 46 specimens are
given in Table 2.
Table 2. Numbers of AE [100] events in tested
specimens
contents of carbon fibres Vf
age levels or load 0 0.26 0.52 0.79 1.5 3.0
28 up to max. load 24 15 25 60
days up to end of test (without SF) 53 1800 2080 2270
up to max. load 15 1 38 570
up to end of test (with SF) 108 1650 2300 2300
56 up to max. load 14 13 9 49 210 200
days up to end of test (without SF) 370 1466 1300 1910 2000 3250
up to max. load 29 59 27 460 950 1050
up to end of test (with SF) 350 1450 2350 2400 3050 3750
In Fig. 3 the influence of fibre content on total AE events (Table 2) and on Wtot for
specimens with and without SF are shown.
From these results it was found that in all specimens tested very few AE events were
detected at low levels of loading before the appearing of the 1st crack. This observation
showed that the AE recording was correctly executed and that the lower discrimination
level for AE signals had been appropriately selected.
Only single cracks occurred before effective initiation of the damage process, as has
been observed by BALAGURU and SHAH [13]. Then, after the change of slope of the
load-deflection curve the higher rate of cracking was recorded by AE.
Appropriate concrete technology 424
As can be seen from Figs. 1 and 2 the influence of fibre reinforcement was considerable.
In specimens with more fibres higher numbers of AE events occurred and this was due to
more intensive but dispersed cracking, Table 2 and Fig. 3. Moreover, the shapes of the
AE-deflection curves are different for plain and reinforced specimens. The variations of
slope of load-deflection curves which are due to microcracking and later to development
of major cracks are correlated with rapid increases in the number of AE events and of
absorbed energy Wtot, Fig.3.
Systematic analysis of flexural toughness of all specimens tested was carried out using
different systems of toughness indices. The most appropriate for estimation of the
influence of the volume of fibres and of silica fume on the mechanical properties was the
relative toughness index It. It is defined as the ratio of the energy absorbed in FRC
specimen up to the end of the test to the similar energy in a comparable beam without
fibre reinforcement, ACI 544 [14]. This index is free of uncertainty of determination of
the 1st crack and, besides, it provides a basic measure of the efficiency of the fibres.
Values of It for tested specimens (Fig.4) show that the toughness was considerably
enhanced for specimens with Vf0.79% but only when SF was added.
Appropriate concrete technology 426
CONCLUSIONS
Significant enhancement of the flexural toughness of pitch carbon fibre composites may
be obtained when multiple cracking ocurs and this requires that the fibre-matrix bond is
improved by the addition of a microfiller and that the fibre volume is adequate.
Appropriate concrete technology 428
REFERENCES
1. OHAMA, Y, AMANO, M, ENDO, M. Properties of carbon fiber reinforced cement with silica
fume. Concrete International, March 1985, pp 5862.
2. AKIHAMA, S, SUENAGA, T, BANNO, T. Mechanical properties of carbon fibre reinforced
cement composites. Int.J.of Cement Comp.and Lightwt.Concr., Vol.8, No.1, 1986, pp 2133.
3. NISHIOKA, K, YAMAKAWA, S, SHIRAKAWA, K. Properties and applications of carbon
fibre reinforced cement composites. Proc.RILEM Symp.Developments in Fibre Reinforced
Cement and Concrete. Sheffield, 1317 July 1986.
Cracking of carbon fibre reinforced mortars 429
ABSTRACT. The concept of using fibres in mortars dates from the past.
Nowadays, fibres from various sources such as asbestos, polypropylene,
cellulose, glass and others are succesfully applied to mortars for repair or
remaking purposes as well as for a great variety of precast components or
thin walls. By this way, plastic and longtime shrinkage defects are
avoided resulting in remarkable longevity of materials. The aim of this
experimental work is the production of good quality fiber cement products
by using cost effective by-products and find alternatives to asbestos
cement industry. Flexural, tensile and splitting tensile strength were
measured as well as the first crack load. Cracking tendency of composites
was estimated by observing with microscope thin plates cured at low
humitidy environment 4045% RH.The parameters which were studied
are: a) type and percentage of fibres b) percentage of cement replacement
by fly ash c) cementitious content d) addition of sand. Based on results it
could be said that there are cost saving alternatives in fiber cement
composites. High percentages of cement from 30 to 60 can be replaced by
fly ash while the total amount of cementitious can be reduced.
Polypropylene seems to render fiber cement grout of good quality.
keywords: Fiber Reinforced cement, Grouts, Fly ash, Natural fibers. Polymeric fibres,
Flexural, Tensile, Splitting, Strength.
Low cost fibre reinforced cement products by using inexpensive additives 431
INTRODUCTION
Fiber reinforced mortars or grouts are widely applied in cast and precast composites.
Especially fibres of low moduli of elasticity such as natural fibres (cellulose, jute, sisal,
bamboo, coconut) and polymeric fibres (polypropylene, nylon polyethylene, PVA) have
been used since 1960 with high efficiency [1] acting as crack inhibitors. These grout
mixes are characterized by high cement content and large volumes of material consumed
in situ. Applications of this type are groutings for support lining for mine tunnels or rock
slopes or for repairing canals or damaged concrete surfaces as well as for construction of
thin section elements or thin (<25mm) plaster coverings.These fibres can also replace
secondary steel reinforcement in thin slabs [2] [3]. In precast cement based industry the
above mentioned fibres may replace asbestos in pipes, roofing sheets, thin walls or boxes
for various uses [2]. Actually they compete asbestos cement products developed in 1900
[3] which according to European Committee directive 91/659 EC L363/91 do not have
future since they are considered dangerous to health.
The great advantages attained by fiber cement mortars when randomly distributed
fibres added are [2] [3] [4]:
Increase stress-strain capacity and flexural strength. Under static loadwhatever the
mode loading- a substantial increase in the area under stressstrain curve is observed.
Improve toughness and post cracking ductility.
Act as energy absorbers.
Reduce plastic shrinkage of first hours after placing and inhibit cracking.
Improve shear strength. An increase of 17% in shear strength due to polypropylene
fibres has been found experimentally [5].
Exhibit better bond to substrate. Therefore the inherent defect of cementitious
matrices, brittleness, is overcome. This is of special importance in case of grouts or
slurries where there are no aggregates to hinder crack propagation.
Appropriate concrete technology 432
The concept of reinforcing low potential materials with fibres dates from the past [6]. Fig.
1 shows the characteristic behaviour of chopped straw in old sunbaked mud bricks. The
fibres cause slow development of a wide fracture zone. They tranfer stress across cracks
by bridging them. Therefore the materials can retain some residual post crack strength
and fragmentation is inhibited [2] [3]. The good performance of fibres in a given mortar
depends mostly on fibres content, Vf (percent by volume of composite), fiber geometry
(aspect ratio: fiber length/fiber diam. and distribution of fibres) [2] [3] [4] [6]. Different
techniques such as Hats check method and wet-mix shotcreting in precast and in cast
application are the most known present day processes of manufacturing fiber cement
composites but much amelioration could be made [2] [3].
SCOPE
The aim of this research programme was firstly to find alternatives to asbestos cement
industry ELLENIT, one of the remaining factories of ETERNIT producing asbestos
cement products. The objectives were a) to reduce the cost of production by replacing a
substantial proportion of cement with low cost fly ash -locally produced- and b) to test
the performance of alternatives in conjuction with cement substitution for fly ash. Before
proceeding to full scale tests in industry, an experimental work was carried out at Lab. of
Building Materials of University of Thessaloniki in order to decide upon proportioning
parameters. A number of 29 grout and slurry mixtures were prepared and tested. The
result of this part of research work could find direct application in cast fiber reinforced
mortar, grout or slurry production.
Low cost fibre reinforced cement products by using inexpensive additives 433
EXPERIMENTAL DETAILS
Materials
Portland cement 145 type was used in combination with greek fly ash wich is a High
Calcium Fly Ash (HCFA). The raw HCFA has a free lime content of 910% but it is
ground and treated in the mill with equivalent quantity of H20 so as the free lime content
of the final HCFA decreases to 3% (CaO+H20Ca(OH)2). The suitability of HCFA in
slurry and grout mixtures was found in previous experimental work [7]. Chemical
composition of the cementitious materials is given in table 1. Limestone filler or river
sand of max. size up to 1 mm were used in grouts. In asbestos slurries dry waste was
added to limestone filler according to practice followed in asbestos cement industry.
Apart from asbestos (chrysotile) used in two densities 25103 and 45103Kg/m3, three
other type of fibres were tested: cellulose, fibrilated polypropylene, and PVA. Although
the mechanical performance of mortars with these fibers is inferior to steel or glass fiber
mortars [3] they were selected because of their better behaviour in grouts and slurries.
Cellulose mortars present high water rententivity and standability while polypropylene
fibres are very effective in reducing shrinkage and have excellent resistance to moisture,
acids and alkalis [2] [3]. Both of them are relatively inexpensive materials. In addition,
HCFA
Table 1. Characteristics of cementitious materials
used in fiber reinforced grouts
Constituents Portlanl cement 145 % Ptolemaida fly ash mean value
SiO2 19.70 32.5
Al2O3 4.80 11.3
Fe2O3 3.70 5.02
CaO 54.10
added to cement mixtures for cement replacement reduces the pH of mixture with time.
This is advatageous for cellulose which is sensitive to high alkaline environment of
hydrated cement.
The characteristics of the used fibres, have been provided by the suppliers.
Polypropylene
Type: PPC3H6 fibrilated
Length: 12mm
Diam.: 18103 mm
Spec, grav.: 0.9103 Kg/m3
Harmless
Tens, streng.: 34x102MPa
Mod. of elastic.: 69103MPa
Natural cellulose
Type: ARBOCEL PZ8
Length: 1, 4 mm
Moist.cont.: 1012%
Spec, grav.: 1.5103 Kg/m3
Harmless
Resistant to diluted acids and alkalis
Tens, streng. 35102 MPa
Mod. of elast. 810103 MPa
Asbestos
Type: chrysotile (ZIDASBEST S.A.)
Diam: 0.055103 mm
Spec.gr.: 2.5103 Kg/m3
Tens, strength: 164103MPa
Dangerous in health when inhaled
PVA
Type: V-10 (UNITIKA NEWLON A.A)
Length: 6mm
Diam.: 1.6d
Harmless
Proportioning
The mix proportions of all mixtures are shown in Table 2. Instructions for mixing the
constituents were given by the supplier of fibres. Usually, fibres were added in form of
slurries. Cement was used in different quantities. Water amount was adjusted for
achieving adequate workability of mortars or fluitidy of grouts and slurries.
Two triplets of 4040160 mm prisms were tested for flexural strength of each age. A
set of four briquettes (known from cement tensile tests) was used for direct tensile
strength determination. Two cylinders 50100 mm were used for splitting tensile strength
loaded according to ASTM C-496. The parts of prisms after testing in flexure were
pressed in order to measure the first crack load. Dynamic modulus of elasticity was
estimated by measuring ultrasonic waves velocity through Ferret prisms by sonometer.
The compaction of fresh mixtures was made at the vibrating table described in cement
Low cost fibre reinforced cement products by using inexpensive additives 435
regulations for preparing cement mortars. With mixtures of high content in fibres and
especially with cellulose, stiffness problems appeared. A superplasticizer of
lignosulfonate origin was used up to 1% by weight of cementitious to anticipate them.
Specimens were cured in a climate chamber at 45% Relat. Humidity and 25C. This
realistic environment was chosen to be similar to curing conditions of precast products
and construction conditions in summer period. In addition, four plates 400400 mm of
6mm thickness made with the four types of fibres (Mixes No 9, 27, 28, 29), were casted
in order to see the appearance and resistance to cracking. The plates were placed at the
above mentioned environment and they were examined microscopically (x100
magnification) at regular periods for over 3 months to see cracks or other deformations.
Mean values of test results are shown in Tables 3, 4, 5 and 6. It can be said that relatively
great deviations from the mean values up to 19% were observed especially in mixtures of
high fibre content and tensile strength determinations. Asbestos fiber reinforced
composites have a specific gravity around 0.55103 Kg/m3 at a total quantity of
cementitious 75% w/w in dry mix. Higher strengths in tension exhibit the slurries with
the lower water/cementitious (w/cmtt) ratio, 1.93 (fig. 2). Comparing mixes No 8 and 9 it
seems that 40% w/w replacement of cement by HCFA results in a 15% reduction in 28-
day flexural strength but at significant increase at 3-months strength. At high w/cement
ratios of 2.47 or 5.80, HCFA addition up to 50% w/w of cement does not differentiate
mechanical characteristics. Net HCFA slurries developed at 28 days flexural strength
around 70% of corresponding of net cement slurries of the same w/cement ratio.
Therefore the substitution of cement with HCFA should be made according to the
required early strength and final strength of the product.
Cellulose fiber reinforced grouts have a specific gravity around 0.60103Kg/cm3 for a
cementitious content in dry mix of 70%. In comparison with asbestos mixtures of almost
the same w/cement ratio higher strength can be achieved with addition of a proportion
0.34% cellulose in dry mix. HCFA percentages up to 60 contribute effectively to flexural
strength development especially at
Table 2. Mix proportions of fiber reinforced cement
grouts and slurries
Mix Cement HCFA Water/ Asbestos Cellulose Polyprop. P.V.A. Limest. Sand Supe
ture cement. ehysot. filler (max. rplas. (%
Code +Dry size 1 w/w of
No waste mm) cement.)
1 1 5.8 0.25 0.08
2 1 6.6 0.25 0.08
3
4 1 5.8 0.25 0.08
5 0.6 0.4 5.8 0.25 0.08
6 0.5 0.5 5.8 0.25 0.08
Appropriate concrete technology 436
days days months days days days days months days days months days days months
24 8.95 9.38 7.20 1.22 1.90 7.30 9.50 7.80 1.33 1.41 1.39 4.65 6.70 6.12
25 2.30 3.50 3.18 1.25 1.13 4.40 6.10 6.10 1.41 1.50 1.50 4.80 8.50 8.56
26 1.00 2.00 1.50 0.21 0.61 2.20 3.20 2.70 1.33 1.40 1.41 1.85 1.82 3.67
27 3.60 5.70 1.10 3.80 6.70 1.60 4.33 4.33
the age of 2 months. Flexural strength and first crack load seem to be more influenced by
cellulose fibres.
Polypropylene fiber reinforced grouts with 48% w/w cementitious content in dry mix,
exhibited the higher strength values of 10 MPa with a proportion of fiber content 0.072
and w/cement ratio from 0.86 to 1.09 (mixes No 14, 15, 16). Even with lower proportion
of fibres 0.040 the values for strength in tension and compression remained higher in
comparison with those of asbestos and cellulose mixtures, although cementitious content
in dry mix was significantly reduced (26% w/w of dry mix). Fig.3 shows the effect of
fiber content in flexural strength. Addition of HCFA up to 50% does not change
Low cost fibre reinforced cement products by using inexpensive additives 439
considerably even the 7-day strength and it seems that HCFA cooperates very well with
polypropylene fibres in early and mainly longterm strength development. Comparing
pairs of mixes No 18, 21 and 17, 22 it is obvious that limestone filler could be
advantageously replaced by sand. Addition of superplasticizer 1% by weight of
cementitious content has a significantly positive effect in strength development by
improving the fluidity of grouts, (mixes No 17, 22). The specific gravity of
polypropylene fiber reinforced composites ranges from 1.25103 Kg/m3 to 1.35103 Kg/m3
according to cementitious content.
All PVA fiber reinforced composites showed a drop in strength (or no increase) at the
age of 3 months. This implies negative effects in the system cement/PVA fibres and
further research is needed. HCFA addition in cementitious content is not advatageous as
in the case of previous mentioned fibres. An obvious reduction in strength is observed
(mixes No 25, 26 and 24, 27). Limestone filler renders higher strength grouts in
comparison with corresponding mixtures with the same content of sand.
None of the plates examined by microscope showed cracks up to the age of 3 months.
CONCLUSIONS
1. In asbestos fiber reinforced slurries, cement could be replaced up to 3040 percent by
HCFA. A reduction of 1520% in strength should be taken into account at early ages
(7, 28 days). HCFA cooperates very well with cellulose fibres and could replace
cement up to 60% by weight. High quality fiber reinforced composites (up to 10MPa
flexural strength) can be produced by using polypropylene fibres. HCFA could be
added up to 60% by weight of cement in polypropylene grouts. Mixture of low
cementitious content in dry mix (26%) developed high flexural strength up to 3.0
Appropriate concrete technology 440
MPa. Sand could be used as an alternative for limestone filler. All these lead to the
conclusion that cost effective and high quality grouts or composites could be produced
by using polypropylene fibres.
2. HCFA seems to cooperate very well with all examined fibres except of PVA fibres.
3. Superplasticisers are very effective in fiber grout mixtures since they improve fluidity
and enhance strength development.
4. Based on data given by asbestos cement industry ELLENIT and a short time market
research the following economical considerations could be made: The replacement of
cement by HCFA up to 40% results in the 17% cost saving independingly the type
fibrous materials used.
The alternative of using polypropylene fibres instead of asbestos rises the cost of
slurry production up to 200% approximately but the resulting final cement
products are of high quality in comparison with the technical characteristics of
asbestos cement products.
REFERENCES
1. NAAMAN A.E Fiber Reinforcement for Concrete. Concrete International, Vol. 7, No3, March
1985, pp. 2125.
2. JOHNSTON C.D Fiber Reinforced Cement and Concrete. Advances in Concrete Technology.
CANMET MSL 926(R). Editor V.M.Mahotra pp. 629697.
3. MINDNESS S., BANTHIA N. Fiber Reinforced Cementitious Composites: Current Practice and
future Prospects, ACI SP-144, Concrete Technology. Past, Present and Future. Editor P.Kumar
Mehta, pp. 417446.
4. ACI. Manual of Concrete Practice, part 5, 544 1R-82.
5. SOROUSHIAN P. and others Mechanical Properties of concrete materials reinforced with
polypropylene or polyethylene fibres. ACI materials. Journal Vol 89, No 6, Nov-Dec. 1992, pp.
535416.
6. PAPAYIANNI I. High Calcium Fly Ash Grouts. Concrete 2000 Economic and Durable
Construction through Excellence. Editors: R.Dhir and M.R Jones, E. and FN SPON 1993, pp.
13051312
BIAXIAL FRACTURE BEHAVIOUR OF
FIBRE REINFORCED CONCRETE
E K Tschegg
M Elser
Technical University of Vienna
S E Tschegg-Stanzl
University of Agriculture
Austria
Keywords: Polypropylene-, glass-, steel- fiber reinforced concrete, uniaxial and biax-ial
fracture behaviour, specific fracture energy, uniaxial and biaxial splitting test
Professor Dr Elmar K.Tschegg is head of the Materials Science Laboratory of the
Institute of Applied and Technical Physics at the Technical University of Vienna. He is
involved in research work on the fracture behaviour of disordered materials like con-
crete, refractories, asphalt, wood and fiber reinforced materials etc.
Dr Michael Elser is project-leader in a Materials Science research center. His main
research interests include experimental and numerical studies of fracture processes in
various cementitious materials.
Appropriate concrete technology 442
INTRODUCTION
Though the compression strength of concrete is high its tensile strength is rather low and
responsible for crack initiation and damage of concrete structures in many cases.
Therefore it is tried to compensate this draw-back by reinforcing metallic or nonmetallic
fibers (fiber reinforced concrete, abbreviated FRC in the following). The main aim of
adding fibers is to improve especially the tensile strength, the impact strength and the
fracture toughness. Many investigations have been performed in order to study the
possibilities of making use of different kinds of fiber materials and fiber types. In this
context, the influence of volume fraction, length and cross section of the fibers on the
mechanical properties have been measured and discussed in the literature. Relatively few
investigations exist on the fracture behaviour in comparison with the other mechanical
properties like compression strength, bending strength etc. Most works (for example [1
5]) treat the uniaxial fracture behaviour of FRC, some exist on two-axial and multi-axial
compressive loading [611] and very few on the fracture properties of FRC during
twoaxial tension-compression loading, though this is the type of loading, which leads to
crack initiation and propagation in most cases. Therefore failure envelopes have been
determined in tension-compression loading experiments of unnotched specimens made of
FRC in [12, 13]. The influence of fiber length and number of fibers on the fracture
properties of Polypropylene FRC [14] and steel-fiber FRC [15] has been measured with
notched specimens. In addition, the biaxial fracture behaviour has been simulated theo-
retically in [16]. The main result of these studies is that, the notch bending strength is
only slightly influenced by a superimposed compression load during combined loading
conditions, whereas the specific fracture energy and especially the post-peak behaviour of
the load-displacement curves is influenced much.
In the present study the influence of a combined bending tension and a compression
load on the fracture behaviour of identical medium-strength concrete with and without
glass fiber or Polypropylene or steel fiber reinforcement has been investigated. The used
specimen shape and the type of loading are shown schematically in Fig. 1. The com-
pression stress 1 is kept constant during the fracture test. A new wedge splitting
procedure according to Tschegg [17, 18] has been used which allows uniaxial and biaxial
tests with cubic specimens in an easy and convenient way.
Biaxial fracture behaviour of fibre reinforced concrete 443
The basis setup for biaxial loading is the wedge splitting method according to [17, 19, 20]
for conventional uniaxial mode I fracture testing as shown in Fig. 2, part A. A slen-der
wedge, which is positioned in a groove on top of the specimen, splits this specimen. The
crack starts from a starter notch and ends on the opposite side along a linear support
piece. Friction losses during load transfer are reduced by using needle bearings, so that
the measuring results are hardly influenced. Load displacement and crack mouth open ing
displacement (CMOD) are measured on both ends of the groove with two displacement
gages (LVDTs). The LVDTs are fixed on a frame, which is connected with the
specimen only. More details are reported in [20].
Appropriate concrete technology 444
The biaxial loading system [21, 22], which is shown in Fig.2, part consists of an uniaxial
basis setup (part A) and an additional hydraulic compression load unit in order to
generate the biaxial stress state in the specimen. The compression load unit consists of
two mechanically independent frames. Three hydraulic cylinders are fixed to both of
them. A constant and homogeneous compressive load 1 is generated in the specimen in
this way [22].
Following construction and functional elements of the uniaxial and biaxial load unit
are the same:
Stiff properties of the whole load transfer equipment owing to the wedge load system
slender wedge
load transfer via a needle bearing with negligible friction losses
load transfer from testing machine to loading system of the specimen
ligament height and area as well as specimen shape, besides upper part of specimen (see
Fig. 3)
Different are:
The load transmission pieces are Lshaped for the uniaxial test. For the biaxial test,
they are part of the load transmission slabs in order to generate the stress 1 in the
specimen.
The linear support is a thin and low bar in the uniaxial test, whereas a long sword-like
piece is introduced between the two lower beams (Fig. part B) in the biaxial test,
without touching them or inhibiting the splitting procedure. Thus the specimen is
supported essentially similar as in the uniaxial test.
Biaxial fracture behaviour of fibre reinforced concrete 445
An important criterion to ensure a homogeneous and well defined biaxial state of stress in
the specimen, is to balance the transverse strain between concrete and steel. This problem
in treated in [22] in a fundamental way. An appropriate and optimum equipment to
balance the transverse strain is obtained by introducing a Teflon foil (approximately 0.5
mm thick) and a cardboard (approximately 12 mm thick) between the polished steel
load transmission slabs and the specimen. This constructive solution is simple,
inexpensive and accomplishes all requirements. More details of the new biaxial loading
system are described extensively and discussed in [21, 22].
Main aim of the present study is to quantify the influence of the fiber reinforcement on
the fracture properties. Concrete composition and storing conditions therefore were kept
constant for all test series. The fibers only were varied as to their material, length and
additions. Composition of the four test materials as well as their strength properties after
28 days water storage are summarized in Table 1.
Appropriate concrete technology 446
Quality and shape of fibers: The diameter of the Polypropylene fibers HAREX-Poly
Con is 1020 m. The glass fibers being called Cern-Fil-AR-Glass fiber have the
same diameter. The steel fibers are called HAREX-Steel fiber. Their diameter is 0.8
mm and they are not anchored at the ends. The authors are aware of the fact that glass
fiber reinforcement usually is not used for concrete but rather for mortar and in much
higher concentrations there than in this study. The results of this study therefore are of
academic value only as to the glass fiber reinforcement. The results, obtained with steel
and Polypropylene fibers on the contrary, are of practical relevance too.
Table 1: Properties of concrete mixes after 28 days
water storage
Type of aggregates: Natural gravel, limestone Maximum aggregate size: 16 mm
Cement content 240 kg/m3
Water content 174 kg/m3
Water/cement ratio 0.725
Density 2430 kg/m3
Youngs modulus ~30 GPa
Specimen shape and dimensions are shown in Fig. 3. Ligament and all other dimensions
as well as place and height of load application lines do not differ for uniaxial and biaxial
specimens. The top of the specimen, however, is different, as the load is applied via slabs
in the biaxial system, which are not needed for uniaxial loading. The specimens were cast
Biaxial fracture behaviour of fibre reinforced concrete 447
to boards (70 cmx15 cmx15 cm), removed from the molds after 2436 hours and stored
in water for 28 days. Before testing the specimens were cut to their final shape (see Fig.
3) and a notch being 2.5 to 3 mm wide was introduced.
Testing was performed in a mechanically driven machine (SCHENCK) with a load
capacity of 100 kN. The cross-head speed was 0.5 mm/min in all cases, which
approximately corresponds to the RILEM [27] recommendation for fracture tests. The
testing temperature was 2022 C. To obtain appropriate statistic values, testing was
performed with 34 specimens for each type of FRC and for each loading condition.
Load-displacement curves for uniaxial and biaxial testing of Polypropylene FRC are
shown in Fig. 4. Compared to plain concrete, FRC is characterized by a pronounced
softening behaviour in the post-peak regime for uniaxial loading (1/fc=0%, with fc being
the compressive strength). The softening behaviour is less pronounced for biaxial loading
(1/fc>10%) and decreases with increasing biaxiality, though this reduction is less
pronounced. The peak heights of the load-displacement curves for the different biaxial
loading conditions are not too different. A similar behaviour has been observed in this
study for glass FRC and steel FRC, so that these load-displacement curves are not shown
separately.
The characteristic fracture values Gf and 2NBS are plotted versus the normalized
compressive stress 1/fc with and without fiber reinforcements as parameter, in Figs. 5
and 6. Fig. 5 shows that the uniaxial notch-bending tensile strength 2NBS is not changed
by the fiber reinforcement. For biaxial loading, the values of the FRC are approximately
10% lower than of plain concrete. The curves are almost horizontal between 10 and 60%
Appropriate concrete technology 448
normalized compressive stress, which is typical for FRC and plain concrete with varying
composition [2225].
The Gf curves in Fig. 6 demonstrate that the crack growth resistance of FRCs are much
higher compared to plain concrete for uniaxial loading (1/fc=0). The ratios are: Gf,
poly:Gf,steel: Gf,glass:Gf,plain=9.4:5.7:1.45:1, which means that the Gf value is increased by a
factor of 10 by the addition of Polypropylene fibers. These ratios are lower for biaxial
loading. For 1/fc being between 20 and 60% the ratios are only 4.5: 1.6:1.2:1, showing
that Gf is approximately 5 times higher for Polypropylene FRC. The maximum decrease
of the specific fracture energy is found for steel FRC. It is worthy to note that glass fibers
increase the fracture properties only slightly. This FRC is not used for practical purposes.
Biaxial fracture behaviour of fibre reinforced concrete 449
It is of interest for the practical engineer to know, that FRC looses more than 50% of
its crack growth resistance under biaxial loading in comparison to uniaxial loading. If Gf
values of uniaxial loading tests are used for calculations, extremely unsafe values are
obtained, as too high Gf values have been assumed. This result demonstrates that biaxial
fracture properties of FRC have to be measured for practical applications, instead of only
introducing uniaxial values or substituting them by subtracting some undefined value.
The remarkable reduction of the specific fracture energy under biaxial loading may be
explained with a simple model according to [1416]. Though the fibers are randomly
oriented in concrete, it is useful to assume three main orientations in order to model the
fracture energy consumption. These three orientations are shown schematically in Fig. 7:
In (a) the fibers are oriented normal to notch root (or crack front) and parallel to crack
growth direction, in (b) parallel to notch root (or crack front) and normal to the crack
growth direction and in (c) normal to notch root (or crack front) and normal to the crack
growth direction.The fiber orientation (b) does not influence the fracture behaviour much,
Appropriate concrete technology 450
CONCLUSIONS
Fracture tests have been performed with a new wedge splitting procedure under uniaxial
and biaxial loading of unreinforced and Polypropylene-, steel- and glass fiber reinforced
medium-strength concrete using cubic specimens with deep notches. The following re-
sults were obtained:
1. The nominal notch bending tensile strength 2NBS of unreinforced concrete and FRC
does not differ much under uniaxial loading conditions. The notch bending tensile
strength values of FRC decrease however under biaxial loading in contrast to plain
concrete in the range of 1/fc=1060% by approximately 10 %.
2. The resistance against crack propagation, which is characterized by the specific
frac-ture energy, is increased by reinforcing Polypropylene fibers by a factor of 10, by
steel fibers by a factor of 6 and by glass fibers by a factor of 1.5, in comparison to plain
con-crete. For biaxial loading in the range of 1/fc=2060%, the Gf values of Polypropy-
Biaxial fracture behaviour of fibre reinforced concrete 451
lene FRC are 4.5 times, of steel fibers 1.7 and of glass fibers 1.2 times higher than for
plain concrete. The results demonstrate that the specific fracture energy is much reduced
under biaxial loading in comparison to uniaxial loading.
3. For the practical engineer it is important to know, that characteristic fracture values
of FRC, which have been determined under uniaxial loading, cannot be used to calculate
biaxial values (as it is done quite often), even if safety differences are assumed. In order
to obtain reliable values, biaxial fracture experiments have to be performed.
ACKNOWLEDGMENT
The authors thank Mr. A.Plzl, BPV Lanzendorf, Austria for the expert manufacturing of
the specimens. The project was financed by the Austrian Nationalbank and the Fonds zur
Frderung der wissenschaftlichen Forschung (Austrian Science Foundation) under
project number P 8885-TEC.
REFERENCES
1. LI V.C.AND LEUNG C.K.Y. Tensile failure modes of random discontinuous fiber reinforced
brittle matrix composites. In Fracture Processes in Concrete, Rock and Ceramics, ed. by
J.G.M.van Mier, J.G. Rots and Bakker, E&FN Spon, 1991, pp 285294
2. WANG Y., BACKER S.AND LI V.C. An experimental study of synthetic fiber reinforced
cemenitious composites, J. of Mat. Science, 22, 1987, pp 42814291
3. MOB ASTER B., QUYANG C., C ASSTRO-MONTERO A. AND SHA S.P. Tensile behaviour
of fiber reinforced concrete. In Concrete, Rock and Ceramics, ed. by J.G.M.van Mier, J.G. Rots
and Bakker, E&FN Spon, 1991, pp 269284
4. GOPALARATNAM V.S., SHAH S.P., BATSON G.B., CRISWELL M.E., RAMAKRISHMAN
V AND WACHARATANA M. Fracture toughness of fiber reinforced concrete, ACI Journal,
Aug. 1991, pp 339353
5. LI V.C.Postcrack scaling relations for fiber reinforced cemenitious composites, J. of Mat. in
Civil. Eng., ASCE, Feb. 1992, Vol. 118, No.11, pp 22462264
6. TORRENTI J.M., DJEBRI B. AND B ASCOUL A. Comparative study of two biaxial presses
for concrete, Materials and Structures, No. 24, 1991, pp. 5260
7 YIN W.S., SU E.C.M., MANSUR M.A. AND HSU T.C. Fiber reinforced concrete under biaxial
compression, Eng. Frac. Mech., Vol.35, 1990, Nr. 1/2/3, pp. 261268
8. YIN W.S., SU E.C.M., MANSUR M.A. AND HSU T.C. Biaxial Tests of Plain and Fiber
Concrete, ACI Journal, Aug. 1989, p. 236243
9. WANG C.Z., GUO A.H. AND ZHANG X.Q. Experimental investigation of biaxial and triaxial
compressive concrete strength, ACI Material Journal, V.84, Apr. 1987, pp. 92100
10. AHMAD S.H. AND SHAH S.P. Complete triaxial stress strain curves for concrete, Journal of
Structural Engineering, ASCE, Vol. 108, No. ST4, 1982, pp. 728742
11. TAYLOR M., TAI M. AND RAMEY M. Biaxial Compressive Behavior of Fiber Reinforced
Mortar, Proc. ACI, No. 9, 1975, p. 496501
12. KO H.Y., MEIER, R.W., EGGING D.E., STURE S. AND FENG C.C. Constitutive properties
of steel fiber reinforced concrete in multiaxial loading, Proceedings, Interaction of non nuclear
munitions with structures, US Air Force Academy, May 1983, Part 2, pp. 7176
Appropriate concrete technology 452
13. MEIER R.W.The Strength and Behavior of Steel-Fiber Reinforced Concrete under Combined
Tension-Compression Loading, M.S. Thesis, Univ. of Colorado, Dep. of Civil, Environmental
and Architectural Engineering, Boulder, 1983
14. ELSER M., TSCHEGG E.K. AND TSCHEGG-STANZL S.E. Fracture behaviour of
polypropylene fiber reinforced concrete under biaxial loadingan experimental investigation,
submitted to ACI, Jan. 1995
15. ELSER M., TSCHEGG E.K., AND TSCHEGG-STANZL S.E. Fracture behaviour of steel fiber
reinforced concrete under biaxial stress, submitted to ASCE J. of Mat. in Civil Eng., Aug. 1995
16. ELSER M., TSCHEGG E.K., FINGER R. AND TSCHEGG-STANZL S.E. Fracture behaviour
of polypropylene fiber reinforced concrete under biaxial loadingmodelling und computer
simulation, submitted to ACI, Jan. 1995
17. TSCHEGG E.K. Equipment and appropriate specimen shapes for tests to measure fracture
values (in German) Patent No. 390328, Austrian patent office, 1986
18. TSCHEGG E.K. AND ELSER M. Biaxial loading equipment for fracture tests. (in German)
Patent 1525/94, Austrain patent office, 1994
19. TSCHEGG E.K. Load introduction equipment for fracture tests in brittle materials (in German),
Patent No. 396.997,1990
20. TSCHEGG E.K. New Equipment for fracture tests on concrete, Materialprfung/Materials
Testing, Vol.33, 1991, pp. 338343
21. TSCHEGG E.K. AND ELSER M. Einrichtung zur biaxialen Bruchprfung (in German), patent
application 3. 8.1994
22. TSCHEGG E.K., ELSER M. AND TSCHEGG-STANZL S.E. Biaxial Fracture Tests on
Concrete, Development and Experience; Cement &Concrete Composites 17 (1995), 5775
23. KOTSOVOS M.D. AND NEWMAN J.B. Fracture mechanics and concrete behavior, Magazine
of Concrete Research (London), Vol. 33, No. 115, 1981, pp 8799
24. ZIELINSKI A.J. Concrete under biaxial compressive-impact tensile loading, in Fracture
Toughness and Fracture Energy of Concrete, ed.F.H.Wittmann, Elsevier Science Publisher,
Amsterdam, 1986, pp 479489
25. WEERHEIJM J., Concrete under Impact Tensile Loading and Lateral Compression, Thesis at
the Prince Maurits Laboratory TNO and Delft University of Technology, 1992
26. TSCHEGG E.K., ELSER M. AND KREUZER H. Mode I Fracture Behavior of Concrete under
Biaxial Loading, Journal of Materials Science, 30 (1995) 235242
27. RILEM Draft Recommendation (50-FMC): Determination of the fracture energy of mortar and
concrete by means of three-point bending tests on notched beams, Materials and Structures, Vol.
18, 1985, pp. 287290
DESIGNING STRUCTURES USING
NON-FERROUS REINFORCEMENT
JL Clarke
P ORegan
Sir William Halcrow & Partners
UK
Keywords: Aramid fibre, Bending, Bridges, Carbon fibre, Composites, Glass fibre,
Reinforced concrete, Resins, Tunnels.
Professor John L Clarke is a Principal Engineer with Sir William Halcrow and Partners
and visiting professor at Southampton University. He has 25 years experience of the
structural behaviour of reinforced and prestressed concrete and has published many
papers on the subject. He is currently leading Halcrows involvement in the
EUROCRETE programme, which is developing fibre composite reinforcement for
concrete.
Mr Paul ORegan is a Structural Engineer with Sir William Halcrow and Partners and
has 10 years design and construction experience of reinforced and prestressed concrete
structures. He is currently involved with the EUROCRETE programme, developing
design methods for non-ferrous reinforced concrete elements.
INTRODUCTION
FIBRE COMPOSITES
Fibres
Currently, the most suitable fibres for the reinforcement or prestressing of concrete are
glass, carbon or aramid. Each is a family of fibre types and not a particular one. They all
have high ultimate strengths (in the region of 3000N/mm2) and stiffnesses which range
from about 50kN/mm2 to over 200kN/mm2. They have a linear elastic response up to
ultimate load, with no significant yielding. The fibres are relatively difficult to handle and
anchor into concrete and hence, with the exception of one aramid prestessing system,
they have been combined with resins to form composite rods or grids, as described
below. Short, chopped fibres such as polypropylene, which would provide little if any
additional strength, are not considered in this paper.
Resins
There is a wide choice of resins available, many of which are suitable for forming
composites. The choice will depend on the required durability, the manufacturing process
and the cost. Thermosetting resins are generally used but thermoplastic resins are now
being developed.
Composite manufacture
The most widely used manufacturing process is pultrusion, which is used to produce a
wide range of structural shapes. The fibres, which may be all of one type, or a
combination of materials are supplied in the form of continuous rovings. They are drawn
off in a carefully controlled pattern through a resin bath which impregnates the fibre
bundle. They are then pulled through a die which consolidates the fibre-resin
combination and forms the required shape. The die is heated which sets and cures the
resin allowing the completed composite to be drawn off by suitable reciprocating clamps
or a tension device. The process enables a high proportion of fibres to be incorporated in
the cross-section and hence relatively high strength and stiffness are achieved. However,
the sections have a smooth surface, which provides insufficient bond for use in concrete.
Hence, a secondary process is required to improve the bond.
The properties of the composite will depend on the type and percentage of fibres used.
Typically with glass the ultimate strength might be 1200N/mm2 and the elastic modulus
40kN/mm2.
Designing structures using non-ferrous reinforcement 455
APPLICATIONS IN PRACTICE
Bridges
The widest use of fibre composites in concrete structures has been for prestressing. This
is probably because changing from steel to the new materials leads to little, if any, change
in the design process. Hence, a number of bridges have been built worldwide, generally
with conventional steel for the unstressed reinforcement. Examples include the
Ulenbergstrasse Bridge in Dusseldorf opened to traffic in 1986 [2], which, along with
other bridges in Germany and Austria, is prestressed with glass-fibre composite cables. In
Japan a number of bridges have been built [3], mainly prestressed with aramid or carbon
composites, while in the USA and Canada [4, 5] various materials have been used.
Tunnels
In Japan, fibre composites have been used for reinforcing the sprayed insitu lining of
several kilometres of tunnel [1]. Here the reinforcement, in the form of a square grid in
place of the usual welded steel fabric, is only required during the early life of the concrete
lining.
Other applications
Fibre composites have been used in the foundations of sensitive electrical equipment,
because of their non-conducting and non-magnetic properties. Other uses include marine
structures and storage tanks.
Introduction
Current design codes consist of a mixture of basic principles and simple rules of thumb.
The latter are based on one hundred or so years of experience of structures reinforced
with steel. The designer is given no guidance as to how such requirements should be
modified when changing to different reinforcement materials. In formulating new design
methods two approaches are possible. The first is to develop completely new methods
and the second is to adapt the existing ones. The former requires considerable
experimental data and would thus tend to delay the introduction of new materials. Hence
the EUROCRETE programme is developing modifications to existing design clauses,
justified by comparison with the limited data available. The resulting designs may not be
the most economic use of the new materials but they should result in safe structures.
Appropriate concrete technology 456
In modifying the design clauses, only those concerned directly or indirectly with the
reinforcement need to be changed. Those covering general principles and those
concerned with the geometry of the structure remain unaltered. The following sections
cover the more significant aspects.
Bending of beams
The basic principles of the behaviour of a beam in bending, such as plane sections
remaining plane, should not depend on the type of reinforcement material provided it has
sufficient bond. It is likely that failure will occur by compression of the concrete and not
by reaching the ultimate capacity of the tensile reinforcement.
A number of researchers have studied the behaviour of beams in bending including
Nakano et al [6] who used carbon or aramid composite bars, Faza & GangaRao [7,8],
Nawy & Neuwerth [9,10] and Masmoudi et al [11] who used glassfibre composites and
Maruyama et al [12] who used carbon fibre composites. Figure 1 shows the test results
plotted against the theoretical predictions of BS8110, using a partial safety factor of 1.0
for both the concrete and the reinforcement. It indicates that the design approach is
satisfactory: a factor of safety of 1.5, a reasonable value, would be required to give safe
design values.
Shear
In line with most design codes, BS8110 uses an empirical approach to determine the
shear capacity of beams. The total capacity is taken to be the sum of the capacity of the
concrete cross-section and that of the shear reinforcement. While this is not a true
representation of the behaviour, it has been proved to give an adequate margin of safety
for conventional steel reinforcement. Initially, the equations have been modified for
beams with non-ferrous reinforcement as follows:
Concrete cross-section: replace the area of non-ferrous reinforcement with the
equivalent area of steel, on the basis of the modulus of elasticity
Shear reinforcement: use a limiting strain of 0.0025 (c.f. a limiting stress of 460N/mm2
in BS8110).
Figure 2 shows the results from Nawy & Neuwerth [9], Maruyama et al [12] and from a
major series of tests by Nagasaka et al [13] and other unpublished data plotted against the
predicted values again with factors of safety of 1.0. Although the scatter of results is
greater than for the bending comparisons, as would be expected, the agreement is still
reasonable. An overall partial factor of safety of 1.5 would lead to safe design values.
Further study is required in this area.
CASE STUDIES
Trial Footbridge
EUROCRETE was invited to build the first footbridge in Britain, and possibly in Europe,
using glass fibre composite reinforcement [14] at Chalgrove in Oxfordshire. The bridge
slab, measuring 5.0m1.5m 0.30m, was precast using Grade 40 concrete by Tarmac
Precast. The bars, which were produced by GEC, were of 13.5mm diameter and were
used as mesh reinforcement with rods placed orthogonally at 150mm spacing on the top
and bottom surfaces. Prior to casting, a total of sixteen vibrating wire gauges were placed
throughout the slab such that monitoring of the inservice strains could be carried out.
After casting, fibre-optic cables were placed at soffit level over the central 2m of span
which also allow inservice behaviour to be measured.
The bridge slab was transported to site by Laing Civil Engineering and placed on mass
concrete abutments, using Neoprene pads as a bearing surface. Glass fibre composite
handrails were fixed to the slab once it was located in position. Load tests were then
carried out to 125% of the design load in accordance with BS8110 using a series of steel
deadweights. The results justified the design assumptions. In addition, some dynamic
tests were carried out. It is planned that these load tests will be repeated. Continuous
monitoring is being undertaken electronically by the recording of a full set of data once
each hour. The completed structure is shown in Figure 3.
Designing structures using non-ferrous reinforcement 459
Cribwall units
Cribwalls consist of small precast units which form a series of interlocking structures into
which suitable free draining granular materials are placed. The result is a near vertical
retaining wall. There is a growing interest in this form of construction for schemes such
as motorway widening. The units are reinforced chiefly for handling purposes; their
dimensions are dictated by the requirement for adequate cover to the steel.
Units from Tarmac Precast Concretes Kriblok system and from Phi Groups Anda-
Crib system were cast with glass fibre composite rods in place of the normal steel. They
were tested at Sheffield University along with conventionally reinforced units and proved
to be very satisfactory. It is planned that a trial length of wall will be built by Tarmac
Precast consisting of a vertical strip of units reinforced with glass fibre rods with
conventional units on either side.
Other structures
As part of the EUROCRETE programme, the construction of other trial structures using
non-ferrous reinforced concrete is planned. These include the following:
a 40m length of post and panel fencing required to have non-magnetic properties
rock anchor spreader plates, for use with a non-ferrous rock anchoring system
concrete cladding units
coastal protection units
a gravity fender unit in a marine location in the Middle East.
Appropriate concrete technology 460
CONCLUSIONS
The trial structures being built as part of the EUROCRETE programme are demonstrating
that fibre composite rods can offer a viable alternative to steel reinforcement in concrete.
Safe structures can be designed using modified versions of the current codes.
ACKNOWLEDGEMENTS
The work described in this paper was carried out as part of EUROCRETE, which is
partly funded by the Department of Trade and Industry (DTI)/EPSRC Link Scheme,
Structural Composites Programme. The Authors are grateful for this support and the
assistance of the other EUROCRETE Participants, (EuroProjects, Laing Technology,
Sheffield University, GEC Reinforced Plastics, ASW Construction Systems, DSM
Resins, Vetrotex International, Du Pont de Nemours International, SINTEF, Norsk Hydro
and Statoil) and for the support of Tarmac Precast Concrete Limited.
REFERENCES
1. SUGITA, M.NEFMAC Grid type reinforcement, Alternative Materials for the Reinforcement
and Prestressing of Concrete (Ed. J L Clarke), Blackie Academic and Professional, London
1993, Ch. 3.
2. WOLFF, R. and MEISSELER, H-J. Glass fibre prestressing system, Alternative Materials for the
Reinforcement and Prestressing of Concrete (Ed. J L Clarke), Blackie Academic and
Professional, London 1993, Ch. 6.
3. TSUJI, Y., KANDA, M. and TAMURA, T. Applications of FRP to prestressed concrete bridges
and other structures in Japan, PCI Journal, 1993, July-August, pp 5058.
4. IYER, S.L. Advanced Composite Demonstration Bridge Deck, Fiber-Reinforced-Plastic
Reinforcement for Concrete StructuresACI International Symposium (ACI SP-138), 1993, pp
831852.
5. ANON, Carbon-fibre strands prestress Calgary span, Engineering News Record, 1993, 18th
October, p21.
6. NAKANO, K., MATSUZAKI, Y., FUKUYAMA, H. and TESHIGAWARA, M. Flexural
Performance of Concrete Beams Reinforced with Continuous Fibre Bars, Fiber-Reinforced-
Plastic Reinforcement for Concrete StructuresACI International Symposium (ACI SP-138),
1993, pp 743766.
7. FAZA, S.S. and GANGARAO, H.V.S. Bending and Bond Behaviour of Concrete Beams
Reinforced with Fiber Reinforced Plastic Rebars, Transportation Research Record 1290, 1992,
pp185193.
8. FAZA, S.S. and GANGARAO, H.V.S. Theoretical and Experimental Correlation of Behaviour
of Concrete Beams Reinforced with Fiber Reinforced Plastic Rebars, Fiber-Reinforced-Plastic
Reinforcement for Concrete StructuresACI International Symposium (ACI SP-138), 1993, pp
599614.
9. NAWY, E.G. and NEUWERTH, G.E. Behaviour of fibre glass reinforced concrete beams,
Journal of the Structural Division, Proc. ASCE 1971 Vol. 97, Sept., pp 22032215.
10. NAWY, E.G. and NEUWERTH, G.E. Fiberglass Reinforced Concrete Slabs and Beams,
Journal of the Structural Division, Proc. ASCE 1977 Vol. 103 No. ST2, Feb., pp 421440.
Designing structures using non-ferrous reinforcement 461
INTRODUCTION
The current interest in non-ferrous reinforcement for use in concrete has occurred mainly
as the result of the extensive problems associated with the degradation of existing
buildings and structures resulting from the corrosion of steel rebar in adverse service
environments[1]. Design practice for enhancing the durability of reinforced structures
involves matching the assessed exposure with an appropriate amount of protection in
terms of concrete grade and cover to rebar [2], [3]. This often produces a concrete with a
performance potential in excess of that required for sustaining imposed service loads.
The use of fibre-reinforced polymers in concrete is not new[4] but improved
manufacturing techniques and modified morphology have enabled a more efficient use of
materials to produce elements which in strength and specific energy relative to
performance excel that for ferrous reinforcement [5], [6]. Furthermore the reinstatement
or strengthening of structures may provide another opportunity for the use of FRP in lieu
of steel [7], [8]. Although, as with steel, at elevated temperatures and during exposure to
fire, loss in structural integrity of FRP reinforced concrete elements could occur mainly
because of the susceptibility of the resin component in the rebar.
This paper presents results from the first stage of an investigation into the behaviour of
medium to high strength concrete beams reinforced with FRP and high tensile steel
rebars and examines their comparative behaviour in the context of models currently used
for conventional reinforced concrete design in the UK.
EXPERIMENTAL DETAILS
Materials
Concrete:Portland cement (42.5) with uncrushed fine aggregate(M sand) and coarse
aggregate (205mm gravel).
Reinforcement:Steel and FRP reinforcement details are contained in Table 1
Table 1 Specified Reinforcement Details
REBAR DENSITY STRENGTH ELASTIC BAR
TYPE MODULUS SIZES
kg/m3 MPa GPa mm
High Tensile 7,850 460 (Yield) 210 8, 12, 16
Steel
FRP 2,045 695 50 8.8, 13.2, 16.4
(Ultimate)
Appropriate concrete technology 464
Test Cubes
3 No. 100mm side test cubes were manufactured from each mix and cured adjacent to the
test beam in the laboratory. The results are in Table 3. column 7
Beams composition
One beam 103mm(breadth)203(depth)2.740mm(length) containing 2 No. steel or FRP
bars was manufactured from each mix; further details shown in Figure 1.
Models
In general these follow the basis of design for steel reinforced members in BS8110 with
minor variations and are summarised below with appropriate explanation (3).
Therefore the the ultimate moment(Mu) may be calculated from concrete cube
strength(fc) rebar strength(fy) and area of rebar(As). It should be noted however that when
applied to FRP where area of rebar is Af this model equation is likely to give an
underestimate of the moment capacity as there is minimal yield of this material at failure.
Failure mechanisms
Sketches of the two principal failure mechanisms are shown in Figure 1. Initially vertical
cracks emanating from the bottom surface within the middle section preceded failure in
all cases. In some cases these were followed by inclined cracks towards the supports. At
the ultimate condition beams failed in two alternate modes.
Flexure failure:resulted from yielding of the steel or sudden rupture of the FRP
followed by crushing of concrete in the compression zone within the mid-span section
The actual ultimate moment (Mu') is calculated on the basis of the maximum load at
failure(F in kN) its lever arm 0.770m and takes into account the self weight of the
beam(kN) and span(m) which approximates to:
Table 3 contains details of the beam references and experimental programme together
with rebar details in columns 14. Results from the experimental work and those from the
theoretical models are contained in columns 5 to 14.
Failure modes
All steel rebar beams failed in flexure except those with 16mm diameter rebars(TB9 &
TB11shear-bond mode) whereas FRP rebar beams failed in shear-bond with the
exception of C60 beam with 8.8mm diameter bars (TB8flexural rupture of bars) and
C40 with 8.77mm diameter bars (TB6simple shear). See column 5.
Failure Load
Failure loads suggest that the steel rebar beams were more sensitive to rebar area and less
sensitive to concrete grade than those reinforced by FRP, however cognisance must be
taken of failure mode in this respect. See column 6.
Performance Quotient
For flexural failure the Performance Quotient for steel and FRP reinforced beams (TB5 &
TB8) were comparable although not for the shear type failures. See column 14.
Load versus deflection behaviour for C40 concrete with 8mm steel rebars (TB5)
compared with C60 concrete with 8.8mm FRP(TB8) rebars are shown in Figure 2. This
suggests that although the FRP reinforced concrete displays a higher deflection at a
similar failure load, the toughness of the FRP beam (indicated by the area under the
Load vs Deflection curve) is much higher than that for the steel beam of a similar
performance quotient for flexural failures.
Table 3 Summary of beam properties and
performance
< Column Number >
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Beam Conc Reber Area Failure Fail fc NA Lever Mu' Mu vs vs( Perform.
Code rete No. Rebar Mode ure Cube Depth Arm Actual Theor Actual max) Quotient
grade Dia/ mm2 Load Str. mm mm kNm kNm MPa Theor Actual
Type kN MPa MPa
TB5 C40 208S 101 Flexure 24 61 12 173 9.64 8.01 0.57 0.93 0.01863
TB1 C40 212S 226 Flexure 50 55 31 164 19.65 17.08 1.20 1.21 0.0411
TB9 C40 216S 402 Shear- 68 61 49 156 26.58 28.85 1.63 1.47 0.04794
Bond
TB6 C40 2 121 Shear 20 69 20 20 169 8.10 14.20 0.98 0.1369
08.77F
TB2 C40 2 273 Shear- 34 57 54 154 13.49 29.23 0.81 1.29 0.02531
13.19F Bond
TB10 C40 2 425 Shear- 29 51 93 136 11.57 40.13 0.69 1.50 0.02216
16.44F Bond
TB7 C60 208S 101 Flexure 24 73 10 173 9.64 8.02 0.57 0.93 0.01574
TB3 C60 212S 226 Flexure 48 72 23 168 18.88 17.44 1.15 1.21 0.03044
TB3 C60 216S 402 Shear- 58 75 40 160 22.73 44.77 1.39 1.47 0.0339
Bond
TB11 C60 2 121 Flexure 26 72 19 170 10.41 14.24 0.62 0.98 0.01678
08.77F
TB4 C60 2 273 Shear- 38 79 39 161 15.03 30.49 0.91 1.29 0.02127
13.19F Bond
TB12 C60 2 425 Shear- 28 76 62 150 11.18 44.24 0.67 1.50 0.01543
16.44F Bond
Key
Steel rebar(S) FRP rebar(F)
A preliminary investigation into the comparative performance of FRP 469
Figue 3 shows a similar curve comparing 16mm steel rebars with 16.44mm FRP rebars
for C60 concrete (TB11 & TB12) and indicates that shear-bond failure tends to favour the
perfomance of the steel rebar beam compared with the FRP rebar beam.
The results suggest that performance of the FRP beams containing the higher areas of
tension reinforcement could be greatly enhanced by improving shear-bond capacity hence
utilising their maximum flexural strength capability. Certain properties of FRP should
enable this to be attained more economically than with steel rebar. It is proposed to
explore new design strategies in future stages of the investigation.
CONCLUSIONS
1. The mode of failure appears to be influenced by rebar type and area although concrete
grade was a confounding factor for FRP reinforced beams.
2. Conventional design models for steel reinforced concrete may require modification for
FRP rebar beams especially for shear failure.
3. For flexural failure FRP beams display a greater capacity to absorb energy than steel
for similar load capacity, although they exhibit reduced stiffness.
4. A performance quotient relating load capacity to section potential for similar
geometry may be a useful efficiency comparator.
5. The development of a design strategy for improving shear-bond performance without
significantly increasing rebar area could be applied to FRP beams to enhance their
performance.
ACKNOWLEDGMENTS
The authors wish to express their gratitude to International Composties and International
Gratings Ltd for the provision of materials and technical advice and information and also
to Sheffield Hallam University for provision of facilities and in particular the invaluable
assistance of the technical staff in the School of Construction.
REFERENCES
1. CLARKE J.L. Alternative Materials for the Reinforcement and Prestressing of Concrete, First
Edition 1993, 201pp.
2. BRITISH STANDARDS INSTITUTION, BS 5328:1991 Concrete. Guide to Specifying
Concrete. Part 1.
3 BRITISH STANDARDS INSTITUTION, BS 8110:1985 Structural Use of Concrete.Part 1.
4. NAWY, E.G. et al. Behaviour of Fiberglass Reinforced Concrete Beams, Journal of Structural
Division ASCE, September 1971, pp22032215.
5. ERKI, M.A. & RIZKALLA, S.H. FRP Reinforcement for Concrete Structures, Concrete
International, June 1993, pp4853.
6. BRITISH CEMENT ASSOCIATION.UK Cement Manufacture and the Environment. 13pp.
7. SAADATMANESH, H. Fiber Composites for New and Existing Structures, ACI Structural
Journal, V91, No 3, May-June 1994, pp346354.
A preliminary investigation into the comparative performance of FRP 471
8. LARRALDE, J. et al. Fiberglass Reinforced Plastic Rebars in Lieu of Steel Rebars, ASCE 7th
Annual Structures Congress, San Francisco, California, May 1989, pp18.
9 MARTIN, L.H. et al. Concrete Design to BS 8110,Edward Arnold,London 1989. 477PP
10. BROWN, V.L. & BARTHOLOMEW, C.L. FRP Reinforcing Bars in Reinforced Concrete
Members, ACI Materials Journal, V90, No 1, January-February 1993, pp3439.
STRUCTURAL BEHAVIOUR OF
BAMBOO REINFORCED CONCRETE
BEAMS AND SLABS
K Ghavami
University Cat. Rio
A J S Min
H C L Junior
N P Barbosa
University Fed. Paraiba
Brazil
INTRODUCTION
Renewable materials are having crescent interest to be used for humanity. In the field of
engineering, bamboo is a vegetal that presents vantages as fast growing and good
mechanical properties. A comparison with other natural and engineering materials
revealed a superior structural efficiency [1]. Since some years ago, bamboo as
engineering material has been studied in a research programme involving Catholic
University of Rio de Janeiro and Federal University of Paraiba, in Brazil. During this
time, some aspects were investigated: physical and mechanical properties of bamboo
species growing in Rio de Janeiro [2] and Paraiba [34], treatments against insects and
bamboo propagation [4], bamboo-concrete bond [5], bamboo special truss [6], bamboo
reinforced concrete elements [710]. This papers presents and discuss results from
experimental tests made in bamboo reinforced concrete beams and slabs.
EXPERIMENTAL DETAILS
Materials
Two kinds of concrete were employed: normal concrete with granite coarse aggregate
and lateritic concrete, with lateritic stones as coarse aggregate. The last presents lower
Young module [11].
Bamboos from species Dendrocalamus giganteus and Bambusa vulgaris were
utilised. Tables 1 and 2 indicates some properties of these bamboos obtained in Brazil.
Beams
Until now, eight 3 m span beams were tested. The first three beams were built in lateritic
concrete. The others in normal granite concrete. The beams, instrumented with electrical
and mechanical strain gages, are show in Figure 1. Different reinforcement tax were used,
trying to improve beams behaviour.
Table 1Mechanical characteristics of bamboo
used
Species Compression Tensile strength Flexion Youngs
strength (MPa) (MPa) strength (MPa) module (GPa)
Bambusa 65 115 131 9
vulgaris
Dendroc. 77 115 152 11
giganteus
Appropriate concrete technology 474
Bamboo-concrete bond
To study bamboo-concrete bond, some tests were made, as show in Figure 3. Despite the
push out test can give only an average value of bond stress, it serves to compare the
behaviour of bamboo rod treated in different way. Some treatments were tested in the
rods with and without node: hot bitumen, bitumen emulsion, oil paint, epoxy. Even a
steel wire was rolled up in some rod treated with bitumen.
Structural behaviour of bamboo reinforced concrete beams and slabs 475
Beams
Initial tests shown that bamboo reinforced concrete beams are more flexible than steel
ones. This is due, further on bond problems, to the lower Young module of bamboo. In
fact, it can be between 1/20 and 1/10 of steel modules. Then, although bamboo tensile
strength is near steel tensile strength, it is necessary to put great reinforcement tax in
bamboo beams. Experimentation shows ideal tax between 3 and 4 %.
Initial bamboo beams cracks presented large opening and were very spaced. A skin
reinforcement was then adopted. This improved behaviour but not sufficiently as
attended. The next step was improve bond using artificial connectors. Two kinds were
tested separately. The first, consisted on two little stones glued with epoxy in each
internode (Figure 2 c). The second, small 3 cm steel construction, also two per internode
(Figure 2 d).
Figure 6 shows curves load deflexion of concrete-bamboo beams in meddle span
where the two types of connectors were used. The load was applied in two cycles: first
until 20 kN e after until rupture of the beam. The beam with stone connectors presented
better performance. Concrete compression strength was near 25 MPa.
Figure 7 shows curves moment-bamboo strain and moment-concrete strain at the
centre of the beam where stones connectors were used. The concrete strain gage was
installed about 2 cm of the beam top.
If a deflexion equal to span/300 is considered as serviceability limit, then the beam
reaches this limit state at a load Fs=32 kN. To produce the same deflexion, it will be
necessary a distributed load 14.5 kN/m. Considering ultimate load (Fu=88 kN), it is
interesting to see that the relationship Fu/Fs=2.75. Comparing with a series of beams
tested, in lateritic and normal concrete, steel reinforced (b=10 cm; h=30 and 25 cm; L=3
m; As /b.h=0.64 to 0,96 %) this relationship was near 1, 81, 9 [14].
Structural behaviour of bamboo reinforced concrete beams and slabs 477
Bamboo-concrete bond
Figure 8 shows some curves average bond stress-displacement, obtained as indicated in
Figure 3. The treatment with products from petrol instead improves bond, it makes worse.
Many other curves were achieved (with normal and lateritic concrete) showing the same
fact. Even the procedures to roll up the bamboo rod with a wire, or pressing treated rod
against sand to give rugosity to the surface, are not very efficient. When epoxy was used,
it was impossible to push out the rod. But it was not used in beams because its cost. Other
curves taken with artificial connectors are been obtained.
Appropriate concrete technology 478
Slabs
Figure 9 shows load-deflexion, moment-concrete strain and moment bamboo strain for
the first slab tested. Load was applied in cycles.
For small load level, using Strength of Materials formulas it is possible to transform
concentrate loads in distributed loads that produce the same deflexion. Considering the
serviceability limit attained when meddle span displacement is equal to L/300, table 1
shows equivalent distributed load, bamboo strain and concrete strain at this displacement.
It is possible to see that even in the worse case (L3) the distributed load (6.7 kN/m2) is
biggest than normal accidental charge for common slabs (1, 5 to 2 kN/m2). Even
considering revetment load and creep effect for self weigh and revetment load, the
serviceability limit is not reached. At this load, Materials present still little strain.
The load employed (Figure 5) produces pure bending in the central third part of slab. In
all slabs, the rupture started by an inclined crack near the load, where there are normal
and shear stress (Figure 10). At this time, the slab is already very deformed.
Depending upon shear stress value, its possible to distinguish three phases when load
is increased: the fist, when load is small, bond for friction and adhesion guarantees that
the two material work together; the second, when shear stress at contact bamboo-concrete
area is greater than friction stress, the slab integrity is warranted by the diaphragms. As
they have some flexibility, they allow small relative displacements..
CONCLUDING REMARKS
The use of bamboo as beam reinforcement is feasible since some precautions are
considered. Skin reinforcement is necessary. Main reinforcement tax must be between 3,5
and 4,5 %. Bamboo treatments with petrol products are not advisable in Northeast of
Brazil, where temperature is always high. Artificial connectors increase load capacity of
beams and improve its behaviour.
Tests shows that bamboo permanent shutter slab is a promising structural element
where this vegetal is abundant.
Under serviceability loads, permanent shutter slabs have a behaviour more then
satisfactory. Even with 10 cm height, limit state is not reached for normal accidental
charges.
Under serviceability loads materials strain are small.
Maximal load is reached with large displacement that prevent its imminence.
There is a large safety factor between rupture and serviceability load.
It is possible to think about an artificial connector for the slabs when bamboo
internode is very long. Some laboratory experiences was already started.
At the Structural Laboratory of Federal University of Paraiba, a 10 cm height slab is
loaded with sand, giving a 2,5 kN/m2 load, since September 1993. The central deflexion
are been observed to have an idea about bamboo concrete slab under long time load.
Until now, creep displacement is about twice elastic one. The slab is in a very good
condition.
To large use of bamboo reinforced concrete structural elements, it is necessary to
study better how to guarantee durability of bamboo into concrete environment. Until
now, observations made in Brazil with bamboo immersed in concrete for 67 years show
that it is in good conditions.
In Brazil Northeast, the only insect that attacks bamboo is a Coleopteran Dinoderus
minutus. The treatment more effective against it consists to put bamboo culms or half
culms immersed in diesel oil.
Bamboo concrete structural elements are feasible to be used in rural constructions in
development countries where this vegetal is available.
ACKNOWLEDGEMENTS
The authors would like to express their acknowledgements to Prof. R.D.Toledo Filho,
Eng. Williams Guimares and the student Margareth Magalhes, from Federal University
of Paraiba. They are also grateful to CNPqBrazilian Research Found, for financial
support.
REFERENCES
1. WEGST, UGK, SHERCLIFF, H F AND ASHBY, MF. The structure and properties of bamboo
as an engineering material. University of Cambridge. Department of Engineering, UK.
Structural behaviour of bamboo reinforced concrete beams and slabs 481
2. CULSONI, RAM. Caractersticas dos bambus e sua utilizao como material alternative no
concrete. Master Thesis. Dept. of Civil Eng. Chatolic University of Rio de Janeiro, Brazil, 1986,
pp 143.
3. BARBOSA, N P. Vigas de concrete latertico reforadas com bambu. Proceedings of the 32th
Annual Meeting of Brazilian Concrete Institute, Santos, SP, Brazil, Aug. 1992.
4. FONSECA K S. Tcnicas de propagao, tratamento e propriedades fisicas e mecnicas do
bambu vulgaris. Report Depart. Agricultural Eng. Federal University of Paraiba, Campina
Grande, Brazil, Mar 1992.
5. LIMA JR, HC. Lajes de concrete latertico em frma permanente de bambu. Report Depart. Civil
Eng. Federal University of Paraiba, Campina Grande, Brazil, Mar 1992.
6. GHAVAMI, K AND MOREIRA, L E. Double-layer bamboo space structure. Space Structures 4,
vol I, Ed. Thomas Telford, Univ. of Surrey, England, 1993.
7. LIMA JR, HC, XAVIER, AC, BARBOSA, NP AND TOLEDO FILHO, R D. Vigas de concrete
reforadas com bamboo. First Eng. Civil Congess Federal University of Juiz de Fora, Brazil,
Aug. 1994, vol I, pp 2232.
8. XAVIER, A.C. Materiais alternatives na construo civil. Report Depart.Agricult. Eng. Federal
University of Paraiba, Campina Grande, Brazil, Aug. 1992.
9. BARBOSA, N P, TOLEDO FILHO, R D AND GHAVAMI, K. Comportamento de lajes de
concreto em frma permanente de bambu. XXVI Jornadas Sud Americanas de Ingenieria
Estructural, Montevideo, Uruguay, 1519 Nov. 1993.
10. LIMA JR., BARBOSA, N P AND GHAVAMI,K.Comportamento em servio de lajes de
concreto reforadas com bambu. V Wood Brazilian Congress, Belo Horizonte, Brazil, 1921 Jul
1995, vol II pp 441454.
11. BARBOSA, N P AND CHAGAS FILHO, M B. Recherche sur beton avecgranulat lateritique.
Concrete in Hot Climates. Proceedings of the Third Intern. RILEM Conference, Torquay,
England, 2125 sept. 1992, pp 4352.
12. GHAVAMI, K. Application of bamboo as a low cost construction material. International
Bamboo Workshop, Cochin, Inde, 1988, pp 115.
13. GHAVAMI, K. Application of bamboo as a low-cost enrgy material in civil engineering. Third
CIB-RILEM Simposium material for low income housing. Mexico city, 610 nov. 1989, pp
528536.
14. BARBOSA, N P. O concreto latertico e seu emprego em vigas estruturais. reforadas com
bambu. Proceedings of the 30th Annual Meeting of Brazilian Concrete Institute, Santos, SP,
Brazil, Aug. 1992.
Theme 6
DESIGN AND
CONSTRUCTION
Chairmen Professor H C Chan
Hong Kong University
Hong Kong
Professor B Teply
Technical University of Brno
Czech Republic
Professor P R Vassie
Transport Research Laboratory
United Kingdom
Leader Paper
Design and Construction
Professor T P Tassios
National Technical University of Athens
Greece
DESIGN AND CONSTRUCTION
T P Tassios
National Technical University of Athens
Greece
INTRODUCTION
For centuries, the structural engineering process was an entire and uninterrupted entity:
The Engineer (one and the same person, the Master-Builder) was responsible for the
conception the plans (very general though) and the execution; Design and Construction
stages were not yet separated. Besides, most of the design was effectuated during
construction, following the method build and see (or rebuild). Consequently, omission
of instructions, and erroneous information from the designer to the contractor did not
make sense; similarly, all feeding-back information (regarding actual conditions in the
job site) was spontaneously collected by the MasterBuilder. Thus, no leakage of
responsibilities could ever be observed at the interface.
Appropriate concrete technology 484
Since the entire production process was (more or less artificially) splitted into a
Design and a Construction stage, we tend to think that these stages are in fact
separated and that only a one way influence remains, i.e. from design to construction.
Everyday experience, however, shows that the structural process is always a real
continuum; two ways interactions are manifested within it. And it is expected that the
entire process will be optimised if all these interactions are carefully identified, sincerely
recognised and facilitated. If not, some characteristic failures may (rather frequently now)
remind us that we cannot disregard the nature of things: A designer proud of his FEM
package is still responsible for
the unbuildability of crucial column-beam joints (because of congested reinforcement),
the cracking of unaccessible slabs, since intensive shrinkage would not be prevented by
means of (expected?) curing, and
the inadequate level of prestressing of long tendons, because of the absence of
appropriate specifications regarding relevant frictionreducing techniques.
For each of these failures, some designers could answer since my analysis is correct, all
this is not my responsibility, the Contractor should have thought of it. But good
designers, keeping the spirit of the ancient Master Builders and recognising the
continuum of the structural process, could have prevented these failures by means of an
appropriate re-dimensioning, additional skin-reinforcement or specific recommendations,
respectively.
Nevertheless, in a total quality environment, early identifications of such potential
risks should be a must for the Contractor as well; he should also be inspired from his
ancestors the Master Builders, and act accordingly towards an upstream care taking.
This is the kind of issues this paper intends to analyse. In Engineering it is not a good
intention or an elegant instrument (be it a mathematical solution or a piece of
sophisticated construction equipment) that really counts: The final engineering product
only matters; and what we are going to discuss here, may equally contribute to the
targeted serviceability and safety of this final product the same way as any advanced
analysis may do.
More specifically, this paper will address the Design/Construction interface problems
appearing in Concrete Structures; maintenance will also be considered.
And in doing so, we will bear in mind the needs of young Engineers; some of them,
under the fashionable land slide of computerised techniques, may tend to forget the
amplitude of our professional duties.
CONCEPTUAL DESIGN
The design process starts always with a conceptual stage; To serve a desired purpose, the
Engineer conceives a first (be it rough) structural scheme; experience and imagination
help him in this creative first step. Most of the characteristics of the final structure are
already (and apriori) decided at this stage! Analytical verifications which follow, may
easily approve different solutions although some of them may not be feasible (i.e. their
doubtful construction or their impossible maintenance may eventually jeopardise basic
performances).
Design and construction 485
Technical Availability
There is an understandable technocratic tendency in every Specialist; design Engineers
could not make exception: Each new design-case may be an opportunity for a modern
and impressive achievement, which may publicise the capacities of the designer (and
make him rich, in glory at least). Independently of any moralistic criticism such an
attitude should receive, it is clear that modernity and impressiveness will not
necessarily optimise the final performance of a structure:
Available materials
The weaponry of concrete technology became quite rich nowadays: Higher performance
concretes, polymer modified concrete, fibre reinforced concrete, shotcretes, etc., offer a
variety of solutions to serve specific requirements. Similarly, considerable development
in reinforcements is made: Completely weldable steel, corrosion protected bars, smart
splicing devices, external prestressing, versatile steel-concrete composites, or
industrialised fixing elements. Without mentioning the various protective materials (from
special grouts and paintings, up to curing compounds), it is clear that a designer keen to
offer the most appropriate solution, should be familiar with these developments and make
a more imaginative selection than just reinforced concrete of a given class of strength
(only!). To this end, modern designers should be educated (intraor extramural
education) in fields broader than mathematical modeling alone. Once again, a good
designer should be an acceptable Material Technologist.
Anyway, it will be very much appreciated if the design of a public building in a region
poor in strong aggregates, would not be based on a high strength prestressed concrete
Appropriate concrete technology 486
Economy in Construction
The obvious basic requirement of economy is supposed to inspire even conceptual
design.
However, economic solutions can only be proposed if the principal designer has a
thorough knowledge of construction techniques and their real costs rather than just unit
prices. As an example, the issue Factory vs. Site production may be irreversibly
influenced if an initial conceptual design was in-situ minded; one cannot successfully
translate a monolithic structural concept into a precast one.
ANALYTICAL DESIGN
Concrete Quality
All the adjectives we use to describe our concrete structures (plain, reinforced,
prestressed, precast, projected etc.) cannot reduce the fundamental importance of
concrete itself. For the benefit of their final products, designers should understand about
concrete something more than just tabulated values of its mechanical properties.
Compressive strength
It is rather instructive to recall here the two basic handicaps of the very concept of
characteristic compressive strength, fck, of concrete:
First, we are all bound to the probabilities accepting a defective lot because of the
economically imposed O.C. lines (related to the compliance criteria).
Appropriate concrete technology 488
Second, all we know about that strength comes from laboratory made and cured
specimens; consequences of transportation, placing, compacting and curing in-situ are not
measured! (Sometimes, cores taken from the bottom of a column and from the top of a
slab, both well concreted and cured, may legitimately differ by more than 50% in
strength).
The combination of these two handicaps of the concept of characteristic strength
should at least remind us what an interest we have as professionals to give further
emphasis to the very details of the construction stage, on which a considerably larger part
of the real safety is depending upon.
Besides, one of the best ways to secure low creep effects is to postpone striking of
falsework, as it should be clearly requested by the design documents, after appropriate
calculations if needed.
Concreting programme
There is no excuse for omitting a thermal analysis (in favour of, say, a dynamic analysis).
Medium massive building-elements under adverse ambient conditions may be drastically
cracked after concreting, because as designers we may have decided that this is a
construction problem, although the selection of concrete pour areas and their sequence
is nowadays the subject of intelligent softwares.
Similarly, uncracked R.C. walls are rare; hygric and thermal shrinkage hindered at the
foundation level, is a construction problem inviting for drastic design provisions.
Curing
This is a decisive construction operation which, despite its numerous structural
consequences, is not remunerated separately. Its importance in design is shortly reminded
hereafter.
The disproportionate scattering of shrinkage values (of real concretes in-situ) should
be beared in mind of every designer; otherwise, the deterministic calculation of an
acceptable distance between consecutive permanent joints may lead to unpleasant
surprises of unexpected cracking. These phenomena may be further accentuated, through
the decades, because of an ill-defined tensional fatigue of concrete under cyclic
hygrothermal conditions.
Design and construction 489
Moreover, the significance of curing on skincrete (the ten to thirty millimeters external
layer of any concrete-element) cannot be overemphasised. For small height building
elements, at the U.L.S., skincrete is the most important part of the compression zone; and
its strength is highly dependent (Fig.1) on curing. More generally, however, this concrete
layer (covering reinforcement and tendons) determines durabi lity in a decisive way:
Permeation properties are highly dependent on curing (Fig.2). By way of consequence,
designers have an interest to rely on good curing rather than on temporary coatings (or
additional safety factors).
In conclusion, against all the aforementioned undesirable structural effects, curing
should be given an important place in design documents: Detailed and demanding
specifications on curing (including specific quality control and separate unit prices)
proved to be an efficient means to this end.
Detailing
Thanks to that part of experience (or chapter of Codes) called Detailing, a lot of
verifications are not carried out by calculation; the relevant requirements are covered via
practical rules. However, this simplification does not reduce the importance of the
respective checks: Most of them are meant to insure safety, e.g.
Of the steel bar itself: Bending should be made with the aid of appropriate mandrels,
or, Welding should not be made near bends of bars.
Of the anchorage and the durability of bars: A specified minimum concretecover
should be provided.
Design and construction 491
Finally, several other detailing rules are intended to cover the inevitable uncertainties
which otherwise would have undermined the applicability of calculation models (e.g.
minimal thicknesses of concrete sections, maximum steel ratios, etc.).
This short reminding of the basic importance of this part of Design, may be useful per
se; but it may also help to further harmonise design-versus-construction procedures: A
designer deing conscious of the fundamental importance of these rules (secondary rules
are called by some non-engineering minded people), will find ways to insure their correct
application during the construction stage. To this end, these are at least two things to be
reminded:
First, make your drawings as attractive and friendly to the Contractor as possible: the
actual weaponry of C.A.D. is very helpful in this respect. Frequent didactic reminders
illustrating important construction principles should be made on these drawings; do not
feel you will offend any body if you repeat well known things, in a well illustrated way
however. Second, establish a special category of checking of design documents, of
buildability that is. The main problem in this respect is to identify areas:
of congested reinforcement, requiring modifications in detailing, or an explicit
warning about the way you envisage to overcome the local difficulty in concreting,
and
of impossible reinforcement, i.e. a geometry of steel bars which cannot be
implemented because other bars occupy the same space; three dimensional coloured
representations of such difficult areas (e.g. beam-column joints) may be used for such
checkings, using special softwares (Fig.4) and (Fig.5) which may automatically make
the necessary modifications of detailing.
One of the most controversial issues in Civil Engineering is the naive view that a
designer in this field of industry must know everything in advance, to such an extent
that a complete and precise budget and time schedule would be prepared and fully
respected. Due, however, to the very nature of this peculiar industry (open air and
nomadic factory, strong dependence on environmental data, no serial-production, non-
certified personnel, etc.), such a complete and precise knowledge in advance, in most
cases, is simply impossible: E.g. extensive soil investigation, adapting job site operations
to medium range meteorological forecasts, exclusive use of certified personnel for every
construction step, and the like, despite their socially unacceptable costs, would never
suffice to overcome the inherent peculiarities of this industry.
Design and construction 493
REFERENCES
INTRODUCTION
The approach almost universally used for the design of reinforced concrete framed
structures was originally developed when design methods effectively assumed that
reinforced concrete was an elastic-brittle material. As our understanding of reinforced
concrete has developed, the design approach has been somewhat modified, but not really
Appropriate concrete technology 496
changed. Is this approach still the most appropriate in the modern age or would
something different be more appropriate? This paper examines some aspects of this
question and concludes that changes could, with advantage be made.
The approach originally developed for analysis and design can be summarized as
follows:
(i) Elastic analyses are carried out for various arrangements of the loads to give a
bending moment envelope. In most countries the arrangements used for obtaining the
bending moment envelope were, and still are:
(a) alternate spans carrying the maximum load while the other spans carry the
minimum load.
(b) each pair of adjacent spans considered in turn carrying the maximum load while
the remainder carry their minimum load.
(ii) Critical sections are detailed to provide the required strength using permissible
stresses obtained by dividing the specified material strengths by appropriate safety
factors.
The realization that reinforced concrete was not elastic but had a degree of ductility
led to the introduction of redistribution (initially commonly up to 15%). The development
of the Limit State approach led to the use of design ultimate loads and material
properties. These were obtained by applying partial safety factors to the loads and
material strengths. Further research led to increases in the permitted amount of
redistribution. A more recent change in BS8110 [1] in the UK has resulted in some
simplification of the load arrangements by substituting a single load case of maximum
load on all spans for the adjacent spans loaded case.
It will be seen, however, that the changes have been incremental. The realization that
reinforced concrete can be designed to be effectively plastic has not led to a full rethink
of the logic of the design process.
Two elements of the design process will be considered in this paper: the employment
of load patterns for the ultimate load analysis and the probabilistic consequences of
continuity. In both cases it will be suggested that current methods are unnecessarily
conservative provided adequate ductility it available. A further matter that will be
discussed is the relevance of current methods of frame analysis to the actual behaviour of
frames at ultimate. It will be suggested that th the relevance is limited as inelastic
behaviour leads to major differences from current analyses.
Assuming that all sections of a beam have adequate ductility, failure of a reinforced
concrete beam will be the result of a mechanism forming. This mechanism will consist of
elements of beam connected by plastic hinges. When a beam is designed, reinforcement
is provided to match as nearly as is practical the strength required at the critical sections.
Usually the sections where the design strength most closely matches the required strength
will be over the supports and near mid-span but, where reinforcement is curtailed, there
may also be other points where the designed strength approaches the required strength. At
The design of framed structures-A proposal for change 497
all other points the beam will have a bending capacity greater than required by the
bending moment envelope. This limits the positions where hinges are likely to form and
hence limits the possible failure mechanisms that need consideration.
When a continuous beam is analysed using the classical load arrangements set out above,
the bending moment diagram for each load arrangement is calculated and the results
superimposed to give a bending moment envelope. Such an envelope is shown
schematically in Figure 1(a). Figure 1(b) shows the moment capacity likely to be
Appropriate concrete technology 498
provided by the reinforcement and the possible locations for hinges. It will be found that
two mechanisms are practically feasible: a mechanism with sagging and hogging hinges
in alternate spans and a mechanism having a hogging hinge at two adjacent supports and
a sagging hinge near mid-span. These are shown in Figures 1(c) and 1(d). It can easily be
shown [2] that the critical mechanism is that shown in Figure 1(c) where all hinges form
in regions of the beam where the design moment is defined by the alternate spans
loaded load arrangement. The support moment, which is defined in the analysis by the
adjacent spans loaded arrangement, would have to be reduced to that required for the
alternate spans loaded case before the mechanism shown in Figure 1(d) became critical.
Thus, provided a ductile failure can be guaranteed, it should only be necessary to carry
out analyses for the alternate spans loaded arrangement. Further simplification can be
achieved by considering the practicality of the mechanism shown in Figure 1(c). In
monolithic framed structures, the supports do not permit free rotation but are monolithic
connections with columns and the failure mechanism can only develop if the beam-
column joint can rotate sufficiently. Furthermore, current detailing rules mean that the top
reinforcement is not curtailed till well beyond where it is needed and hence a hinge at the
curtailment point of the top steel will have a strength well in excess of that required by
the bending moment diagram. It thus seems that a mechanism requiring hinges in more
than one span is most unlikely to occur. If so then the only practically likely failure
mechanism is the three hinge mechanism shown in Figure 1(d). To ensure that this
mechanism has adequate strength it is only necessary to ensure that, for beams supporting
uniformly distributed loads, the support and span moments satisfy the relation:
[1]
i and j are respectively the ratios of the left-hand and right-hand support ultimate
moments to the ultimate span moment, Mu.span.
This can conveniently be achieved by the simple expedient of analysing the beam for
the single load arrangement of maximum design load on all spans.
Thus, provided adequate ductility can be ensured, design for ultimate in flexure does
not require analysis for multiple load arrangements and the construction of a bending
moment envelope. The effect of this is a significant reduction in reinforcement quantities.
PROBABILISTIC CONSIDERATIONS
It has been argued above that the only failure mechanisms which needs to be considered
are those occurring within one span (the three hinged failure). A statically determinate
structure only needs to reach its ultimate capacity at one section to fail. From a
probabilistic point of view, it will be seen that a continuous structure will have a lower
probability of failure than a simply supported beam because the probability of three
hinges being simultaneously understrength must be smaller than the probability of a
The design of framed structures-A proposal for change 499
single understrength section. If, therefore, the level of safety currently required for simply
supported beams is considered adequate, then the safety of continuous structures must
currently be excessive and a lower safety factor should be admissible. A simplified
analysis can be carried out to illustrate this.
For a continuous beam with equal support moments to carry the same load as a simply
supported beam with the same dimensions,
Mu.ss=Mu.sup+Mu.span=wul2/8
[2]
where Mu.ss= the ultimate moment capacity of the mid-span section of the simply supported beam.
Mu.sup= the ultimate moment capacity of the supports of the continuous beam
For typical values of the ratio of support to span moment, it can easily be shown from
Equations 2 and 3 that the ratio of the standard deviation of the capacity of the continuous
beam to that of the simply supported beam is about 0.6.
Assuming that the distributions of strength are normal then, for a constant risk of
failure, Eurocode 1 Part 1 [3] states that the partial factor may be chosen such that a
constant value is obtained for the safety index where:
=(wum.wud)/w
in which
wum = mean ultimate load capacity
wud = design ultimate load capacity
w = standard deviation of ultimate strength.
Using the normal definitions for characteristic and design strengths and assuming that the
load capacity is proportional to the material strength, it can be shown that the required
partial safety factor for a continuous beam is given by:
Insertion of reasonable figures into this formula suggest that a reduction in safety factors
of up to about 15% might be reasonable.
A more detailed study of this issue was carried out by Beeby for the Reinforced
Concrete Council (unpublished). This took account of the variabilities of steel and
concrete and of the section geometry. The effects of correlations between the properties
at the three critical sections was also considered. This suggested that, if correlations
between the three critical sections are ignored then a reduction in safety factors of about
15% would be reasonable. Taking into account the maximum likely degree of correlation
leads to roughly a halving of this reduction. Combined with the reductions in steel area
resulting from discarding multiple load arrangements, an overall saving in flexural
reinforcement of up to 30% seemed possible without reducing the safety below that
accepted for determinate structures.
OTHER FACTORS
There are a number of other issues being discussed nationally and internationally which,
if accepted would lead to a further reduction in the quantity of reinforcement used in
frames. The following lists the main areas of discussion.
(i) reduction in the partial safety factor on reinforcement from 1.15 to 1.05. The
British Standards Institute (BSI) has recently published a draft amendment to BS8110
proposing this. Its effect would be to reduce the areas of main flexural steel by around
8%.
(ii) reduction of the partial safety factor on dead loads. There is considerable
discussion within the Eurocode drafting committees of the possibility of reducing the
partial factor on dead load from 1.35 to possibly 1.15. The argument for this is that the
dead load can be assessed with greater accuracy than is indicated by a factor of 1.35. The
pressure for a lowering of safety factors in the Eurocodes arises from interaction with the
old Eastern Bloc countries which use, and have used for many years, considerably lower
overall levels of safety than those used in the West. They claim that these lower safety
levels have not caused any problems.
(iii) increase in the specified characteristic strength of reinforcement from 460 N/mm2
to 500 N/mm2. The European prestandard for reinforcement [ENV 10080] only
recognises 500 grade reinforcement and it seems likely that this will be accepted. This
does not directly affect safety but does have an effect on the stress in the reinforcement
under serviceability loads.
DISCUSSION
The acceptance of any, or all the proposals discussed above lead to a reduction in the
areas of flexural steel employed in reinforced concrete frames. This, in turn, leads to an
increase in the stress levels in the reinforcement in service which leads to an increase in
service deflections and crack widths. Also critical to the proper performance of frames in
service is the requirement that the reinforcement shall behave elastically under service
conditions. Any inelastic behaviour will lead to the development of large permanently
The design of framed structures-A proposal for change 501
open cracks which will generally be unacceptable. Both BS8110 [1] and Eurocode 2 [4]
limit the stress in the reinforcement under service load to 0.8fy to avoid this possibility.
At present it is not generally necessary to check for this since the safety factors under
ultimate conditions and the limits on redistribution generally ensure that the limit is not
exceeded. Unfortunately, this is not the case if the safety factors are reduced to the extent
which seems possible for framed structures and the check on the service stress becomes
the controlling factor in the design, (i.e. the area of flexural steel is governed by the
limitation of the stress under service load to 0.8fy and not by the ultimate flexural
strength.)
It has been recognised for very many years that, if an under-reinforced member is
restrained against lateral expansion, its strength can be substantially enhanced due to the
development of in-plane forces which effectively prestress the member [eg 5]. The forces
develop because when a reinforced beam cracks and, more especially, when yield
develops at any section, the beam, if unrestrained, will expand. If this expansion is
inhibited by restraint at the supports then an in-plane force will develop which will be a
function of the difference between the expansion which would occur in an unrestrained
beam and that which occurs in the restrained beam. In slabs where restraint to expansion
may be provided either by other surrounding slab panels or by stiff elements in the
structure such as core walls, the increase in strength resulting from restraint can be very
large and has been the subject of extensive research. It has not generally been
incorporated into design codes because of the difficulties in defining the restraint
conditions, the enhancement in strength of beams in framed structures could be expected
to be generally small as restraint is commonly only provided by fairly flexible columns.
The possibility of using this strength enhancement in design has never seriously been
considered. What does not appear to have been recognised is that, regardless of whether
account is taken of the possible increase in capacity of the beams or not, if lateral forces
develop then they have to be supported by the columns. The columns must also be
capable of sustaining the deflections resulting from expansion of the beams. An extensive
research programme is currently in progress in the Civil Engineering Department at
Leeds on the actual ultimate behaviour of framed structures. The results so-far from this
project illustrate very clearly that the actual conditions under ultimate conditions in a
frame are very different from those which current analysis methods would predict. An
outline of the test programme and a summary of the results so-far are given below.
EXPERIMENTAL PROGRAMME
The test specimens aim to model conditions in the end columns and end span of a
rectangular reinforced concrete frame (see Figure 2). The specimens are designed to be
roughly half scale models of a practical frame. The overall height of the column between
points of contraflexure was chosen as 3.6 m and the length of the beam (corresponding to
approximately 2/3 of the span) was 2.2 m. It is believed that, at this scale, size effects
Appropriate concrete technology 502
should not be significant. The test rig is shown in Figure 3. The objective is to be able to
load the beam and also provide an axial load in the column while permitting any axial
load in the beam which develops to be measured. Considerable difficulties were
encountered in arriving at a set up which would permit reliable assessment of the axial
loads but this was eventually achieved.
Results will be presented for a series of four tests where the beam and column breadth
was 150mm and the overall depth of both members was 250mm. The axial load in the
column and the reinforcement ratio in the beam were the principal variables investigated.
Table 1 shows the principal results.
It will be seen that the capacity of the beams have been increased somewhat, with greater
increases with lower reinforcement ratios, and that significant membrane forces have
developed. The forced developed have been between 1 and 2 N/mm2 at ultimate. The
most significant effect on behaviour is, however, the bending moments imposed on the
columns. This is illustrated in Figure 4 which compares the bending moment diagrams
obtained from analyses ignoring the membrane forces with those which were measured at
maximum load on the beam.
It will be seen that the actual column moments may be up to seven times the calculated
values if membrane forces are ignored. Specimens 1 to 3, which had low axial loads on
Appropriate concrete technology 504
the column, all failed finally by failure of the beam. Specimen 4, and other tests where
the higher column load has been applied, failed in the column. Thus, not only does the
normal method of analysis not predict the moments which occur in the frame near failure,
it may also not correctly predict the mode of failure.
Other Investigations, such as the tests by Lahlouh and Waldron [6] show similar
results.
It has been found that more rigorous, non-linear analyses using finite elements
(ABAQUS) can predict the behaviour of the frames with reasonable accuracy provided
allowance is introduced into the analysis for the development of bond failure in the hinge
regions.
CONCLUDING DISCUSSION
It has been argued that there are a number of changes that could logically be made to the
design of reinforced concrete framed structures for flexure which could result in
significant economies and that there are other developments nationally and
internationally which could lead to further reductions in the areas of flexural
reinforcement. The study carried out for the Reinforced Concrete Council shows that any
significant reduction in the area of flexural reinforcement will lead to the requirement
that there is no yield of the reinforcement under service conditions being the critical
design condition rather than ultimate strength. The research on membrane effects in
frames shows that the normal type of analysis used for framed structures, which ignores
membrane effects, leads to some conservatism in the estimation of the strength of the
beams but a major underestimation of the moments which may be imposed on the
columns. This underestimate is sufficiently large that, while the design method assumes
that failure will occur in a ductile manner in the beams, there are likely to be cases where
the actual failure will be a brittle failure of the column. The present method of analysis is
thus misleading at the ultimate limit state.
It seems arguable, therefore, that the current analysis for the ultimate limit state is not
particularly helpful and that there might be advantages in carrying out design for the
serviceability limit state instead. The proposal is the following for beams:
(a) carry out the analysis of the structure under the service loads using the classical
load arrangements of alternate spans loaded and adjacent spans loaded.
(b) calculate the required reinforcement using an elastic, modular ratio method with a
limit of 0.8fy on the stress in the reinforcement.
(c) check cracking, deflections and any other relevant service conditions using the
service steel stress.
(d) check shear. In principle, this check should be done under ultimate load conditions
but it is likely to be easier to recast the shear check in service load terms.
The advantages in this approach are:
(a) a significant saving in flexural reinforcement quantities. The Reinforced Concrete
Campaign study suggests that this may be about 25%.
(b) by concentrating on serviceability, design attention is concentrated on aspects of
performance which actually concern the client and the user such as excessive deflection
The design of framed structures-A proposal for change 505
and which can be measured rather than the almost entirely hypothetical conditions that
might occur near failure.
(c) use of an elastic analysis will ensure that the requirements for ductility at critical
sections at ultimate is a minimum and hence no special measures beyond some overall
limit on neutral axis depth will be necessary to ensure that adequate ductility is provided.
The disadvantage, which may be fatal to the suggestion, is that it will require a major
revision to current design procedures. This will be resisted fiercely by designers.
CONCLUSIONS
(1) From studies of the behaviour of beams it is concluded that, at ultimate, there is no
justification for the consideration of the current load arrangements in the analysis.
(2) Consideration of the probability of failure of continuous beams relative to that of
determinate members show that the safety factors for continuous members could be
significantly reduced.
(3) If (1) and (2) above were implemented then serviceability considerations would be
found to control the quantity of flexural reinforcement rather than ultimate.
(4) Consideration is currently being given to reducing the partial safety factor on
reinforcement strength, reducing the partial safety factor on dead loads and increasing the
specified strength of reinforcement. Each of these, if adopted would further increase the
criticality of serviceability conditions and reduce the criticality of strength considerations.
(5) Research shows that the current methods of analysis for the ultimate limit state,
which ignore the influence of membrane forces, can provide very misleading predictions
of behaviour.
(6) Actual problems with structural behaviour are practically always serviceability
problems rather than safety.
(7) Taking all the above into account suggests that design might more sensibly be
carried out for the service condition rather than ultimate.
REFERENCES
1. BS8110:1985; Code of practice for the structural use of concrete. British Standards Institution,
London 1985.
2. Beeby, A.W. The analysis of beams in plane frames according to CP110. Development Report
No. 1, The Cement and Concrete Association, 1978.
3. ENV 19911 Eurocode 1, Basis of design and actions on structures, Part 1; Basis of design. CEN
1994.
4. DD ENV 199211:1992, Eurocode 2, Design of concrete structures, Part 1: General rules and
rules for buildings. British Standards Institution, London 1992.
5. Braestrup, M. Dome effect in RC slabs: rigid-plastic analysis. Journal of the Structures Division,
Proceedings of ASCE 106 No. ST6 June 1980.
6. Lahlouh, E-H. and Waldron, P. Membrane action in one-way slab strips. Proceedings of the
Institution of Civil Engineers, Structures and Buildings November 1992.
DEVELOPMENT OF MOMENT
CONTINUITY CONNECTION FOR
SIMPLY SUPPORTED PRECAST
CONCRETE BRIDGES
G Davies
K S Elliott
A A H Arshad
University of Nottingham
UK
INTRODUCTION
There are many advantages in making short or intermediate span bridge decks simply
supported during the construction phase. These include rapidity of construction and
economy. However it is increasingly recognised that maintenance problems are
associated with extensive below-deck deterioration due to ingress of drainage water
through expansion joints, particularly when laden with de-icing salt. Additionally smooth
jointless construction improves vehicular riding quality and diminishes vehicular impact
stress levels. There is therefore considerable interest in reducing the number of expensive
expansion joints, producing continuous deck construction or integral bridges. The basic
building block is still the pre-tensioned beam, and consideration has been given to how
continuity can best be accommodated practically, to ensure that both the hogging and
sagging moments can be adequately designed for the limit states of strength and
serviceability. There is interest in maintaining a competitive market edge for future
bridges, as well as the upgrading of existing bridges. Sagging moments at the supports
may be due to loading, settlement, creep, shrinkage, and temperature effects. The paper
reviews existing methods, indicating advantages and disadvantages, and proposes an
alternative method which is appropriate for both old and new structures. In particular the
paper describes preliminary tests carried out to establish sagging moment capacity and
the moment transfer mechanism for such connections.
The following Types of construction shown in Figure 1 have already been used:
1. Wide Insitu Integral Crosshead. 4. Continuous Separated Deck Slab
2. Narrow Insitu Integral Crosshead. 5. Tied Deck Slab
3. Integral Crosshead Cast in Two Stages. 6. Prestressed Deck Continuous Over the Pier
Limitations/Problems
Method Limitations/Problems
Type 1. Requirement of temporary supports.
1
2. Projection of bottom continuity reinforcement beyond the ends of precast beams
resulting in
a. Congestion of reinforcement at the beam ends
b. Reduces the benefit of standardisation
c. Health and Safety hazards during transportation and erection
d. Difficulty in launching of beams
Appropriate concrete technology 508
Type 1. Spaces between beam ends are very narrow which cause difficulty in placing high
2 quality concrete.
2. Projection of bottom continuity reinforcement at the ends of precast beams causes:
a. d. as for Type 1.
Retrofit Bridges
None of the existing methods are suitable to convert existing simply supported span
bridges into continuous/integral bridges.
Description:
Continuity over the pier is confined to the diaphragms and deck slab.
Negative moment: reinforced concrete deck slab and precast beam.
Positive moment: reinforced insitu concrete diaphragm.
Differences with existing methods. The connection is similar to Type 2 except that:
There is no bottom continuity reinforcement at the ends of precast beams.
The shape of the insitu reinforced concrete diaphragm is trapezoidal so the change of
curvature between the two different cross-sections is smoothed.
The reinforced concrete diaphragms are designed for positive moment.
Advantages:
No continuity reinforcement at the ends of the precast beamsproblem of congestion,
health and safety hazards avoided.
No temporary work is required.
Minimal additional formwork and ease of construction.
Can be adopted for converting existing bridges.
Suitable for most bridges consisting of beams with bottom flanges side-by-side such a
T-beams, inverted T-beams and M-beams.
Appropriate concrete technology 510
Introduction
It is quite difficult to simulate a rigorous test for sagging moment resistance at a support
in a continuous beam, particularly if primarily caused by temperature gradients or
shrinkage. However the essential requirement is the ability to transfer the beam span
moments through the interfaces of the insitu support connection satisfactorily. Since the
real life situation could not be completely idealised it was decided to make initial
comparisons on the basis of simply supported composite beams under four point sagging
momentFigure 3.
recorded from transfer stage to failure. The cube strengths of diaphragm and deck slab
concrete and beams were 58 N/mm2 and from 44 to 54 N/mm2, respectively.
Test Procedure
The beams were supported temporarily at the mid-span position and end until the in-situ
infill concrete had hardened. The general test arrangement for Initial and M-beam Model
Tests is as shown in Figure 3. Two linear potentiometers (LP1 and LP2) were placed at
either side of the mid-span position. Loads were gradually applied up to failure. During
testing vertical deflections at the midspan and crack opening at the bottom of diaphragm
were measured and recorded.
TEST RESULTS
Unlike the other two tests, Test 1.1 behaved non-linearly up to a point approximately half
of the ultimate failure moment. The failure was sudden at a bending moment of 10.5
kNm. In Test 12, linear behaviour was observed up to almost three quarters of the
failure moment, where following extensive cracking the flexural stiffness had fallen one
tenth of the initial stiffness. The failure moment was 15.1 kNm, i.e 44% greater than Test
1.1. Test 1.3 (which is practically a continuition of Test 1.1) behaved in a similar manner
to Test 1.2 except that failure moment occured at 12.0 kNm, i.e only 14% greater than
Test 1.1.
Figure 7 also shows the calculated mid-span elastic deflection for Tests 1.1 to 1.3. The
results for Test 1.2 and 1.3 were broadly similar. The theoretical applied-moment vs
central deflection data for composite beam-slab section agrees almost exactly with test
results up to a moment of 5.9 kNm. Also, the final secant stiffness in the test agrees with
the calculated stiffness based on the infill cross-section.
Appropriate concrete technology 514
Cracks appeared at a total bending moment of 9.6 kNm and started from the soffit of the
bottom flanges of the precast beams near the ends of the solid infill diaphragm and
propogated upwards as the bending moments increased as shown in Figure 10. No cracks
appeared elsewhere in the connection. Strain distribution in the links of the precast beams
at various distances from the connection are shown in Figure 10(a).
DISCUSSION
is capable of transferring the sagging moment from one span to another. Initial
considerations indicate that moment continuity is produced by the direct bearing action of
the insitu diaphragm against the precast beam flangesthe resisting couple acting about
the centre of rotation I, as shown in Figure 11(b). Figure 11 (a) shows the strain
distribution in the links of the precast beams adjacent to the diaphragm. As the applied
bending moment increased, the tensile strains in the links gradually increased and the
centre of rotation of the couples created in the connection moved away from the edge of
the beam toward the ends of the solid infill diaphragm. At the failure bending moment,
tensile forces in the links on one side of centre of rotation were balanced by a point
compressive force exerted by the infill on to the bottom flange of the beam which caused
cracking in the precast beamsshown in Figure 10. This prevented the development of the
theoretical ultimate moment of resistance of the infill of 19.51 kNm. The ratio of
test/calculated bending moment was 0.62.
CONCLUSIONS
The ultimate sagging moment capacity of the connection formed of JJ3 model precast
beams, i.e. the initial tests, but excluding shear reinforcement was higher than its ultimate
value. The mode of failure, however, involved a single large web crack and an
unacceptably sudden failure, indicating a general weakness of the unreinforced web zone
adjacent to the joint. It was decided that it would not be possible to develop the research
using these units. Shear reinforcement must be present in the web of the beams in order
to resist the upthrust force exerted on the shoulder of the top flange of the precast beam
by the insitu concrete infill and prevent end splitting.
In the case of web and interface connection provided with sufficient shear
reinforcement, the cracks develop more gradually in the precast units away from the
connection and near the ends of the solid infill diaphragm. Due to the pesence of shear
reinforcement, applied moment was transformed into couples in the connection, i.e.
tensile forces in the links balanced by compressive force exerted on the flange of the
beam.
The reinforced connection using the infill diaphragm is seen to be effective in
transferring the sagging moment from one span to another. An extensive series of tests is
now under way to investigate the effect of the joint parameter, and to improve
understanding of the moment transfer mechanism.
REFERENCES
3. PRITCHARD, B P. Bridge design for economy and durabilityconcepts for new, strengthened
and replacement bridges. Thomas Telford, London, 1992.
4. TRENT JETFLOOR LIMITED. Jetfloor standard, October 1992.
EVALUATION OF HONEYCOMB
CONCRETE
G W Seegebrecht
S H Gebler
B G Stejskal
Construction Technology Laboratories Inc
USA
INTRODUCTION
Girder A
Girder A was designed with the following parameters:
Besides the problem of the aggregate size being too large for the parallel reinforcement,
the average clear spacing between bars was 2.4 in. (61 mm). Taking into consideration
that laps are required in this area and the lengths of these laps are approximately 4 ft-6 in.
(1.4 m), this results in a clear spacing between bars of 1 in. (25 mm). ACI 318 Code
requires that the clear spacing between bars be greater than or equal to db, the bar
diameter.
Girder B
Girder B was designed to contain the following elements:
The resulting average clear spacing between steel reinforcement was 2.32 in. (59 mm).
This steel arrangement met the Code. However, intersecting with this member is a
column that contained a curtain of No. 4 ties and No. 10 vertical bars. The net result is
that clear spacing between bars was reduced to 3/4 in. (19 mm). Besides failing to meet
code requirements for db, the aggregate size becomes a more significant overriding factor
in the formation of rock pockets. By Code the NMSA should not have been greater than
9/16 in. (13 mm).
RECOMMENDATIONS
ACI Committee 309, Consolidation, has published several documents concerning how to
construct honeycomb free concrete. ACI 309.3R [2] Guide to Consolidation of Concrete
in Congested Areas, presents examples of what is considered to be congested
reinforcement. The Guide simply states:
Appropriate concrete technology 524
Where requirements dictate that highly congested sections are needed, the designer and
contractor should meet to resolve constructability issues.
We have shown through these two case histories that honeycombs or rock pockets
could have been avoided using some simple rules and common sense. The following list
of recommendations is presented to reduce or eliminate the occurrence of honeycombing.
Re-evaluate the member design, especially overall dimension and connection to other
members.
Consider tolerances for cover and bar spacing in the design stage.
Use alternate splicing methodsmechanical vs. laps.
Consider appropriate mix design especially aggregate size, admixtures, appropriate
workability and thorough consolidation.
Maintain form tightness to prevent paste or mortar leakage.
Ensure that consolidation methods can be effectively accomplished.
The recommendations provided in this paper and those detailed in ACI 309.3R should be
followed to reduce the risk of costly repairs associated with these defects. The associated
costs for using modified mixtures, smaller coarse aggregate and/or redesigning the
member will more then offset the cost of remediation.
Using a concrete mix containing too large a coarse aggregate for bar spacing, cover,
and steel density contributes to bridging of aggregate particles and segregation. Stiff
harsh mixes require more compactive effort for proper consolidation and are difficult to
pump.
Another consideration to offset the effects of densely placed reinforcement is to attain
access for vibrators to achieve proper consolidation. If restrictions on the spacing and
density of the steel are so difficult it may also be necessary to employ smaller diameter
vibrators in conjunction with the normal size vibrators to consolidate the concrete. In
extreme cases, form vibrators may be helpful in conjunction with internal vibrators.
REFERENCES
1. Whiting, D., Seegebrecht, G.W., and Tayabji, S., Effect of Degree of Consolidation on Some
Important Properties of Concrete, American Concrete Institute, Special Publication SP-96,
Detroit, Michigan, pp. 125143, 1987.
2. Guide to Consolidation of Concrete in Congested Areas, ACI 309.3R-92, reported by
American Concrete Institute Committee 309, ACI Manual of Concrete Practice 1993, Detroit,
Michigan, 1993.
FLEXIBLE REINFORCED CONCRETE
BEAMS ON ELASTIC RUBBER
FOUNDATION MODEL
A Hassani
Tarbiat Modarres University
Iran
D W Cox
University of Westminster
UK
INTRODUCTION
It is known that a beam composed of laminations placed freely upon one another deflects
more under a given load than does a solid beam of the same cross section under the same
loading conditions. The difference in deflection is caused by the fact that in the solid
beam all longitudinal planes can support shear, whereas a freely laminated beam can not
resist shearing load on the planes that separate the individual laminations. This paper
discussed a laboratory investigation which was performed to determine the response of a
concrete laminated beams bonded together with a means of a thin layer of bitumen. In
order to reduce the internal stresses while maintaining the benefits of a thicker concrete
pavements.
This inability to resist horizontal shear has led the use of various devices such as
riveting, welding, or bolting, to connect the laminations so that the laminated beam can
be made to act similarly to a solid beam.
Due to the viscosity of the bitumen layer the resulting laminate acts as a whole when
suddenly loaded by the traffic, and as individual thin flexible laminate when slowly
internally stressed by thermally induced volume changes in the concrete, or loss of
support from the soil below.
SUB-GRADE MODEL
The model sub-grade material was required to simulate two types of behaviour.
a) a relatively high recoverable deflection under load, due to weak sub-grade material.
b) a deflection due to long term irrecoverable subsidence of the sub-grade independent
of the applied load.
Previous investigations into the behaviour of layered systems have widely adopted
support conditions using soil [1,2]. It has been generally found however, that a uniform
condition is difficult to achieve and in any case ensuing moisture changes lead to
variation in support. Other investigations used a Winkler foundation, consisting of a set
of independent springs arranged in a circular pattern as the support for a series of tests on
bituminous beams and slabs. Other investigators adopted support conditions using
closed-cell expanded natural and solid rubbers used for support layers with cemented and
bituminous materials [3,4,5 and 6].
This investigation is concerned with the study of the behaviour of laminated beams on
an elastic rubber foundation model under static and rapid (traffic) loading.
Flexible reinforced concrete beams on elastic rubber foundation model 527
PROCEDURE
Beams preparation
Initial preparation started with laying out the rubber pads to the required dimensions and
pattern. The rubber foundation for the beams considered of a row of 21 positions with
either one 6mm, or 36=18 mm, or 96=54 mm rubber pads each 15080mm in plan at
pre-determined intervals. Steel shuttering was used for the sides, which was bolted to
supporting frames using an angle section. When the mould had been erected and
positioned relative to the reaction frame, 2 number 8mm diameter high tensile
reinforcement bars were laid in position, (at centre). The load footprint used throughout
this investigation was a rectangular 150100 mm steel plate 30 mm thick which was
bedded on a layer of plaster to give an even load distribution. Load was transmitted via a
calibrated proving ring with 50 kN capacity from a manually operated screw jack with 50
kN capacity, for application of a constant rate of deformation.
Appropriate concrete technology 528
The laminated beams were also subjected to different constant load applications for
creep measurement by means of a lever arm method. Data were recorded by observation
of dial gauges over the length of the beams at different time periods up to 24 hours. In
each case on un-loading the procedure was the same sequence as for loading.
In laminated beams, after casting, the beam was left to cure for 3 days in a wet condition
and the bitumen layer with 1.5mm thickness (as discussed in more details in Ref 7), was
commenced on the fourth day. At this stage the side moulds were extended up for the top
concrete layer, as for the bottom layer. Twenty four hours after casting the top layer the
mould was stripped down and the concrete cured as for the bottom layer for three days
and the final preparation for testing commenced on day four. This include the positioning
and fixing of transducers and dial gauges on the beam side near the lamination level and
on the top surface, see Figure 1 for general view.
For Solid beam the reinforcement bars were laid in position at 50 mm and 150 mm
depth to comply with configuration of laminated beams. After casting the same procedure
was followed as for laminated beam.
Flexible reinforced concrete beams on elastic rubber foundation model 529
TEST METHODS
Beams were tested using various constant rates of central deflection (recording the load
necessary to maintain the constant rate) and also by applying various constant loads for
chosen durations (recording the central deflection, and deflection profiles along the
beam). Constant load tests were repeated on a solid beam with no lamination. The main
variable in all cases were central deflection (and deflection profile), load and time, (either
as duration of loading or rate of deflection).
All beam tests were carried out on three different model sub-grade stiffnesses (soft,
medium and stiff consisting of 9, 3 and 1 thicknesses of rubber pad).
weight), for method of creating the settlement profile see [7]. From Figure 3, 4 it can be
seen that the surface deflections are more or less the same under traffic loading. This
shows no or very little loss in stiffness of the laminated beams. In the slow loading the
surface deflections were greater (10%) over 24 hours when artificial settlement was
introduced under three rubber pads and 14% greater when artificial settlement was
introduced under 9 rubber pads. The foregoing discussion has presented an evaluation of
laminated beam response under a known settlement profile. In particular it has been
shown that the behaviour of the laminated beam supported over a subsidence sub-grade
model is in agreement with the predicted behaviour, that the beam will tend to deform
with the subsidence, so that deflection under load is relatively unaffected by the ground
movement.
It is also shown that, with longer loading time the deflections along the laminated
beams increased significantly especially on softer sub-grade.
CONCLUSIONS
When designing pavement it is assumed that the slab is in contact with the soil and the
slab distributes the load directly to the sub-grade. If the soil subsequently deforms, as it
often does, this leaves the slab acting as a bridge between high spots. The thicker the slab
the more rigid it is and hence the greater the span between the
Appropriate concrete technology 532
high spots. The results showed that this was not the case with laminated model beams,
which retained its flexibility due to the laminations and maintained contact with ground
between high spots. This enabled the traffic load to be transmitted directly to the sub-
grade without over stressing the pavement, while maintaining the benefit of the thicker
concrete pavement. The other principal findings arising from this work are as follows;
1. The laminated beam tests showed that the beams retained 75% of the stiffness of a
solid beam of the same cross section when loaded rapidly. In long term loading the
laminated beam stiffness was reduced by 30%. The experiments also showed that, the
stiffness of laminated beams under the same loading system increased with reduced
support or sub-grade stiffness (thickness), whereas with the solid beam this effect was
insignificant.
2. The experiments showed that in the laminated model, the internal stresses are
reduced as a result of laminating the thicker concrete pavement with bitumen which
allows each lamination to slowly change length with temperature, reducing ground
friction.
Appropriate concrete technology 534
REFERENCES
1. BUSH, D.I Behaviour of flexible pavements subjected to dynamic loads. PhD thesis, University
of Nottingham, 1969, England.
2. WANG, M C AND MITCHELL, J K. Stress-deformation prediction in cement-treated soil
pavements. Highway Research Board, Highway Research Record 351, Washington DC, 1971,
pp 93111.
3. CARLTON, P F AND BEHRAMNN, R M. A model study of rigid pavement behaviour under
corner and edge loadings. Proceeding 35th annual meeting. Highway Research Board,
Washington DC, 1956, pp 139146.
Flexible reinforced concrete beams on elastic rubber foundation model 535
4. KEKWICK, S V. Test on wet-lean concrete in relation to use for road construction. PhD thesis,
1980, University of Surrey, England.
5. DIJIK, van W. Practical fatigue characteristics of bituminous mixes. Paper presented at 1975
annual meeting of the Assn of Asp Pavement Technologists, Phoenix, Arizona, Feb 1975.
6. RAMSAMOOJ D V, MAJIDZADEH, K AND KAUFMANN. The analysis and design of the
flexibility of pavements. Proceeding 3rd Int Conf on Structural Design of Asphalt pavements,
London, 1972, England.
7. HASSANI, A. Bitumen laminated reinforced concrete pavements. PhD thesis, University of
Westminster, London, 1991.
PORTAL A-FRAME VS PORTAL H-
FRAME STRUCTURES A CHOICE
BASED ON TOTAL COSTS
D R Lemelin
Hydro-Quebec
Canada
INTRODUCTION
Different types of structures or towers can be used to anchor the 3-phase transmission
lines at the entrance of a substation. For esthetical reasons, tubular section structures are
usually preferred over lattice-type structures. The three types of structures most
commonly used at Hydro-Quebec and elsewhere include a set of three isolated poles, an
A-frame portal structure or an H-frame portal structure. The choice of one type over the
other depends on several factors including the horizontal angle of arrival of the lines, the
layout of the electrical equipment around the structure, space limitations and costs. When
the oncoming angle of transmission lines is less than about 15, portal type structures can
be used. If so, the choice between an A-frame and an H-frame structure is mainly
dependant on costs.
The first part of this paper gives the loading criteria for the 120 kV transmission lines
arriving at the Groulx substation. The second part presents the designs of an A-frame
structure with its superficial foundations and an H-frame structure with its caisson
foundations for a soft clay soil. Total costs are detailed and, finally, a comparison of the
two solutions is provided.
LOADING CRITERIA
The loading conditions on the structures are governed by the sag and tensions of the
transmission lines which are dependant on the climatic conditions prevaling at the site.
Wind and ice loads were obtained from Hydro-Quebec technical specification SN-40.1
[1]. They are limit loads corresponding to an acceptable risk of being exceeded in the
useful lifetime of the lines. The 120 kV conductors have a diameter of 27.8 mm and a
unit weight of 14.9 N/m. The ground wire has a diameter of 11.0 mm and a unit weight of
5.7 N/m. The span of the lines is 205.5 m. The sags and tensions for the different loading
conditions are given in Table 1. The angle of the oncoming lines is 10. Another loading
condition to consider includes earthquake forces, which in the present case do not govern.
Table 1Sags and tensions
Loading conditions
45 mm ice, 0C 0 20 mm ice, 0C 0 mm ice, 10C 0 mm ice, 30C
kpa wind 0.65 kpa wind 1.2 kpa wind 0 kpa wind
Cable Horizontal Sag Horizontal Sag Horizontal Sag Horizontal Sag
tension (kN) (m) tension (kN) (m) tension (kN) (m) tension (kN) (m)
Conductor 72.2 7.7 51.1 6.2 38.0 5.1 24.4 3.2
Ground 48.0 8.3 31.2 6.8 15.6 4.9 9.0 3.3
wire
Appropriate concrete technology 538
where q is the soil pressure, P is the vertical load, A is the effective footing area, Mi and
Si are the bending moment and the footing effective section modulus about i respectively.
For the compression foundations, the permanent loads give qmax= 30.5 kpa vs 46.5 kpa
allowable and qmin=19.5 kpa>0 (no overturning). The ultimate loads give qmax=54 kpa
which is the allowable pressure (22.5+17.51.8) and qmin=21 kpa>0 (no overturning).
Sliding and torsion are resisted by the Rankine passive pressure at the level of the footing
using a coefficient, Kp, of 3. The maximum pressure applied is 48 kpa resulting in a S.F.
of 2 not considering the base friction. The uplift foundations are treated in the same
manner and will not be discussed here. The maximum bending moment and shear force
in the footing are respectively 88 kN.m/m and 150 kN/m. The maximum bending
moment and shear force in the beams are respectively 553 kN.m and 353 kN. The
concrete design is according to CSA standard CAN3-A23.3 included in the PCA
Concrete Design Handbook [3].
Total costs
The mass of the steel structure is approximately 11 500 kg including anchor bolts. The
unit cost of fabrication is $3.00/kg, resulting in a structure cost of $34 500. The volume
of concrete for each footing is 12.7 m3. The unit cost of construction is $700/m3 which
includes the concrete, the forms, the reinforcing bars and the styrofoam filling (or
granular backfill). The cost of the four superficial foundations is then $35 560 and the
total cost amounts to $70 060.
Portal A-frame vs portal H-frame structures: A choice based on total costs 539
The design of the caissons illustrated in Figure 2b is based on the soil horizontal reaction
modulus (Ks) to lateral pressures, which in turn is dependant on the undrained shear
strength of the clay (Cu), which is 30 kpa at the critical depth (zc= 3D) of 5.5 m. The
reaction modulus for a 0.3 m wide caisson (Ks1) as a function of Cu is taken from
Marche [5]. The variation of Ks with depth (z) is expressed by the exponential form
given below and takes into account the diameter of the caisson (D) and a surface
correction,
Ks(z)=Ksc (z/zc)0.3 and Ksc=Ks10.3/D where Ksc is the reaction modulus at zc
One way to analyse a caisson is by replacing the soil with springs having stiffnesses
equivalent to the Ks values. The numerical model is shown in Figure 2c, along with the
loads obtained with the 45 mm ice critical condition. The required length of the caisson is
determined so that its maximum rotation is limited to 0.2 and the allowable pressure of
the soil is not exceeded. The allowable pressure (Pa), is fonction of the creep pressure
(Pf) and the limit pressure (PI) taken from Marche [5] with appropriate safety factors.
Hence, at zc, Plc=9Cu, Pfc=Plc/1.5 and Pac=Pfc/2=90 kpa. The variation of Pa with depth
is given by,
Pa(z)=Pac(z/zc)0.3 for zzc and Pa(z)=Pac(2+7 z/zc)/9 for 0<z<zc
The design of the caissons is according to the Handbook of Steel Construction [2] and
the PCA Concrete Design Handbook [3].
Appropriate concrete technology 540
Total costs
The mass of the steel structure is approximately 9 200 kg including the anchor bolts. The
unit cost of fabrication is $3.70/kg resulting in a cost of $34 040 for the steel structure.
The two steel caissons are 13 mm thick and have a total mass of 24 800 kg and their cost
amounts to $91 760. The unit cost of construction is set at $1 200/m which includes the
mobilisation of the crew and the equipment, hammering the caissons, cleaning the inside,
field welding, the concrete and the reinforcing bars. For a total caisson length of 42.6 m,
the cost of construction is $51 120. The cost of the caisson foundations amounts to $142
880 producing a total cost of $176 920 for the structure and the foundations.
CONCLUSIONS
For practical purposes, it can be assumed that the total cost of steel is the same for an H-
frame or an A-frame portal structure, considering that the latter is only 1.4% more
expensive. However, for the weak soil considered here, the superficial foundations are
four times less expensive than the caisson foundations. The total cost of an H-frame
structure with caissons is then twice and a half more expensive than an A-frame structure
with spread footings. Consequently, the latter becomes clearly the best choice. In
addition, some electrical equipment can be mounted on the A-frame structure, therefore
saving ground space and equipment foundations. As the soil quality increases, it can be
expected that the cost margin between the two types of foundations will narrow. The rate
of decrease would need to be looked into further.
ACKNOWLEDGEMENTS
The author would like to gratefully acknowledge the constructive comments received
from Mr Estevao Amorim, ing., who took the time to review an initial draft of this paper.
Although most of his comments were adopted, any remaining errors
Appropriate concrete technology 542
or omissions, of course, are the responsability of the author. The author would also like to
express his kind appreciation to his superior, Mr Marcel Beauchamp, for his support and
encouragement throughtout this work. Also, sincere thanks go to Mr Robert Desrochers
for drafting the figures. Financial support for this study was provided by Hydro-Quebec,
Postes de rpartition ouest, Montral, Qubec, Canada.
REFERENCES
INTRODUCTION
A shearhead is a fabricated structural steel assembly placed within the slab depth on the
column support. Shearheads have been used widely in the United States and on the
Continent for around thirty years but their use has been very limited in the United
Kingdom. Shearheads used in the States are generally of a cruciform arrangement with
steel members passing through the columns. Generally the top reinforcement passes over
the shearhead. The bottom reinforcement may either pass under the shearhead or be
curtailed and supported on the bottom flange.
Structural steel shearheads offer many advantages in flat slab construction. The most
important of these are:
i) they increase the punching shear resistance,
ii) they give the opportunity for providing large openings close to columns,
iii) they are an efficient and practical method for transferring shear and bending forces to
columns especially edge columns,
iv) they increase the maximum column shear capacity,
v) they enable prefabrication of column-shearhead assemblies on large repetitions
projects, and
vi) they eliminate the need for using column heads and drops.
Rules for design of the cruciform type of shearhead are given in the ACI Code(1). These
are based on work carried out at the Portland Cement Association and reported by Corley
and Hawkins(2). Three design criteria are considered:
i) shearheads increase the punching shear perimeter;
ii) the flexural strength of the shearhead arm needs to be adequate;
iii) the slab reinforcement for the negative moments over the columns can be reduced.
In 1989, a project was started at the BCA with support from DTI, Square Grip Ltd and
Ove Arups to develop shear reinforcement systems for flat slabs. The first phase of this
work led to the development of the Shearhoop system (3,4). The second phase was
concerned with structural steel shearheads with the aim of developing a practical system
and producing design guidance for UK conditions. This project had to be halted owing to
organisational changes at the BCA; recently, the project has been revived at Imperial
College following a grant from the DoE and with support from the RCC. This is
scheduled for completion early in 1997 with the publication of a Design Manual.
TEST WORK
Extensive test work was carried out to assess the performance of slab reinforced with
steel shearheads.
Three series of tests were carried out:
i) Model tests on internal slab-column connections; slabs were 120 mm deep.
ii) Full scale tests on internal slab column connections; slabs were 250 mm deep.
iii)Large scale model tests on edge slab column connections; slabs were 150mm deep.
Appropriate concrete technology 546
Details of the test specimens have been published elsewhere(5) and are not repeated here.
The results are being analysed in detail to confirm the design rules being proposed.
Shearheads of the type shown in Figure 1 have been developed as the optimum design
in terms of structural efficiency. This is prefabricated from channel sections (or I sections
for large loads) with the primary arms placed either just outside the column steel or
passing through the column section. Welding of these units is to BS 5135 and in
accordance with recognised QA procedures.
This paper is concerned with draft design recommendations for the type of shearhead
developed in this project. It should be made clear that these are preliminary rules subject
to confirmation with the BCA and other test data currently being analysed. In due course,
tables of standard shearheads with given load capacities calculated on the basis of the
final approved recommendations will be produced.
The following points should be considered in the design of shearheads:
location.
control perimeter.
plastic moment capacity of steel section.
provision of openings.
maximum column face shear.
maximum shear capacity.
minimum stiffness of steel section.
span depth ratio.
transfer of moment to column.
reduction in support steel.
prestressed slabs.
These are discussed in turn.
1. Location
i) Shearheads are positioned within the slab depth so that the top reinforcement passes
over the shearhead. The bottom reinforcement may either pass under the shearhead or
Preliminary design rules for structural steel shearheads in concrete slabs 547
be curtailed along the line of the bottom flange. Where the bottom reinforcement is
curtailed along the line of the bottom flange and the edge of the shearhead lies in the
region of sagging moment, the bottom reinforcement needs to be adequately tied to the
flange by welding or other means.
In some circumstances, shearheads may be positioned directly on the slab
formwork. In such cases, the durability and fire protection aspects must be
considered carefully. Fire engineering methods can generally be used to
demonstrate adequate fire resistance.
ii) All compression flanges of the steel section shall be located within 0.3d from the
compression surface of slab and tension flanges within 0.3d of tension surface of slab.
iii) The primary arms of the shearheads can be located to pass through the columns.
Where this interferes with the column steel, it is possible for these arms to be located
just outside the column provided the column face shear is not greater than the critical
value (see section 6). The channel section can have a secondary plate welded to it so
as to form a shear key; it should be noted that the section could be positioned within
the cover region of the column.
Small holes are provided within the web of the shearhead to facilitate concrete
being provided within the channel and to improve composite action.
2. Control Perimeter
In the ACI standard, enlarged values of the critical perimeter are specified.
Extensive test work was carried out at the BCA to investigate this particular
aspect of shearhead design. The tests show that, for these shearheads, it will be
conservative to take the control perimeter at 1.25d from shearhead face and the vc
values given in BS8110 may be used.
There has to be a maximum limit on the size of the control perimeter. This will be
covered under section 5 on maximum shear capacity.
3. Plastic Moment Capacity of Steel Section
The load transfer mechanism of these shearheads is quite different from ACI type
shearheads and design rules based on first principles have to be developed for
checking the moment capacity of the shearhead arms. The intention is to ensure
that the steel section does not fail prior to a punching shear failure outside the
shearhead.
The assumption made is that the shear forces along the secondary arms on the
perimeter of the shearhead are evenly distributed. The magnitude of this force is
where V is the total column shear and is the product vc bo d where bo is
the perimeter of the shearhead. The forces from the secondary arms are then
transmitted directly through the primary arms to the columns.
The moment capacity required, Mp, of each primary arm is then given by
Generally, the moment capacity of the secondary arms does not have to be
checked separately if they are fabricated from the same section. Also the primary
steel arms have to be checked for web shear and buckling.
4. Provision of Openings
Openings may be provided anywhere within the shearhead. The sum of the
lengths of the openings in any direction should not be greater than one third of the
length of shearhead measured in the same direction. The required bending
reinforcement interrupted by the openings should be placed either side of the
openings up to a maximum distance of d, the effective depth, outside the
shearhead.
When openings are provided, the primary arms need to be stiffer because the
shear forces carried by the steel section and hence the moment capacity required
are increased. in section (3) above is reduced according to procedures given in
CIRIA 110 (6) for provision of holes. The calculations for the plastic moment
capacity can then proceed as in section (3) above; alternatively it is conservative
to take as zero.
5. Maximum Shear Capacity
This is a problem area as far as the drafting of rules is concerned. The facts are:
i) European literature claims the punching shear resistance can be quadrupled by
providing shearheads.
ii) The ACI Code has a limit of 1.75 Vc for shearheads (1.5 Vc for conventional
reinforcement).
iii) EC2 refers to design by testing for shearheads. There is, however, a limit on the
size of loaded area to 3.5d which is very small as far as the size of shearhead is
concerned.
The reason for imposing this limit is the lack of test data on large shearheads. It is
suggested that the extension of the primary shearhead arms is limited to 3d from
the column face. In practice, this will mean that shearheads are provided within
the range 1.5d to 3d from column face. For a 3d extension, the control perimeter
is 4c+34d as compared to 4c+12d for a slab without shearhead giving a limit of
around 2.5 Vc for the maximum shear force with shearhead. It should be noted
that the current limit using normal reinforcement in BS8110 is 2Vc and 1.6Vc in
EC2 (2Vc in NAD).
6. Maximum Column Face Shear
Where the primary arms are placed on the column face, the face shear should be
limited to the lesser of 0.8fcu or 5 N/mm2. Where this limit is exceeded, the
primary arms can be located within the column; some of the shear force is then
transferred to the column by direct bearing and the remainder along the column
face. The maximum shear capacity may be doubled by this means.
7. Minimum Stiffness of Steel Section
The ACI Code requires the ratio v between the stiffness of the shearhead arms
and that of the surrounding composite cracked slab section of width c= to be
greater than 0.15. This requirement is considered unnecessary for these
Preliminary design rules for structural steel shearheads in concrete slabs 549
shearheads where the stiffness will be determined from the required plastic
moment (see Section 3).
8. Span: Depth Ratios
The BCA test work has shown a significant increase in the stiffness after cracking
of the test slab with shearheads (around 50% increase). However, it is likely that
in real structures the effects would be less significant owing to continuity of the
flat plate beyond the line of contraflexure. A grillage analysis carried out to
investigate this further has proved to be inconclusive.
Until further information becomes available, it is recommended that rules for span
depth ratios given in BS8110 are used for the design of slabs with shearheads.
9. Transfer of Moment to Column
There is some American test data on ACI type shearheads to suggest that
conventional methods for dealing with moment transfer such as in BS8110 are
appropriate for the calculations of shear stresses caused by transfer of moments
even when shearheads are used. No additional recommendations are necessary.
For edge columns, when unbalanced moments are considered, the shearhead must
have anchorage to transfer the moment Mp (see section 3) to the column. This can
be done simply by bearing within or outside the edge column.
10. Reduction in Support Steel
The ACI standard allows the support moment for designing the reinforcement to
be reduced by the moment carried directly by the shearhead. Some European
literature claims that this can be halved in many cases.
There seems no reason why the support moment should not be reduced by the
calculated value Mp although in practice this makes little difference to the steel
arrangement.
11. Prestressed Slabs
The principles outlined in this note can be applied to prestressed slabs. There
may, however, be difficulties in preparing tables of standard shearheads with
given load values for prestressed concrete slabs. Further work will be required in
this area.
CONCLUDING REMARKS
Preliminary design rules for providing steel shearheads in concrete slabs have been
described. Calculations such as those required can be fairly time-consuming and the next
phase of this work is to provide a Design Manual with tables of standard shearheads with
given load capacities. This Design Manual should be available in early 1997.
REFERENCES
Keywords: Reinforced concrete beams, Shear resistance, Exposure to fire, Centre bars,
Non-linear FE analysis
Mr Satish B Desai is a Principal Civil Engineer with the Building Regulations Division
of the Department of the Environment, London. His main responsibilities include
management of research projects in support of the British Standards and Eurocodes on
Actions, Geotechnical Design and Concrete.
Dr Nutan K Subedi is a Senior Lecturer in Civil Engineering at the University of
Dundee, Scotland, UK. His main interests are in the field analysis and design of concrete
structures; for example, deep beams, panel structures, tall buildings, shear walls, and
concrete elements subjected to hydrostatic pressure effects.
Professor Kuldeep S Virdi is Professor of Structural Engineering and Director of
Structures Research Centre, City University, London, UK.
Appropriate concrete technology 552
INTRODUCTION
Concrete is known to provide good protection to the steel reinforcement, from the
environmental effects as well as from exposure to fire. The cover to the steel, therefore, is
a main consideration in the design of reinforced concrete members. The prescriptive
recommendations for fire rating of structural elements include the cover to the
reinforcement and minimum thickness of members. Also, the rules for design against
shear require structural elements to have certain minimum dimensions, for example,
members with thickness less than 200 mm which cannot be reinforced against shear using
links.
In this paper, a design method is developed for estimating the load carrying capacity
of members under normal as well as fire exposure conditions, using the centre bars as
shear reinforcement. This method could allow flexibility of choice of dimensions of
members and the strength and properties of the constituents of the members.
Centre bars would serve as shear reinforcement for members irrespective of their size.
Also, their contribution could be available for a longer period of exposure to fire, due to
the protection provided by the surrounding concrete. The centre bars could assist in
reducing the requirement of links and improving the detailing of steel at supports by
avoiding congestion of steel. In addition to the extra shear carrying capacity, members
provided with centre bars could have an effective tying and an additional ductility which
is important for design against accidental loading and for providing an improved residual
resistance.
The design of reinforced concrete members against shear failure mainly concerns
provision for avoiding a brittle and sudden failure. Although there is no conclusive
solution available to the problem of shear, some general principles can be deduced from
the research by Kani [1], Regan [2] and Taylor [3]. A reinforced concrete member is able
to carry the applied shear through three modes of shear transfer; reactive stresses in the
compression block concrete, aggregate interlock across a shear crack and the dowel
action of the tension steel. Also, the reduction in the depth of compression block is
considered to be the main damaging effect of the increasing shear force.
These shear transfer modes are influenced by the components of the section; concrete,
the tension steel and the web steel. Their influence changes with the increase in applied
shear. However, the increase and decrease in the individual contributions of the
components could compensate each other. For example, the aggregate interlock provided
by the concrete could be partly taken over by links. The links would assist in bridging
over the local weakness in concrete section and redistribute the stresses and also enhance
Shear resistance of reinforced concrete members at high temperatures 553
the resistance to dowel-splitting in the tension steel region. The tension steel provides the
stiffness which enables the member to resist the worsening of cracks. This depends on the
amount of tension steel and its modulus of elasticity. These contributions of the
constituents of the member are combined in a rule to give a part of the ultimate shear
resistance of concrete (Vcu). Vcu depends mainly on the strength of concrete and, also, on
the amount and modulus of elasticity (Est) of the tension steel.
Vcu=0.0046 ( Estfcu)1/3 (400/d)0.25 bd
(1)
b width of beam (mm)
d effective depth of the beam (measured from the extreme compression fibre to the (mm)
centroid of tension steel)
fcu Characteristic cube strength of concrete (N/mm2)
100Ast/bd, where Ast is the amount of tension steel
With the value of Est of 200 kN/mm2, this rule gives the corresponding BS 8110 rule. The
links have a complex role in enhancing the shear carrying capacity of a section.
Ideally, the links should be treated as reinforcement acting in conjunction with the
concrete and enhancing the shear resistance of a member. For the sake of convenience,
the ultimate shear resistance (VDU) of a member reinforced with links is expressed as
addition of two components; one provided by concrete, VCU given by the previous
equation and the other provided by links, VLU.
VLU=(Aswfyvd)/s
(2)
fyv yield stress for steel reinforcement (N/mm2)
Asw area of cross-section of links (mm2)
S spacing of links along the length of the member (mm)
(5)
(6)
(7)
i) the ambient high temperature which is a function of the fire exposure time (t, in
minutes);
ii) b, the width of the cross-section (mm); and
iii) r, the ratio of the overall height to the width of the beam.
The following cubic equation is proposed and the values of D, A, B, and C are obtained
by solving a number of simultaneous equations constructed to represent, as closely as
possible, the trends given in Wades charts and to accord with measurements by Lin and
the data obtained from the tests done under this project.
T=(DAx+Bx2Cx3)/r0.25
(8)
B=0.085 C=0.000221
These equations can be used to plot Graphs, to give temperature distribution in beams
with various widths and for various fire exposure periods. Using these equations, the
formulae for VDU, VCT, VLT and VBT are set in a LOTUS Spreadsheet program which is
easy to operate.
Beam Tests
Ten rectangular simply supported beams were tested at Veseli, near Prague under fire
exposure conditions. The soffit and two vertical faces of the beams were exposed to fire
conforming to the standard time-temperature relationship. During the tests, failure modes
of the beams were noted and mid-span deflections and temperatures developed in the
beam were recorded. The beams were 200 mm wide and 300 mm deep and the span was
1400 mm. The cover to main steel was 25 mm. The other details of the specimens are
shown in Table 1 and Figure 1.
All the beams failed in shear and the ratio of West/Wt is close to 1.00 for all beams
except beams B101 and B402. The specimen B101 was used as a trial specimen and the
maximum applied load was 56 kN. The trial test was useful in taking some additional
measures for the other tests; for example, additional insulation was provided on the top of
the beam for the subsequent tests to protect it from heat from the furnace. Beam B402
failed earlier than expected and the temperatures developed
Table 1 Details of beam specimens
Spec No fcu N/mm2 Top steel Tension steel Centre steel Links
B101 39.6 2T12 3T20
B102 39.5
B301 34.4 1T16
B302 34.0
Appropriate concrete technology 556
in the beam at failure were much lower than the other beams. This could be attributable
to some local weakness resulting in severe spalling near the support, which was noted
during the test. The spalling could have caused direct exposure of tension steel to fire and
a rapid reduction in the strength and modulus of elasticity of the steel.
A non-linear finite element method was used to analyse the beams in order to verify the
contribution to the shear resistance capacity made by the provision of centre bars. The
software used was ABAQUS. Due to symmetry only half the beam was idealised using,
C3D20R, three-dimensional brick type elements with 20 noded reduced integration. The
REBAR facility available in ABAQUS allowed reinforcing bars to The test results are
compared with the estimated load carrying capacity as shown in Table 2.
Wt : The centrally applied load (kN). Wt is twice the applied shear, calculated as 60% of the load
carrying capacity of the beam at room temperature using a notional value of fcu of 30 N/mm2.
West Estimated load-carrying capacity corresponding to the duration of the test (tt, minutes) [twice
the value of the estimated shear VDT];
m: The mid-span deflection (mm)
Shear resistance of reinforced concrete members at high temperatures 557
be specified anywhere in the solid elements. The reinforcing bars were defined
individually at the specified locations in the cross-section. The details of the beam are
shown in Figure 2. For this study three of the beams containing the centre bars are
discussed. These are B300, B400 and D200. Beams B300 and B400 contained 116 mm
and 120 mm centre bars and beam D200 contained 116 mm centre bar plus 6
mm vertical links @ 200 mm centres (Table 1).
The material non-linear properties adopted for concrete is defined by a uniaxial stress-
strain relationship as shown in Figure 3. The peak stress of 0.67 fcu represents the
maximum stress in concrete under uniaxial stress condition. For the reinforcing bars a bi-
linear stress-strain relationship, Figure 4, was used to define its uniaxial behaviour in the
elastic and plastic regions.
Appropriate concrete technology 558
The FAILURE RATIOS parameters define the ratios of concrete strength in tension and
in multi-axial stresses. The TENSION STIFFENING curve defines the softening of the
concrete behaviour after the formation of cracks. The SHEAR RETENTION parameters
define the deterioration of shear stiffness in terms of concrete strain after the event of
cracking. The parameters defining the failure ratios, tension stiffening and shear retention
are important parameters which control the analysis process of concrete beams. For this
study the parameters were selected as follows :
*FAILURE RATIOS
1.16, 0.12, 1.28, 0.33
*TENSION STIFFENING
1.0, 0.0
0.0, 3.5 E-3 (in some cases 3.55103 was used)
*SHEAR RETENTION
1.0, 0.0075, 1.0, 0.0075
The details with regard to data preparation and input file may be fond in ABAQUS User
Manual [6].
Appropriate concrete technology 560
FE ANALYSIS RESULTS
The results from the finite element analysis and the ultimate shear capacity of the beams
calculated from the equations (1) to (4) are compared in Figures 5 and 6. In Figure 5
beams B300 and B400 series are analysed for centre bars and
representing 0.38, 0.59, 0.92, and 1.51 percents respectively. In Figure 6 D200 series
beams (with links) are compared.
In both figures, 5 and 6, the slopes of the curves represent the proposed contribution of
the centre bar from zero up to a maximum of 40% of Vc for 1% reinforcement. In Figure
6 the contribution of links is about 73% of Vcu for series D200 beam. In all three series of
beams studied here the FE analysis confirms that the proposed design equations are good
and agree well with the analysis. From the FE analysis, although it is not possible to
extract exactly the amount of contribution made by the centre bars, the overall estimate of
shear resistance capacity are in agreement with those obtained from the design rule.
Shear resistance of reinforced concrete members at high temperatures 561
CONCLUSIONS
1. The centre bars provide enhanced shear resistance of reinforced concrete beams at
normal as well as at high temperatures.
2. A method of design based on simple equations is proposed. The method provides an
estimate of the ultimate shear resistance capacity of beams taking into account the
contribution of the centre bars.
3. A non-linear finite element analysis confirms the adequacy of the estimate of the
overall capacities of beams given by the proposed design method.
REFERENCES
1. Kani on shear in reinforced concrete: Kani, Huggins and Wittkopp; Department of Civil
Engineering; University of Toronto: 1979.
2. Shear Study Group Report: The Institution of Structural Engineers, London, January 1969.
Appropriate concrete technology 562
3. Taylor H P J, Investigation of the forces carried across cracks in reinforced concrete beams in
shear by interlock of aggregate, Cement and Concrete Association, 42.447, November 1970.
4. Wade C, Method for fire engineering design of structural concrete beams and floor systems,
Technical Recommendation 8, C1/SfB/HQ4 (K), New Zealand, February 1991.
5. Lin T D, Ellingwood B and Piet O, Flexural and shear behaviour of reinforced concrete beams
during fire tests, NBS-GCR-87536, US Department of Commerce, National Institute for
Science and Technology, Centre for Fire Research, Gaithersburg, MD, USA, December 1988.
6. HKS Ltd, ABAQUS User Manual Vols I & II
REPEATED AND CYCLIC LOADING
TESTS ON PRECAST CONCRETE
BEAM-TO-COLUMN CONNECTIONS
Y C Loo
Griffith University
Australia
B Z Yao
BHP Steel Building Products
Hong Kong
S Takheklambam
India
Ms Bao Zhong Yao obtained her master of engneering degree in 1993 from the
University of Wollongong NSW, Australia. She is a structural engineer with BHP Steel
Building Products (Hong Kong) Ltd currently working in Southern China.
Mr Sukumar Takheklambam graduated from the University of Wollongong with a
master of engineering degree in early 1995. He is a practising structural engineer in India.
INTRODUCTION
Connection design is one of the most important considerations for the successful
construction of precast reinforced concrete structures. The detailing and structural
behaviour of the connection affect the strength, stability, constructibility as well as load
redistribution of the building under loads.
Despite the fact that precast concrete connections have been used all over the world
since the 1950s, a very limited number of studies has been conducted on the
performance of precast concrete connections. Although the PCI manuals [1,2] contain the
descriptions of approximately 40 beam-to-column connections fulfilling many functions,
published test results are available for only a few of them. Reliable connection behaviour
can only be assessed by physical testing.
A laboratory study has been conducted of the strength and deformation behaviour of
beam-to-column connections suitable for use in precast reinforced concrete building
frames. In all, 24 half-scale model connections were designed, built and tested. They
include,
for the unidirectional repeated and cyclic load tests, 2 models each of the monolithic
and precast connection Types A and B [3,4];
for the additional static load tests, 4 monolithic models and 4 each of the precast
connection Types [3,5].
The behaviour of the connections under static and repeated loads has been reported
elsewhere [6]. Thus only a summary of the findings is given here for completeness. This
paper presents the results in detail of the cylic load tests on the precast connections.
TEST MODELS
The design of the models was based on the structural requirements of a five-storey
reinforced concrete frame forming part of a low-cost residential building system [7]. The
two types of precast connections were designed according to recommended guidelines
[1,2,8,9]. The reinforced concrete design and manufacturing process complied with the
Australian Standard [10,11].
In all, 24 half-scale connection models were fabricated, making eight groups of two
precast specimens (Types A and B) and one monolithic specimen. All models had the
same overall dimensions but different groups had different concrete strengths and/or steel
ratios. Of the models, twelve were tested under unidirectional repeated and cyclic
loading. The remaining twelve were investigated under static loading. Each model was
identified by two letters and a number. The first letter, C, R or S, indicates cyclic,
Repeated and cyclic loading tests on precast concrete beam-to-column connections 565
Monolithic
The loading apparatus consisted of three floor-mounted steel portal frames, as shown in
Fig.4. The required loads were applied by the hydraulic jacks. The top end of the column
was fixed to the cap, which was fixed to load frame 1 through jack 1. The free end of the
connecting (precast) beam was loaded by the vertical double-acting actuator or jack 2
(INNERPAC RAH-306) with jack 3 providing the balance for jack 2 at load frame 3. The
loads were measured using Interface load cells (model 1220-BF with 113.5 kN
Appropriate concrete technology 568
capacity). The vertical deflection of the connecting beam directly under the loading point
was measured by dial gauge 1. The beam and column deflections and concrete strains
were recorded manually up to failure. The load-deflection curves of the models under
repeated and cyclic loading were drawn with the aid of a Hewlett Packard plotter. The
strains on the reinforcing bars at the connecting zone were measured using 10 mm
electrical resistance strain gauges. The strain values were recorded using a Hewlett
Packard 3054A automatic Data Acquisition/Control System.
For every model test, an axial load (Pc) was first applied on the top of the column.
This load, which was equal to 10% of the design axial strength of the column, was kept
constant throughout the test. Then a vertical load, Pb, was applied to the connecting beam
stage by stage until failure of the model occurred. For the repeated and cyclic loading
test, the load Pb was controlled by the magnitude of the vertical deflection, A, measured
at the tip of the beam (dial gauge 1, Fig. 4). The vertical tip deflection was increased in
multiples of y, where y is the deflection at first yield. Typical load history diagrams for
the repeated and cyclic loading tests are shown in Figs. 5 and 6 respectively.
In between load applications, visual inspection and manual marking of cracks and
crack propagation were carried out. Failure was indicated by a marked increase in beam
deflection accompanied by a.rapid decrease in the vertical load Pb.
Based on the model test results, the following observations were made [6]:
Under both static and repeated loading, the precast connections attained a higher
flexural strength than monolithic connections.
Under static loading, the ductility performance of Type B precast models is satisfactory
when compared with that of the monlithic connections. In this respect, Type A
connections are superior to Type B connections and the monolithic models.
Appropriate concrete technology 570
Under repeated loading, the ductility characteristics of both types of precast connections
are satisfactory, although Type B connections performed marginally better than Type
A connections.
Both the precast connection types, under repeated loading, possessed larger energy-
absorption capacities than the monolithic models.
Flexural Strength
The measured and predicted ultimate loads of the connecting beam are presented in Table
2. Similar to the repeated load tests, it was found that Pu of the precast models under
cyclic loading were greater than those of their monolithic counterparts (with the
exception of model CA1). The improved performance was due mainly to the strength of
the cast-in-place concrete which was much greater than the concrete strength of the
component beams and columns and the corresponding monolithic models (see Table 2,
column 2).
models with a lower tensile steel content (Group 7), the precast connections performed
better in ductility than the monolithic one.
A typical load-deflection curve for the group 7 models under cyclic loading are given
in Fig. 7. From all of these curves (for groups 7 and 8 models) [4], the following
observations were made:
The load-deflection curves of the precast models are very similar to those of the
monolithic models. This indicates that the deformation behaviour of the precast
connections is similar to that of the monolithic ones.
The precast models sustained more cycles than their monolithic counterparts. They
generally experienced no more degradation in load carrying capacity than the
monolithic ones.
The Type B connections showed a higher load carrying capacity and smaller crack
width at failure. They performed marginally better than Type A.
Energy Absorption
For all the models, the values of the cumulative energy absorbed are calculated and
presented in Table 2. It is clear that both types of precast models performed better than
Appropriate concrete technology 572
CONCLUSIONS
Based on the test results of the half-scale beam-column connection models, the following
conclusions can be drawn.
(1) Under cyclic, repeated or static loading, the precast connections, in general, attained a
higher flexural strength than the monolithic ones.
(2) Under cyclic loading, the precast models were capable of sustaining large beam
rotation and deflection. Hence the two types of precast connections are considerd to
have performed satisfactorily.
(3) Both the precast connection types, under cyclic loading, possessed a larger
energyabsorbing capacities than the monolithic models.
REFERENCES
INTRODUCTION
The theoretical study of plain concrete is based on the Dutch code NEN 6720 (1991) [1].
It is however obvious that any other code can be adopted, provided that it is sufficiently
developed with regard to plain concrete design. The stragegy is to exploit concrete to the
ultimate and to use standard solutions for traditional reinforcement where needed. The
optimal result is gained when SISMO walls are designed as braced construction
elements whose horizontal loads are supported by other, bracing elements belonging to
the same construction, e.g. shear walls. In that case, the outer walls of a (high rise)
building are subjected to axial load with a given excentricity, and windload
perpendicularly to the plane of the wall. According to NEN 6720, such walls can be
designed in plain concrete, provided that md+mt.i.mu (1), where
md design value for the limit state of collapse, of the maximum bending moment per unit length,
due to the loads liable to act on the structure (Nmm/mm)
mu ultimate bending moment per unit length occurring with the design value of the axial load
applied at the center of gravity of the cross section (Nmm/mm)
mt.i. design value for the limit state of collapse, of the accidental restraint moment per unit length
( of the adjacent moment of span for beams and one-way slabs and of the adjacent
moment of span for two-way slabs, according to NEN 6720, 7.3.3.).
In most cases mt.i. is too high to meet condition (1). Therefore, the Dutch code allows a
lower value if this can be proved by applied mechanics, e.g. by proving that a position
equilibrium is possible for the limit state of serviceability.
For both limit states the stress-strain curves for concrete according to NEN 6720 are
given in Figure 2.
SISMO-building technology: Plain concrete in high rise buildings 577
The curves are obtained from the conditions of equilibrium of a typical cross section.
Equations have been worked out in the Handbook for Plain Concrete SISMO Walls
[2].
floor-connection is sufficient to follow the rotation of the floor slab (pl). Figure 4 shows
the flowchart for assessing the cross section of a wall at the wall-to-floor connection.
where
eo begin excentricity of
wd design value of wind load
Appropriate concrete technology 580
The bending moment curve along the wall is either linear (no wind load) or parabolic
(wind load or soil pressure). For the wall segment above and below the floor slab, the
slope and deflection at any cross section are calculated, knowing the bending moment
SISMO-building technology: Plain concrete in high rise buildings 581
Finally, if wPl, there is position of equilibrium for the wall-to-floor connection, so that
the wall may be designed as a plain concrete element.
As shown in the Handbook for Plain Concrete SISMO Walls [2], this calculation
must be done for each possible load combination.
The study of an extensive number of examples proved that in most cases the
conditions of equilibrium within the range of the appropriate stressstrain diagram, can
be satisfied by a plain concrete section in both limit states (collapse and serviceability).
Sometimes however the axial force n is too small to generate enough bending moment
resisting capacity. This can only occur at the roof level : it is obvious that the maximum
rotation capacity of a wall is obtained when the greatest possible number of cross sections
crack and yield, i.e. when mrep is defined by the ODFH-curve of the axial force-bending
moment interaction diagram (Figure 3).
If the axial load n is smaller than only the sections immediately above and below
the floor slab will be able to crack and yield. All the other sections will deform elasticly.
The cracks in the initially cracked sections will grow at those very sections until the
structure collapses.
This is different from reinforced concrete where reinforcement spreads the crack over
a zone in various small cracks. Therefore our (provisional and very safe)
Appropriate concrete technology 582
EXPERIMENTAL STUDY
From the above we concluded that plain concrete can generally be used for walls and
wall-to-floor connections.
One theoretical exception to this are the wall-to-floor connections where
where we recommended reinforcement. In order to minimize the negative impact of this
on the work on site, we developed easy to set SISMO standard reinforcement for these
connections, which can be calculated according to tested recommendations. The program
of the experimental study is summarized in Tables 1 and 2.
The test program covered three subjects :
The bending moment resisting capacity of a SISMO wall-to-floor connection under
vertical load (V)
The bending moment resisting capacity of a SISMO wall-to-floor connection under
horizontal load (H)
Preliminary tests on the shear strength of a SISMO wall-to-wall connection (S)
Table 1. Tests on bending moment resisting
capacity of a SISMO wall-to-floor connection
Type Nr Reinforced / SISMO Thickness Thickness Diameter
Plain walls type wall (mm) slab (mm) reinforcement
(mm)
V 1 R 250 150 150 10
V 2 PL 250 150 150 10
V 3 PL 250 150 150 14
H 1 R 250 150 150 10
H 2 PL 250 150 150 10
Concrete: C20/25 according to Eurocode 2 [3]
Steel: S500 according to Eurocode 2 [3]
SISMO-building technology: Plain concrete in high rise buildings 583
Figure 7.
Appropriate concrete technology 584
It shows test piece V1 (wall-to-floor connection under vertical load) just after collapse of
the right-hand wall-to-floor connection.
The SISMO standard reinforcement of a wall-to-floor connection is shown at Figure
8.
The results of the tests carried out on the bending moment resisting capacity of a
SISMO wall-to-floor connection are summarized at Table 3.
Table 3. Comparison of calculated and measured
limit state bending moments
TEST Cracking moment Mr * Moment of collapse Mc * (kNm)
(kNm)
Nr. Angle Theoretical Measured Theoretical (3) Measured (4) M/T
(1) (2) (5)=(4)/(3)
V1 Left 21.652 22.098 61.048
Right 21.675 23.418 61.048 55.407 0.91
V2 Left 20.209 21.379 61.112 61.606 1.01
Right 20.217 19.695 61.112
V3 Left 23.460 25.059 Structure collapsed through brittle failure of wall at end
of angle reinforcement. The SISMO standard
Right 23.451 23.240 reinforcement, diameter 14 mm is too strong for a plain
concrete wall of 150 mm (see at Test nr. 3, 3.3)
H1 Left 20.209 22.737 56.140 between 53.427 0.95 and 0.99
and 55.407
Right +20.206 +22.291 +51.094 greater than >1.09
+56.040
H2 Left 20.202 23.173 56.204 between 54.308 0.97 and 1.10
and 61.606
Right +20.174 +24.038 +51.258 greater than >0.95
+48.576
Sign convention : +opening moment
closing moment
Table 2 already gave the test results for the shear strength of SISMO wall-to-wall
connections as found in short span three point bending tests.
Test S1 showed a fair correspondence between experimental and theoretical (i.e.
calculated) values.
Tests S2 and S3 however did not confirm that correspondence. Research is continued
to find out the reason(s) for the different behaviour of those tests.
Appropriate concrete technology 586
For a given percentage of reinforcement 1%, the bending moment resisting capacity,
expressed as function of the theoretical moment of collapse of the weakest member of the
wall-to-floor connection, is not less than 95% for a positive bending moment (closing
moment) and 91% for a negative bending moment (opening moment). These conclusions
confirm similar tests performed in Germany, that are reported in Beton-Kalender 1989,
Part II, E, 4.4.
This leads to following recommendation :
The SISMO standard reinforcement of wall-to-floor connections with
(steel S500 and concrete C20/25 according to Eurocode 2 [3]), may be computed like a
traditional, continuous angle reinforcement.
For this purpose an additional partial safety factor s for loads, equal to 1,15, shall be
applied on top of the normally used load safety factor(s).
This applies both for positive and negative bending moments.
where
N.B.
If the wall is designed as a plain concrete wall, the bending moment applied by the floor
at the wall through the wall-to-floor connection shall be kept as small as possible.
Therefore the percentage of reinforcement of the wall-to-floor connection shall not be
greater than 1 %.
GENERAL CONCLUSIONS
The experimental and theoretical studies described in this paper show the great potential
and flexibility of the SISMO building system.
For braced structures design becomes extremely simple.
Computation of plain concrete walls can easily be done in a spreadsheet taking in
account limit states of collapse and serviceability. When necessary, easy to place wallto-
floor connection standard reinforcement can be designed according to experimentally
confirmed recommendations.
SISMO-building technology: Plain concrete in high rise buildings 587
REFERENCES
[1] NEN 6720 (Sept. 1991). Regulations for concreteStructural requirements and calculation
methods. 1991 Nederlands Normalisatie-instituut, Delft, Netherlands
[2] Handbook for Plain Concrete SISMO Walls (1995) 1995 SISMO Engineering and
K.U.Leuven, Civil Engineering Department, Kalken/Leuven, Belgium
[3] ENV 199211:1991 (Dec. 1991), Eurocode 2: Design of concrete structuresPart 1: General
rules and rules for buildings CEN 1991, Brussels, Belgium
FACTORY AND SITE PRODUCTION
A COMPARISON
H P J Taylor
Costain Building Products Limited
UK
INTRODUCTION
Prefabrication is one of our oldest and most successful construction techniques. In all
forms of construction, apart from reinforced concrete and masonry, it dominates. On an
industrial basis it is accepted in civil engineering, ship building, car and aircraft
manufacture without question. For smaller equipment, including consumer goods and
computers, modular methods of design and manufacture dominate. It is only in the
building industry, perhaps with its tradition of hand craft work on site, with stone
masons, bricklayers, carpenters, etc, that we find resistance to the good sense of its use.
Prefabrication has an intellectual justification in the thoughts and theories of the
leaders of the industrial revolution. Adam smith expressed the benefits of the specialised
work force and factory as:
Adam Smith pointed out some of the advantages to be obtained from maximising factory
based construction.
Enhanced skill operatives
Development of machines
Speed
He left unsaid the other main advantage of precasting
Improvements in quality
HISTORICAL DEVELOPMENT
The Victorian Engineers took their lead from the earlier pioneers of the Industrial
Revolution and tackled many projects in ways which were then innovative, to produce
structures which could not be made without prefabrication.
Stephenson in 1845 moved the main box girders for the Britannia bridge across the
Menai Straits on barges to be then jacked up on their final abutments. Brunel did the
same thing with the tubular tied arch on the Tamar Bridge.
These massive structures were not manufactured in factories, but were assembled in
safe locationsprefabricatedand moved to their final location. It was probably these
spectacular successes that encouraged civil engineers to favour prefabrication. Concrete
structures have also been similarly constructed and moved in a variety of ways.
Transportation by floatingMulberry Barbours and offshore oil platforms, by launching
and by towing are all common.
DISCIPLINE OF PREFABRICATION
Prefabrication brings advantages of better buildability and speed in the whole structural
range, from the most massive to the most simple. It is only necessary to visit a precast
concrete frame structure being erected on site for one to appreciate the benefits of the dry
form of construction, no deterioration of site stored materials and the absence of a large
and often badly controlled workforce. The erection process from factory to truck to final
location on site minimises opportunity for damage and for inferior work to be built in.
The description of Prefabrication also brings many benefits from an often disregarded
opportunity in design.
Advantages of the discipline
Some of the most inspiring works of human creativity comes from the skilled use of a
limited language. In music for example a limited number of notes and instruments is
capable of providing the most stirring of results. Indeed, there appears to be more beauty
in the use of the musical language in a disciplined classical way than in more recent, freer
Appropriate concrete technology 590
use of sound. Prefabrication has this same opportunity, the observer notes the shape of
the small components and appreciates how these can be built up with only minor
modifications to a totally different, but readily understandable whole.
The aesthetic discipline can also pick up and emphasise, to advantage, a further
practical necessity, the handling of joints.
DESIGN CONSIDERATIONS
In all forms of prefabrication the presence of joints brings three immediate requirements;
the need to have sufficient strength, the need to avoid abrupt failure modes and the need
to cope with manufacturing and construction tolerances.
Strength of joints
Durability
Tolerance
From the aesthetic point of view a structure which is proud of its joints, in which joints
are well articulated and detailed to express their function and load paths can be very
successful.
In simple frame structures where competition with other materials is intense this may
not be the case and demand is for a structure which does not have a precast look. In
bridges and other structures exposed to the external environment, joints must be detailed
for maximum durability; sometimes they are expressed and sometimes not. When joints
can be expressed, however, very attractive structures can result.
All structures have joints and these create local problems of concentrated forces and
force resultants in the area immediately around the joint zone.
Saint Venant stated this principle clearly:
Away from joint zones, precast concrete design is no different from monolithic
reinforced or prestressed concrete design. Even within joint zones the same rules for
bearing and bond apply. A beam to column joint, column to column joint, or beam to
beam joint, can all be understood in terms of internal struts and ties and a successful
design will usually result if these concentrated forces and any further secondary forces
from their interaction are satisfied. The only difference in precast construction is the
additional problem of tolerances. It is easy to imagine how a physical displacement of a
supported member from its correct support position can result in higher flexural forces on
a cantilever bracket for example, for which it may not be designed. Similar displacements
of reinforced members may result in there being no overlap of reinforcement for one
member to the other. Joint design must always take account of the likely construction
variations, whether intended or not.
These aspects of joint design are not unique to precast concrete; steel or timber
construction also have similar problems which are, of course, solved in ways appropriate
Factory & site production-A comparison 591
or each material. Bolted or welded steel connections always require a detailed analysis in
the Saint Venant zone.
ASPECTS OF CONSTRUCTION
The prediction of the development of machines has certainly come to pass in precast
concrete factories. The development of machines to mix transport and place concrete has
always been important. In the precast concrete industry the machines are becoming
increasingly product specific, perhaps apart from highly engineered and architectural
products reducing the precast manufacturer to the role of mere operator. In the concrete
products area manufacturers all have similar machines and tend to compete in
transportation and service as well as quality. It is this part of our industry where the
importance of building a reputation related to brands is most clearly appreciated.
Innovation in concrete products is now not just from the precast firm, but also from the
machinery manufacturer.
The manufacture of engineered and architectural precast concrete certainly needs
refined and well controlled equipment, but is also heavily reliant on engineering,
management and craft skill. The development and retention of this skill was another of
Adam Smiths predictions. The use of factory conditions and the retention of craftsmen
now enables a quality of bricklaying and stone masons skill to be applied to structures in
areas which would be prohibitively expensive, if not impossible to achieve on site.
The lead taken by Civil Engineers in welcoming prefabrication on site is now being
taken up increasingly in the building industry. Structural components, frames, floors,
stairs, cladding, etc have long been prefabricated. Attention is now being placed on the
assembly of modular service pods, for risers, lifts and other heavily serviced areas. The
prefabrication of service areas has been one of the great advances of the last decade and
further work in this area promises more gains in speed and quality.
Our achievements in fast construction seem to be very greatbut are they? A case
study of one of the earliest large prefabricated structures brings out interesting parallels
with prefabricated structures erected today.
In June 1850, Joseph Paxton, the Duke of Devonshires Head Gardener, prepared a
simple sketch proposal on blotting paper for the hall to be used to house the Great
Exhibition in Hyde Park, due to open in May 1851. This was an alternative design, but
was, nevertheless, accepted by the Exhibition Committee, which had already rejected the
original official design. This started the race to design, detail, fabricate, erect and fit out
one of the Worlds greatest prefabricated structures. Paxton was by no means unused to
construction of large projects. He had already built some massive greenhouses for the
Duke of Devonshire at Chatsworth, one of which, the Great Conservatory, at 82 meters in
length, 37 in width and 20 metres high, was the largest glasshouse in the world. From that
to the Crystal Palace with its dimensions of 564 metres124 metres33 metres was
simply a practical difficulty!
The contractor, Cox Henderson, was taken on board in early July 1850 and the first
column was erected in mid August. A period of only 10 weeks from concept design to
start on site is more than impressive, even to this day. The programme for detailed design
and construction of nine months was achieved and in that time 92,000 sq m2 of building,
Appropriate concrete technology 592
or 1.2 million m3 of volume was enclosed with a structure comprising cast iron columns,
cast and wrought iron girders, timber girders and floors and glass. More than 2,000 men
were employed in Hyde Park at the peak of building and fit out.
Typical erection speeds for precast frames are nowadays 1000m2 per week per crane
with a site team of 8 men. Three teams with three cranes could, therefore, erect 92,000m2
in 30 weeks (fig.1). The time is the same, then as now, but now with the labour
requirements dramatically reduced. As another measure, a highbay warehouse using
double tees as wall and roof units to dimensions 156m16m 16m was erected in 1982 in
17 working days with a single crane. Thus, 2400m3 was enclosed per day (fig.2).
The Crystal Palace structure enclosed some 1.2 million m3 in 210 working days at a
rate of 5700m3 per day. The building rate at the Crystal Palace was clearly very fast, but
it would be easily achievable to this day with three or four experienced erection teams of
610 men each.
The achievement of speed in design and procurement was remarkable and it is difficult
to see how it could be matched now even for such a regular design to an established
system in a green field site.
Some of the details of the design were equally remarkable and show interesting
parallels with modern structures. The glazing support beams of the Crystal Palace were
composite, iron and timber. The timber compression block was profiled to collect water
from the glass and channel it to gutter beams which in turn passed rainwater to the
columns, which passed it on to drains through holes in their centres. Small grooves were
also cut in the side of the gutter block beneath the glazing to take condensation from the
underside of the glass down to the main gutters. In erection, the gutter beams were used
to support a wheeled gantry from which men placed and sealed the glass.
A similar detail has been used in cantilever roof beams in a number of grandstands in
the UK and elsewhere, designed by the Jan Bobrowski Partnership. In this case too the
gutter was used to support a wheeled erection gantry for the roof sheeting.
Appropriate concrete technology 594
Another similarity was more tragic, the floor of the Crystal Palace was boarded with
spaced timber planks so that every night it could easily be cleaned by brushing all the
litter through to the void beneath. It was the accumulation of this material that
contributed to the severity of the fire that eventually destroyed the building in 1936,
which by then had been re-located in South London.
Recent fires at Bradford Football Club and at Kings Cross Underground Station,
which both had a tragic loss of life, were similarly fed with the accumulation of debris
beneath the structures.
Adam Smiths final predication was on quality. This has already been mentioned, but
deserves re-emphasis; for both speed and quality of finish, precast cladding is hard to
match. Modern brick facades are becoming more ornamental and using prefabricated
techniques, brick panels with corbelling and bricks applied to horizontal soffits are all
possible. Combinations of natural and reconstructed stone in panellised form also now
have produced a new revival of attractive traditional elevations in buildings.
Far from being an offshoot of the regularity of dull architectural modernism the
prefabrication of concrete structures is developing into well engineered systems involving
external envelope, often using natural materials, structure and services. Off site quality is
now being integrated speedily into an attractive consistent whole.
CONCLUDING REMARKS
With the positive message given in this paper, why is precast concrete not used much
more widely? The answer to this question lies in a consideration of culture and cost.
The culture of our UK industry is for the design professions to have the main contact
with the client and to feed client wishes to contractors in the form of designs and
specifications. There is a clear self interest for this approach to continue. It is constructive
to consider methods of procurement in other countries where Clients talk directly to
Contractors and where the Designers are often in bureaux owned by the Contractors. The
debate is not straight forward, the design professions rightly claim that their
independence militates towards good design and good advice, although there is now an
increasing realisation that with the few structures now being built and fee competition in
design, this quality of service may not be universally provided.
On the cost side, the issue is more straightforward. For well integrated prefabrication
techniques and systems, bridge beams for example, in all economic climates
prefabrication is an economical option.
In the building industry the solution is slightly different. Despite the better production
from mechanised factories, the mature, skilled labour force is normally paid at a sensible
stable rate. On site, in times of boom, as was experienced in the late 1980s, labour rates
can literally go through the roof. This drove a great number of Contractors to favour
prefabrication often carried out by re-designing work already designed in steel or insitu
concrete. In hard times factory labour cannot easily be laid off and will earn similar rates
Factory & site production-A comparison 595
to those in the boom. Site labour costs, however, plummet and suddenly all sorts of work,
that for other reasons may be required to be precast, is converted to insitu.
The conclusion to all of the considerations in the paper is that the precast industry
would be best advised to promote its benefits directly to clients. The new awareness of
major clients of the benefits of engaging in dialogue and in partnering with suppliers is a
positive move in this direction.
DEVELOPMENT OF SELF-CURE
CONCRETE
RK Dhir
P Hewlett
T D Dyer
University of Dundee
UK
INTRODUCTION
It was decided that the best way to enhance water retention in concrete using a chemical
addition was by exploiting a deviation from Raoults law displayed by certain non-ideal
solutions. Raoults law states that the chance of finding a molecule of a component in the
vapour above an ideal solution is equal to the mole fraction of the component in the
solution itself [1], and is described by the equation below:
where PA is the partial vapour pressure of a component A in an ideal solution, the mole
fraction of A is represented by xA, and P*A is the vapour pressure of pure A. This
relationship is shown in Figure 1(a). This is only true of a closed system, and in the case
of concrete, PA will give an indication of the waters tendency to evaporate into the
atmosphere.
Appropriate concrete technology 598
Three criteria were identified for the effects of a chemical addition to a cement paste to
be described as self-curing :
When cement pastes containing the chemical are cured in air they should display
improved water retention with respect to identically cured controls.
As a result of any improvement in water retention, the degree of cement hydration
should also be enhanced.
An increase in the pastes compressive strength with respect to air cured controls should
be observed.
Table 1. Main characteristics of polymeric
materials
Main Characteristics of Polymeric Materials
Chemical Description Maximum Solubility at Functional Groups
20C
-OH -
Groups OGroups
A1 Synthetic. Av.M.wt. 200 100 Yes Yes
A2 Synthetic. Av.M.wt. 100 Yes Yes
1500
A3 Synthetic. Av.M.wt. 100 Yes Yes
5000
A4 Synthetic. Av.M.wt. 60 No Yes
Appropriate concrete technology 600
10000
A5 Synthetic. Av.M.wt. 65 Yes Yes
20000
A6 Natural. Av.M.wt. 5 Yes Yes
>20000
To establish which of the selected chemicals complied with these criteria, a series of
cement paste tests were carried out. The findings of these tests are shown in Figure 3,
which shows the improvements in water retention (determined by measuring weight-
loss), degree of hydration of the alite phase (measured using XRD) and compressive
strength observed in 28 day old Portland cement pastes.
It is clear from the results that an improvement in cement paste water retention does
not necessarily produce a proportional increase in the degree of hydration. Indeed it
would appear that the degree of alite hydration decreases with an increase in average
molecular weight. Improvements in compressive strength seem also to be, in many cases,
influenced by more than just increased water retention. Chemical A1 performs the role of
a self-cure admixture to the best extent and it is the performance of this chemical that will
be discussed subsequently. It has been established that there is, in fact, a group of
polymers similar to chemical A1 that all have similar properties.
Development of self-cure concrete 601
Shrinkage
Drying of concrete leads to shrinkage. Whilst concrete will always shrink to a certain
extent after placing as a result of the uptake of water during hydraulic cement reactions,
evaporation has the effect of greatly increasing shrinkage. This can lead to dimensional
problems and, especially at early ages, cracking. Figure 5 shows the dimensional changes
displayed by three Portland cement paste prisms up to an age of 180 days whilst kept in a
20C/60%RH environment. Two controls were used: one sealed in film for the duration
of the test and one exposed to the air. The self-cure specimen was made from paste
containing a 0.100M dosage of admixture.
For up to around the first 28 days the self-cure specimen displays shrinkage behaviour
similar to that of the film cured control. After this age its shrinkage increases at a greater
rate than the sealed specimen. This is presumably because, in the film cured specimen, by
28 days the majority of hydration reactions have ceased. Therefore, beyond this point
shrinkage is almost solely due to water evaporation, which will not occur in the sealed
prism. At all times it is clear that the performance of the self-cure specimen is better than
the air cured control.
Development of self-cure concrete 603
Surface Quality
The results of 10 minute initial surface absorption tests (ISAT) [3] conducted on self-cure
concrete at 28 days are shown in Figure 6. Concrete mixes containing 100% PC, 40%
granulated blast furnace slag (GBS)/60% PC and 30% pulverised fuel ash (PFA)/70% PC
were used. Results for both film and air cured controls are also shown. Film curing
provides the highest quality surface, although the higher dosage of self-cure admixture
gives very good results. Both self-cure mixes display better surface characteristics than
the air cured control.
It is possible that the microstructural changes produced by the self-cure admixture also
affect surface permeability. However, the majority of the improved surface quality is
most likely due to the improvement in cement hydration compared to the air cured
control.
The slices were dried in air, ground in a ball mill and analysed using both TG/DTA and
XRD techniques. Figure 7 shows the depths to which sodium chloride and Friedels salt
were detected. Friedels salt is the low chloride content calcium chloro-aluminate
(C3A.CaCl2.10H2O) and results from the reaction of chloride ions with the C3A and C4AF
cement phases and their hydration products [4]. This reaction renders the chloride ions
immobile at pH levels experienced in most uncarbonated, cement matrices [5]. It is clear
that the depth of penetration of chlorides into the pastes is much lower when film curing
is used, compared to the air cured control. However, results virtually identical to the film
cured control can also be seen in the self-cure pastes.
The depth to which carbonation has occurred such that the cement matrix is rendered
ineffective in pacifying corrosion can be assessed by spraying a solution of
phenolphthalein onto a freshly fractured concrete surface [6]. The results of carrying out
this procedure on concrete specimens kept in a carbonating environment for 2 weeks is
shown in Figure 8. Again it is apparent that the addition of the self-cure admixture has a
similar effect on carbonation behaviour as film curing.
Figure 9 shows the corrosion potential of steel reinforcement in cubes kept in the same
chloride spraying environment described previously. Corrosion potential was measured
using the half-cell potential technique [7]. As would be expected from the previous
findings, the film cured control displays the lowest potential to corrode, whilst the air
cured control shows the greatest. As self-cure admixture dosage increases, the potential
for reinforcement to corrode is lessened, due to the improved surface qualities and overall
reduction in permeability. It should be noted that in this case the concrete is not likely to
have carbonated to any great extent and that the observed corrosion potential is,
therefore, almost exclusively due to chloride ingress alone.
Appropriate concrete technology 606
Strength
Whilst, in general, the properties of self-cure concrete show an improvement in
comparison to air cured controls, an increase in compressive strength is not observed.
Since cement paste strength is increased by the addition of the admixture it is clear that it
is the presence of aggregate that is responsible for the lack of improvement in concrete
strength. The most likely reason for this is the change in calcium hydroxide morphology
discussed previously. It is reasonable to assume that the thinner calcium hydroxide
crystals will be weaker, and since it is these crystals which contributes to the bonding
between the cement paste and aggregate particles, the strength of the bond will be
lessened.
Whilst this is situation is obviously less than ideal it should be noted that the
improvement in surface quality far out measures the lack of improvement in strength: the
permeability per unit strength for self-cure concrete is always better than that of air cured
controls [8].
Development of self-cure concrete 607
CONCLUSIONS
1. The addition of a suitable chemical admixture can improve water retention in
hardening concrete and, as a consequence, increase the degree of hydration in cement
pastes cured in air compared to identically cured controls.
2. The manner in which self-cure pastes shrink is close to film cured controls up until
around 28 days. However, beyond this age the extent of shrinkage gradually
approaches that of the air cured control.
3. Improvements in the degree of hydration in air cured concrete have the effect of
improving surface quality. Consequently, both resistance to chloride ingress and
carbonation are improved in comparison to air cured controls, making self-cure
concrete effective in protection of steel reinforcement from corrosion.
4. Whilst an improvement in concrete strength with respect to air cured controls cannot
be expected in self-cure concrete, its other qualities far outweigh this disadvantage.
REFERENCES
1. Murrell. J.N. and Jenkins, A.D., Properties of Liquids and Solutions, 2nd Ed., John Wiley,
Chchester, 303pp, 1994.
2. Dhir, R.K., Hewlet, P.C., Lota, J.S. and Dyer, T.D., Journal of Materials and Structures, 27,
pp606, 1994.
3. Dhir,R.K., Hewlett, P.C. and Chan, Y.N., Magazine of Concrete Research, 39(141), pp183,
1987.
4. Suryavanshi, A.K., Scantlebury, J.D. and Lyon, S.B., The Binding of Chloride Ions by Sulphate
Resistant Portland Cement, Cement and Concrete Research, Vol.25, No.3, pp581, 1995.
Appropriate concrete technology 608
5. Tritthart, J., Chloride Binding in Cement. II. The Influence of the Hydroxide Concentration in
the Pore Solution of Hardened Cement Paste on Chloride Binding, Cement and Concrete
Research, Vol.19, pp683, 1989.
6. Bungey, J.H., The Testing of Concrete in Structures, 2nd Ed, Blackie, Glasgow, 228pp, 1983.
7. ASTM C87687, 1987, Standard Test Method for Half-Cell Potentials of Uncoated Reinforcing
Steel in Concrete.
8. Dhir, R.K., Hewlett, P.C., and Dyer, T.D., Cement and Concrete Research, Vol.25, No.6,
pp1153, 1995.
VIBRATORS FOR COMPACTING
CONCRETEA RADICAL LOOK
P F G Banfill
Heriot-Watt University
UK
INTRODUCTION
Vibration has long been recognised as necessary for effective compaction of concrete and
arising from much early work done on the effect of vibration on the properties of the
hardened concrete recommendations for current practice exist in many countries. The
Appropriate concrete technology 610
general consensus seems to be that the higher the acceleration at moderate frequencies
the more effective the vibration. However, it is much more fundamentally sound to base
recommendations on the performance of fresh concrete and advances in the
understanding of the rheology of fresh concrete have permitted this. This paper discusses
this approach and some important implications.
The importance of vibration as a means of compacting concrete has been recognised for a
very long time and much progress was made as a result of early studies of the effect of
vibration on hardened properties [1, 2, 3], which are embodied in current practice
recommendations e.g. [4]. The application of vibration, of an appropriate frequency and
amplitude, fluidifies the concrete enabling air pockets to be filled and air to bubble to the
surface. An initial gross increase in bulk density is observed, after which prolonged
vibration permits slow reorientation and packing of the particles in the concrete. The
main consequence of the increased density is higher strength. The air content of the
freshly mixed concrete may exceed 10% by volume and removal of this amount of air
from a test cube by vibration will double the measured compressive strength. Associated
with the strength increase is the achievement of the other desirable hardened properties.
With advances in the understanding of its rheology came the possibility of studying the
effect of vibration upon fresh concrete. It is now well established that fresh concrete, over
the range of shear rates important in practice, conforms to the Bingham model and that
the yield value and plastic viscosity can be measured conveniently in the two-point
workability apparatus [5]. In the test the torque required to turn an impeller at several
speeds in a bowl of fresh concrete is measured. From the resulting data a graph of torque
T against speed N shows a straight line relationship
T=g+hN
where g and h can be shown theoretically to be proportional to yield value and plastic
viscosity, respectively, the Bingham constants of the material. In practical terms, the fact
that fresh concrete has a yield value explains why it can support its own weight, as for
example in the slump test. Under vibration concrete loses that ability to support itself and
flows freely under its own weight to pass between reinforcing bars and fill the corners of
a mould. This suggests that the vibration either applies sufficient shear stress to exceed
the yield value or causes a change in the nature of the material such that the yield value is
reduced to a low level. Extensive work summarised by Tattersall [6] confirmed that the
latter was the case.
Vibrators for compacting concrete-A radical look 609
Tattersall and Baker [7] mounted the bowl of the two-point workability apparatus on an
electromagnetic vibrator and constructed flow curves with and without vibration over a
range of frequencies and amplitudes. They found that the flow curve was instantaneously
and reversibly altered by the application of vibration and that the yield value decreased to
zero (Figure 1), i.e. the constructed flow curve fits a power law model passing through
the origin. This enables the concrete to flow under its own weight. Consideration of the
region near the origin in Figure 1 suggests that, under vibration, and at low shear rates,
fresh concrete behaves as a Newtonian liquid. This means that the viscosity of the
vibrated concrete or its inverse, the fluidity, will be the key parameter in determining how
it will flow.
It was then possible to study the effects of vibration by considering changes in the ratio
v/ and the difference (v), where the subscript denotes the value determined under
vibration. They found that, subject to exceeding a small threshold frequency, the most
important characteristic of the vibration to be its peak velocity.
Constructing a flow curve from individual readings of torque and speed is
cumbersome and the data analysis is complex, so Tattersall and Baker concentrated on
the region near the origin of the flow curve where the slope of the curve gives a measure
of the fluidity. This suggested that a simpler apparatus could be successful and in a
comprehensive series of tests using a much more convenient vertical pipe apparatus they
determined the fluidity of the concrete from the rate of efflux from the bottom of the pipe
into the vibrating receptacle [8]. Under vibration, the height H of the column of concrete
in the pipe decreases with time as concrete flows from the bottom according to
dH/dt=bH
where F is an upper limit of frequency (between 250 and 500 Hz), Ao is a threshold
amplitude, which in turn depends on frequency, and c is a constant [8]. This equation
Appropriate concrete technology 612
shows that there is a threshold amplitude below which vibration has no effect in that yield
value is not reduced sufficiently for flow to occur, and that the effect of vibration
decreases markedly as frequency approaches some upper value F. If A>>Ao and f<<F
this equation reduces to
b=KAf=Kv
since peak velocity v is the product of amplitude and frequency. The constant K is
characteristic of a particular mix and may be termed the vibrational susceptibility. It may
be understood as follows. The larger the value of K, the larger the value of b and the
faster the concrete flows out of the pipe when it is subjected to vibration of peak velocity
v, as would be expected of concrete with a higher vibrational susceptibility.
Summary
The characteristic of a vibration determining its effect on fresh concrete is the peak
velocity and it is linearly related to the fluidity under vibration. The proportionality
constant, termed the vibrational susceptibility, is related to the workability of the concrete
and is greatest at low frequency and declines as frequency increases.
PRACTICAL IMPLICATIONS
Frequencies between 50 and 200 Hz are in common use for vibrators [4]. Table and
surface vibrators operate towards the lower end, formwork and internal vibrators towards
the top end. In contrast, Figure 3 shows that the vibrational susceptibility at 1630 Hz is
about four times that at 200 Hz. Thus, in principle, for a given mix and flow geometry the
fluidity and hence the rate of flow at low frequency could be four times that at high
frequency, provided the same peak velocity can be achieved. This could mean a fourfold
increase in the throughput of concrete in the vibratory compaction stage of concrete
placement. In practice, because peak velocity is the product of amplitude and frequency it
is only possible to achieve high velocity at low frequencies if the amplitude is very high
and therefore amplitude is likely to be the limiting factor in any equipment. The
significant effect on site productivity implied by the greater throughput could be realised
if concrete technologists and equipment manufacturers work together to carry out
vibratory compaction at low frequencies in equipment with maximum amplitude.
CONCLUSIONS
Application of vibration to fresh concrete removes its yield value and permits it to flow
under its own weight. The fluidity of the vibrated concrete is directly proportional to the
peak velocity of the vibration where the constant of proportionality, the vibrational
susceptibility, is a characteristic of the material which decreases as frequency increases.
Vibration equipment should therefore be operated at the combination of lowest frequency
and greatest amplitude possible in order to maximise the fluidity of the vibrated concrete.
Vibrators for compacting concrete-A radical look 613
ACKNOWLEDGEMENTS
I am grateful to Dr G H Tattersall for granting free access to his original results and to M
H Gharooni for assistance with processing the data.
REFERENCES
1. KIRKHAM, R.H.H. (1951) The compaction of concrete slabs by surface vibration: first series of
experiments. Mag. Conc. Res., Dec 1951, pp7991.
2. VENKATRAMAIAH, S. Measurement of work done in compacting a known weight of concrete
by vibration. Mag. Conc. Res., Jan 1951, pp8996.
3. CUSENS, A.R. (1958) The influence of amplitude and frequency in the compaction of concrete
by vibration. Mag. Conc. Res., Aug 1958, pp7986.
4. ACI COMMITTEE 309 Behaviour of fresh concrete during vibration. Journal of the American
Concrete Institute, Proceedings, 1981, 78(1), pp3653.
5. TATTERSALL, G H., and BANFILL, P F G. The rheology of fresh concrete. Pitman, 1983,
pp353.
6. TATTERSALL, G.H. Effect of vibration on the Theological properties of fresh cement pastes
and concretes. Rheology of fresh cement and concrete (Banfill, P.F.G. ed.) Spon 1991, pp323
337.
7. TATTERSALL, G.H. and BAKER, P.H. The effect of vibration on the rheological properties of
fresh concrete. Mag. Conc. Res. 1988, 40, 7989.
8. TATTERSALL, G.H. and BAKER, P.H. An investigation into the effect of vibration on the
workability of fresh concrete using a vertical pipe apparatus. Mag. Conc. Res., 1989, 41, 39.
9. WOOD, B.R.H. The effect of vibration on properties of fresh concrete, University of Liverpool,
B.A. Dissertation (unpublished), 1993.
10. MILLAR, W.S. Vibration of fresh concrete, University of Liverpool, B.A. Dissertation
(unpublished), 1994.
THE PLANNING AND DESIGN OF
CONCRETE MIXES FOR
TRANSPORTING, PLACING AND
FINISHING
G G T Masterton
R A Wilson
Babtie Group
UK
INTRODUCTION
Background
The report supersedes CIRIA Report 17, Handling and Transporting Concrete: Time and
Temperature Limitations [1] and is one of a series of state-of-the art guides dealing with
good practice in concrete construction in the U.K. and other countries with temperate
climates.
Since the publication of CIRIA Report 17 there have been significant developments in
construction techniques and in the constituent materials available for use in concrete.
These have given designers, contractors and concrete suppliers greater flexibility to
match concrete mix designs with the particular requirements of modern transporting and
placing methods in a wide range of locations and ambient conditions.
Concrete has a proven track record as a successful material in the construction of the
highest towers, the longest tunnels, the deepest shafts and the largest of manmade
structures and in conditions as extreme as the Arctic tundra and Middle Eastern desert.
With adequate planning in all stages of the construction process there are now few
conceivable limitations on the use of concrete.
The report provides guidance to designers, specifiers, planners, concrete suppliers and
contractors to help ensure that the whole concreting process in planned and implemented
in such a way that the required quality is achieved. Guidance is provided on the fresh
state Theological properties required of concrete for various operational conditions, and
how these properties can be controlled and measured.
Scope
The report provides guidance on the properties on concrete appropriate to a range of
construction methods, operations and environments and on how these properties can be
achieved in practice. Much of the advice will be relevant to other countries, particularly
those with temperate climates, but in more extreme climates specialist references should
be consulted[2], or advice sought.
Familiarity with the contents of BS 5328: Parts 1, 2, 3 and 4[36] is desirable.
Specialist techniques and applications such as roller compacted concrete [7],
underwater concreting [8], road paving [9][10], slip-forming, microsilica concrete [11] and
sprayed concrete[12] are not dealt with in the report.
Design
The procurement and production of concrete which is appropriate for requirements is an
integrated process involving the designer, the construction planner and the concrete
producer, see Figure 1. Similarly, the specification of concrete must also be integrated
between the parties. In practice, the designer begins the process by defining requirements
Appropriate concrete technology 618
and the construction planner, in conjunction with the producer, completes the
specification, taking account of construction methods, environmental conditions and
economic use of materials.
Method of Construction
Experience or advice should be used to anticipate the likely method of construction and
take due account in design, making it clear in the specification if any assumptions have
been made with affect the permanent works. In some cases, the imposition of constraints
on the construction process may be unavoidable.
The designer should however be receptive to alternative proposals by contractors and
be prepared to consider the implications of methods which may not have been anticipated
at the design stage.
Environmental Conditions
The designer should be aware that changes in environmental conditions affect the
behaviour of fresh concrete during transporting and placing and that corresponding
refinements in mix design are desirable to achieve consistency in properties at the point
of discharge into the formwork.
Appropriate concrete technology 620
COMPLIANCE TESTING
CONSTRUCTION PLANNING
The designer begins the specification process for concrete by defining design
requirements. The construction planner completes the specification by taking due
consideration of the chosen construction methods, plant, personnel, the site location and
environmental conditions.
The planning and design of concrete mixes for transporting, placing and finishing 621
The planner must ensure that plant in the supply chain of productiontransport
placingfinishingcuring, is matched to the supply rate necessary to achieve continuity
of placing with minimal risk of cold joints.
In all concreting operations continuity of supply at a rate to match the capacity of
other elements in the supply chain is essential to success. Communication with the
concrete producer is essential, with advance notice being given, particularly for larger
pours or those likely to exceed normal working hours. Some 90% of in-situ concrete in
UK is supplied ready-mixed. On-site batching is unlikely to be economic for volumes of
less than 20,000 m3 and production rates less than 200 m3 per day [14]. There may be
circumstances, e.g. where access is unusually difficult, where onsite batching of smaller
production volumes and rates is economically feasible.
Truck mixers are the usual delivery vehicle for normal concretes. Truck mixers
provide the facility for keeping the concrete agitated during long haul distances. They
also allow remixing of the fresh concrete with superplasticisers added on site. In many
situations it is possible to position the truck alongside the work and discharge directly
into place, for which a concrete of high workability would be appropriate. There must be
clear access that is firm enough to support a truck weighing, generally about 24 tonnes.
Particular attention should be paid to temporary works design when truck mixers are
required to be parked close to the edge of an excavation. Where the concrete is required
to be placed above ground level it may be discharged into skips and handled by crane.
Alternatively, the concrete can be discharged directly into the hopper of a mobile
concrete pump. The most common truck mixer used in the UK holds 6m3 of concrete,
although 4m3 truck mixes are available in some areas and may be useful where access is
restricted.
Placing by crane and skip enables concrete to be placed at the lowest practical slump
consistent with compaction by vibration, and remains a popular technique. Typical rates
of placing are 8123/hour. Selection of the appropriate crane to provide safe access to all
parts of the pour is essential.
In contrast to the low rates of placing concrete using a crane and skip, placing rates of
50m3/hour are regularly achieved by pumping, and rates up to 100 m3/hour are
possible[13]. Pumping has the following advantages:
removes the need for staging, hoists, dumpers, skips etc
overcomes access problems
fast, therefore can be cost effective
pour rates can be maintained at heights
Pumping is one of the most efficient ways of moving concrete. Most pumps can be
continuously fed from a truck mixer with a load of 6m3 (or even two together), and the
concrete pumped in 56 minutes.
The high rates of placement require careful planning to ensure adequate resources of
operatives and plant are available to compact and finish the concrete.
Other methods of placing include chutes, conveyors, drop pipe or trunking and tremie
placing.
The Report considers several other planning aspects including compaction, access,
duration or environmental constraints and finishing.
Appropriate concrete technology 622
The key to successful mix design is to comply with all specified criteria while attaining a
workable concrete which satisfies the job-specific requirements of transporting, placing
and finishing. The skill of the mix designer is in achieving these requirements in the most
economic way making use of the range of materials at his disposal.
Characteristics to be considered
In addition to the designers specified parameters, the key properties to be selected and
controlled for efficient concrete construction are:
workability
stiffening time
cohesiveness
The planning and design of concrete mixes for transporting, placing and finishing 623
The required workability and cohesiveness at the time of placing are related to the nature
of the pour and the placing equipment. Initial workability and cohesiveness may require
to be modified to suit the method of transporting, the ambient conditions and the transit
times of the fresh concrete. The required stiffening time must take account of the planned
time between successive lifts, and the effect on formwork pressures and finishing
methods.
All properties are likely to be affected by environmental conditions. Other
characteristics such as heat evolution are particularly important in large volume pours and
are discussed in CIRIA Report 135[13].
Section 4.3 of the Report discusses the implications of transporting, placing, access
duration and environment, and finishing methods on the desired characteristics of the
fresh concrete, providing guidelines for pumping both normal and lightweight concrete,
tremie placing, and for varying ambient conditions. Measures to reduce bleeding, crazing
and cracking are also suggested.
Section 4.4 of the Report discusses the influences of mix constituents on the fresh
concrete characteristics and describes ways in which these can be tailored to match
specific requirements and environmental conditions. The approach adopted is to consider
the effect of changes in materials or conditions from an initial reference condition.
The anticipation of all potential variations in factors which influence the behaviour of
fresh concrete is, however, not generally necessary. Concrete is a tolerant material and in
its fresh state a particular mix will perform adequately over a range of varying influences.
However, its performance can be improved if the major influences are anticipated and
appropriate refinements made to mix design.
Workability
A primary influence on initial workability is water content. Given that strength is likely
to be a basic governing requirement and that there is a correlation between strength and
water/cement ratio, changes in water content will require corresponding changes in
cement content. The initial mix design process will be based on a resolution of strength
and workability requirements based on this relationship.
The following are discussed in detail in this section of the Report:
water and retempering
cement and blended cements
sand
admixtures and their effect on workability
temperature
cohesiveness
stiffening time and control with admixtures.
Appropriate concrete technology 624
Temperature 10C 10
Increased to 25C 5
30C 15
40C 30
Period of agitation Up to h after Increased to between: 5
after mixing adding water 1h
12h 10
23h 15
33h 20
*Note: pfa and ggbs may have no, or very little, water reducing properties if water reducing
admixtures are also used.
There are many influences on the workability of fresh concrete, each of which can
potentially be used to refine the characteristics of concrete as requirements, or ambient
conditions, change. There is a shortage of data to allow firm guidelines to be given to
enable mix designers to quantify these refinements, and the many interacting variables
and changes in local conditions and site-specific requirements may make it difficult for
anything other than general guidelines to be given. With this caveat, Table 1 gives a
summary of the main factors affecting changes to the water requirement of concrete for
constant slump, and provides indicative adjustments in water quantity to a notional
reference condition.
CONCLUDING REMARKS
to be pumped into place. The mix designer adopts a composite cement with ggbs to
provide better control of air. See Tables 2 and 3.
The problem, essentially, is how to sustain the 4.5% entrained air for such a long haul
time, particularly under winter conditions and originally with such a high cement content
(550 kg/m3). Reducing the cement content to 440 kg/m3 helps, together with the more
angular shape of the granite coarse aggregate and the sharper nature of the sand. The
cohesiveness of the mix will be improved by the air-entrainment, but while the cement
water paste would have good cohesiveness, the sand proportion has to be raised to
produce sufficient mortar to give the mix a pumpable quality. It is the combination of
reducing the cement and water contents, balanced by raising the sand content, leading in
turn to a reduction in coarse aggregate that helps reduce what is potentially a very sticky
mix. Care needs to be taken with the trial mixes and in monitoring the effect of mix
temperature on the amount of air entrained.
Table 2: Case Study 4 Air-entrained Concrete in
Bridge Deck
Preliminary MixDesign Limitations kg/m3
Cement PC 42.5 550
pfa
ggbs
Aggregate Nominal max. size 20mm
Sand BS882 425
Coarse BS882 1,095
Admixture
Water (Free) 210
Density (Fresh) 2,280
Other Provision for water addition on site MUST be Winter conditions favour long
made to adjust workability for pumping. haul time of 45 minutes.
General Considerations
Ready Site Type Reinfor Size Acces Climatic Speed Com Finishing Frame
Mix Tran of cement of s to Cond of pactio n Consid work
or spor t Elem congestio Pour Form ition s Const M erations Comp
Site ent & n (m3) work ruction ethod lexity
Mix Thick
ness
Ready Pump Bridge Congested 45 From North Normal Internal Wood Intricate
Mix Deck Top European Vibration Float
Winter
Special Considerations
Deep Large Hot & Dry Severe Winter Long Top Slip Early
Lifts Volume Conditions Conditions Hall Forming Forming Striking
Pours Time
No No No Yes 45 No No No
Appropriate concrete technology 628
Minutes
REFERENCES
1. BROOK. K M. Handling and transporting concrete: time and temperature limitations. CIRIA
REPORT 17, 1970.
2. CIRIA. The CIRIA guide to concrete construction in the Gulf Region. Special Publication 31,
1984.
3. BRITISH STANDARDS INSTITUTION. Concrete. Part 1: Guide to specifying concrete.
BS5328:1991.
4. BRITISH STANDARDS INSTITUTION. Concrete. Part 2: Methods for specifying concrete
mixes. BS5328:1991.
5. BRITISH STANDARDS INSTITUTION. Concrete. Part 3: Specification for the procedures to
be used in producing and transporting concrete.
6. BRITISH STANDARDS INSTITUTION. Concrete. Part 4: Specification for the procedures to
be used in sampling, testing and assessing compliance of concrete. BS5328:1990.
7. COMMISSION INTERNATIONALE DES GRANDES BARRAGES (ICOLD). Roller
Compacted Concrete for Gravity Dams. 1989.
8. THE CONCRETE SOCIETY. Underwater Concreting. Technical Report No. 35. 1990.
9. MAYHEW, H C and HARDING, H M. Thickness design of concrete roads. Transport Research
Laboratory Research Report 87.
10. KENNEDY, J. Cement-bound materials for sub-bases and road-bases. BRITISH CEMENT
ASSOCIATION Publication No. 46.027.1989.
11. THE CONCRETE SOCIETY. Microsilica in concrete. Technical Report No. 41, 1993.
12. EFNARC. Specification for sprayed concrete (Final Draft). European Federation of National
Associations of Specialist Contractors and Materials Suppliers for the Construction Industry
1993.
13. BAMFORTH, P B and PRICE, W F. Construction of deep lifts and large volume pours. CIRIA
Report 135, 1995. READY-MIXED CONCRETE BUREAU. The essential ingredient:
production and transport. British Cement Association. Publication 97.326. 1993.
15. JACKSON, N. Civil engineering materials, 3rd edition. MacMillan (London), 1986.
ROLLER COMPACTED CONCRETE:
DEVELOPMENT AND USE IN
QUEENSLAND, AUSTRALIA
H Reid
Hilton Reid & Partners
Australia
INTRODUCTION
use high quality aggregates, are mixed in continuous mixing plants and placed using
heavy duty paving machines.
At this time the construction of RCC pavements was being undertaken elsewhere in
Australia (1) and RCC was being used for dam construction (2). A review of published
material from other sources revealed that no consensus had been achieved as to mix
design, testing procedures and construction techniques, particularly in relation to RCC
used in pavements. The Author therefore decided to apply his experience with CTB
pavements to independently developing appropriate techniques for RCC pavements. To
date 16 projects or project stages have been undertaken in Queensland and these have
enabled most aspects of RCC pavement design and construction to be investigated and
satisfactorily resolved.
RCC MATERIALS
Cementitious Material
RCC mix designs having flyash contents varying from none (7) to 60% (8) have been
reported. Cement Type GB (Flyash Blend), composed of 75% Ordinary Portland Cement
Roller compacted concrete: Development and use in Queensland, Australia 631
and 25% Flyash, is readily available in Queensland at a cost about 10% below that of
Portland Cement. Because the use of flyash has the beneficial effect of reducing early
thermal rise and consequent shrinkage Type GB cement is normally used in Queensland
for this, and economic, reasons.
Aggregates
The Queensland Department of Transport Specification MRS11.05 (9) covers aggregate
for granular and CTB pavements. The highest quality fine crushed rock, Type 1.1, is
required to meet stringent aggregate strength and stability requirements, and is produced
by quarries in parallel with concrete aggregates.
The grading of Type 1 aggregates has been arrived at as being that required to
maximise the density of the compacted material i.e. it is the grading that allows
progressively smaller particles to fill the voids left by the larger particles. This grading
and that of a typical combined grading curve for conventional concrete aggregates is
shown in Table 1. Type 1 is finer on all sieves and contains a significant proportion of
material finer than 75 microns.
Since 1987 Type 1.1 aggregates from a number of quarries have been tested to
determine concrete strength and significant differences have been recorded. As the
soundness of the aggregates are similar, and certainly high enough not to be a factor in
determining compressive strength, some characteristics of the fines is the variable.
Aggregate with a grading closer to conventional concrete aggregate has been trialed and
this grading is also shown in Table 1.
Table 1 Grading of RCC and Concrete Aggregates
Sieve Size mm Type 1 Percentage Passing Brisbane 40MPa Trial Blend
26 100 100 100
19 87 100 99
9.5 69 58 68
4.75 54 37 42
2.36 39 32 27
0.425 18 10 11
0.075 7 0 4
This trial however established that aggregate grading has a strong influence on fresh
material behaviour, the trialed material exhibiting a stickiness that made it difficult to
place. Therefore the suitability of the grading of RCC aggregate is determined by a need
for the grading to be such that the cement paste is sufficiently dispersed over the surfaces
of the various particles in the mix for the non-cohesive characteristic to be retained. The
Type 1 grading, by coincidence, appears to be close to ideal.
It should be noted that in Type 1.1 aggregate the presence of clay fine particles is
acceptable provided that the plasticity index of the fraction passing 0.425mm is less that
Appropriate concrete technology 632
4. However, conventional concrete experience has shown that clay fines are detrimental
to strength. The desirable material for fine particles appears therefore to be rock flour,
material produced only by the crushing of a stone meeting the quality requirement of the
course aggregates.
MIXING OF RCC
The Aran continuous mixer is used extensively for both CTB and granular pavement
work in Australia. The earlier model, the ASR200, is readily available and has been used
on most RCC projects. Although later model machines are very sophisticated,
incorporating belt weighers and recording electronics, the performance of the machine is
best appreciated by realising that it essentially volumetric in its operation, as the feed
rates of aggregate and cement are determined by the opening size and belt speed of the
respective feeders.
The detention time of the material in the mixing chamber is low (about 1020
seconds) and hence the efficiency and uniformity of mixing needs to be verified. Two
examples of the performance are given in Table 2, the first being for an RCC project and
the second being a performance trial carried out for high strength, zero slump, concrete.
The results in Table 2 clearly indicate that the Aran pugmill produces consistent mix and
hence any variations that may occur in a project are not caused by deficiencies in mixing.
Table 2 Uniformity of Mixing of Aran Continuous
Mixer.
Model No. of Samples Mean Standard
(1 Pair Cylinders) Mpa Deviation
ASR200 8 32.4 1.64
ASR250X 4 65.25 2.01
Pugmills can be calibrated only approximately by means of setting the volumetric feed
rate of the cement and aggregate. There is the need therefore to have an accurate test to
monitor the actual cement content of the mix. The Queensland Department of Transport
has for many years used the Cement Content of Cement Treated Material (Heat of
Neutralisation) Test Method Q116B-1978 (10) for this purpose. The test method
determines the cement content by measuring the heat rise caused by the reaction between
the cement in a sample of mixed material and a buffer solution of a blend of sodium
acetate and acetic acid. A curve of cement content versus temperature rise is prepared in
the laboratory using mixes with known cement contents, and this curve is then used in the
field to monitor the work.
There have been some variations in results reported using this method. However, in
the absence of a practical alternative, this has been developed for RCC use. Because of
Roller compacted concrete: Development and use in Queensland, Australia 633
the higher cement content of RCC as compared to CTB, more buffer is required to ensure
that there is sufficient for the reaction to consume all of the cement. It has been
established that the original Department of Transport buffer composition, being 150
g/litre of sodium acetate and 240 g/litre of acetic acid is suitable. A number of different
ratios of buffer and mix have been investigated with the aim of ensuring that the test is
accurate, while minimising the cost of testing, the chemicals being rather expensive.
The proportions that allow cement contents of between 10% and 20% (weight
cementitious/weight dry aggregate) to be accurately measured has been established at:
RCC Mixture 2 Kg
Buffer 3 litres
In the Department of Transport Test Method the initial temperature is taken as the mean
of the buffer and mixture temperatures. However this does not take into account the
specific heat of the components, which can be considered as follows:
Tm (MgSHg+MSSHS)=TgMgSHg+TsMsSHs
Where Tm=mean temperature of combined material
Tg=RCC mixture temperature
Ts=Buffer temperature
Mg=Mass of RCC mixture
Ms=Mass of Buffer
Shg=Specific heat of RCC
Shs=Specific heat of Buffer
This equation has been confirmed experimentally in the laboratory by adding gravel to
buffer at varying temperatures and recording the stabilised temperature of the mixture,
with the constant found to be 0.15.
A typical laboratory prepared calibration curve is shown in Figure 1. Test points with
widely differing gravel and buffer temperatures are shown in Table 3, which clearly
demonstrates the accuracy of the test method. These results, and experience from
preparing calibration curves for a number of projects, has confirmed that the test is
accurate to within 0.5% cement content. This represents about 10Kg of cementitious
material per cubic meter of compacted RCC which is certainly sufficiently accurate for
practical purposes.
Appropriate concrete technology 634
STRENGTH/MOISTORE/DENSITY RELATIONSHIP
Concrete cylinders for laboratory trials and construction control purposes are prepared by
compacting RCC into 300mm high by 150mm diameter concrete cylinders using a
Standard Compaction hammer and 100% Standard compactive effort (5 layers 80
blows/layer). Standard Compaction has been selected as it accords with general
Department of Transport practice and makes the process easier to perform.
It has been established that the best placement moisture content is between 1% to
1.5% below OMC Standard. A laboratory trial to determine the effect of varying moisture
contents and compaction levels on strength has been conducted and the results are
Roller compacted concrete: Development and use in Queensland, Australia 635
presented in Figure 2. It can be seen from Figure 2 that the desirable placement moisture
content coincides with that which gives optimum strength, and that strength is sensitive to
density.
Figure 2 Strength/Moisture
Content/Density
MOISTURE CONTENT
Because the accurate measurement of the moisture content of the mix is critical to
controlling both water addition and measuring pavement density (expressed as a Relative
Dry Density) it needs to be determined how the hydration of the cement affects the
apparent moisture content. Department of Transport practice for CTB work is to define
time limits within which tests must be conducted. In RCC work in Queensland the use of
retarder has been found to be necessary to allow finishing to be completed prior to initial
set. Therefore the effect of retarder on material behaviour and apparent moisture content
has been investigated.
Laboratory moisture content testing has been carried out with samples incorporating
retarder. These tests have shown that retarder is effective in preventing apparent moisture
loss due to cement hydration for a period of at least 4 hours from mixing. Using a
microwave oven very good correlations between added and calculated moisture contents
have been obtained. Further, it has been established that a set retarding admixture (not a
water reducing admixture) has no affect on the MDD/OMC curve and the behaviour of
the material in the field is unaffected.
Appropriate concrete technology 636
CONSTRUCTION
Moisture Control
It is necessary to maintain the moisture content within about +/0.5% of the target
moisture content. The moisture content in the aggregate stockpile may contain a
significant proportion of the total requirements and can easily vary by more than the
desirable range for the mixed RCC. Although, with experience, it is possible to manually
assess the fresh RCC moisture content reasonably accurately it has proved difficult to
keep within the desirable range when the aggregate stockpile moisture content varies. By
premixing a variable stockpile prior to loading into the pugmill the need for continual
adjustments to the amount of added water can be eliminated.
Cement Content
Prior to commencing work on site (or when a new batch of buffer is prepared) a check
point on the calibration curve is obtained on site by testing a mixture prepared from
known weights of on-site materials. This provides fail safe verification of the test
procedure. The test takes about 15 minutes to perform and because of the direct
relationship between volumetric and mass feed rates in a pugmill only one test is required
to establish the correct ratio of settings between cement and aggregate and one further
test to confirm the cement content after adjustment of the settings. Further cement content
testing is carried out throughout production as part of a Quality Assurance program.
Density
Department of Transport practice is to accept a particular lot of work in accordance with
a Characteristic Value formula which in effect requires all results to be above the
nominated level. Sand replacement density tests (with microwave oven drying) or nuclear
densometer testing gives rapid test results, enabling rolling to be adjusted as necessary so
that the minimum density of 100% RDD Standard is achieved.
Concrete Cylinders
Cylinders are prepared on site from fresh mix. Because the moisture content is below
Standard Optimum, densities are marginally below 100% RDD and hence cylinder
strengths are below the in-situ strengths. This is therefore a very simple and conservative
means of strength verification. Provided the sample is protected from drying out during
cylinder preparation minimal density and strength differences occur between the two
cylinders in a sample.
CONCLUSIONS
1. The optimum grading for RCC aggregates has been established.
Roller compacted concrete: Development and use in Queensland, Australia 637
2. The reason for strength variation between different sources needs further investigation.
3. An accurate method for determining the cement content of the fresh mix has been
developed.
4. The desirable moisture content for placement and that for maximum strength coincide.
5. During construction simple techniques and tests are used to control and verify the
quality of the finished pavement.
REFERENCES
1. BRETT, D.M. RCC Pavement in Tasmania, Australia. Proceedings, Roller Compacted Concrete
II, ASCE, San Diego, California, 1988, pp 369379.
2. FORBES, B.A. RCC Dam in Australia. Proceedings, Roller Compacted Concrete II, ASCE, San
Diego, California, 1988, pp 323339.
3. GOMEZ, J AND SCHOLER, C.F. Fatigue Strength of Roller Compacted Concrete. Proceedings
4th International Conference on Concrete Pavement Design and Rehabilitation, Purdue
University, pp 505518.
4. PORTLAND CEMENT ASSOCIATION. Structural Design of Roller Compacted Concrete for
Industrial Pavements IS233.01p. 1987 pp 8
5. ANDERSSON, R AND CARLSSON, U. Interaction between Mix Design, Construction and
Properties for RCCP. Proceedings, 6th International Symposium on Concrete Roads, Madrid,
1990, pp 2939.
6. NATIONAL ASSOCIATION OF AUSTRALIAN ROAD AUTHORITIES. Pavement Design,
1987.
7. NAIMAK, T. Use of Roller Compacted Concrete in Norway Proceedings, 6th International
Symposium on Concrete Roads, Madrid, 1990, pp 115123.
8. JOFRE, C, VALUERO, J, LOPEZ-PERONA, R, GARCIA-ROJO, A AND ORTIZ, F. RCC
Pavements for Motorways and Main Highways: The Spanish Practice. Proceedings, 6th
International Symposium on Concrete Roads, Madrid, 1990 pp 155165.
9. QUEENSLAND DEPARTMENT OF TRANSPORT. Standard Specifications Second Edition,
1993.
10. QUEENSLAND DEPARTMENT OF TRANSPORT. Materials Testing Manual Volume 1.
INDEX OF AUTHORS
Abdul-Salam M A 207212
Adhikari R P 115122
Alldred J C 391400
Arnaouti C 105114
Arshad A A H 507516
Arya C 411420
Banfill P F G 607612
Baradan B 369374
Barbosa N P 473482
Barnard D P 131140
Beeby A W 495506
Bellers R 575586
Blight G F 181188
Boer G P L den 7586
Brandt A M 421430
Cabrillac R 225232
Cairns J 401410
Carter W G 123130
Chana P S 545552
Chatveera B 299308
Chishty M S 197206
Choudhary M A 197206
Clarke J L 453462
Clarke M A 255264
Covarrubias J P T 213224
Cox D W 525536
Davies G 507516
Desai S B 8594 553564
Dhir R K 595606
Dyer T D 595606
Economou N 431442
Elliot K S 507516
Ellis C 463472
Elser M 443454
Fathibitaraf F 495506
Feng N-Q 329336
Index of authors 639
Ftikos Ch 361368
Gebauer J 149160
Gebler S H 517524
Gemert D van 575586
Georgiades A 361368
Ghavami K 473482
Hassani A 525536
Hewlett P C 595606
Junior H C L 473482
Karayannis C G 343350
Kearsley E P 233244
Knight D 3748
Konsta M S 343350
Krige G J 189196
Krivenko P V 309314
Kruger R A 171180
Kucharska L 421430
Lemelin D R 537544
Leonovich S N 95104
Leventis D 431442
Loo Y C 565574
Lopati J 6168
MacCraith S 1322
Madrid G G 161170
Malami Ch 281288
Malou Z 225232
Marines J 361368
Marsh B K 4960
Masterton G G T 613626
Min A J S 473482
Morris J 171180 181188
Narayanan R S 6974
Newman J B 141148
Nimityongskul P 299308
Nixon P J 4960
ORegan P 453462
Oltean-Dumbrava C 2330 123130
Ortzar J de D S 3136
Osborne A 141148
Papayianni I 431442
Index of authors 640
Pascual J M D 213224
Pelaez L G 161170
Philippou Th 281288
Pirathapan G 411420
Pullen A D 141148
Quillin K C 289298
Raji N K 105114
Raksha L V 309314
Raksha V A 309314
Rankine R G D 189196
Reid H 627636
Rensbergen E van 575586
Retamal J P 3136
RigopoulouV K 281288
Robu I 337342
Sabuni E I 315328
Saje F 6168
Samarin A 265280
Sarja A E 375390
Seegebrecht G W 517524
Sideris K K 343350
Solminihac H de T 3136
Somerville G 112
Stejskal B G 517524
Stoelhorst D 7584
Stroeven P 315328
Subedi N K 553564
Swann L 3748
Takheklambam S 565574
Tassios T P 483494
Taylor HP J 587594
Torrent R 149160
Tsakiridis T 281288
Tschegg E K 443452
Tschegg-Stanzl S E 443452
Ulas E M 463472
Uzoegbo H C 351360
Wimpenny D E 245254
Wittmann F 149160
Wong A F L 105114
Xantakas G 431442
Yao B Z 565574
This index has been compiled from the keywords assigned to the papers, edited and
extended as appropriate. The page references are to the first page of the relevant paper.
Bamboo 473
Bar-sizing 391
Beam to column connections 565
Bearing capacity 537
Bending 453
Bond 401
Brick powder 369
Bridge beams 507
Bridges 453
arch 189
continuous integral 507
Brief 49
Building technology 61, 575
blending 315
content 627
low energy 2811
hydration equation 343
high alumina 289
mortar 189
Ordinary Portland 171, 181, 245, 337, 361
Portland 213, 329
Portland pozzolana 213
pozzolanic 213, 343
rapid-hardening portland 525
rejuvenated 265
sulfoaluminate 281
Centre bars 553
Changeable structural system 61
Chemical activation 265
Chloride 207, 595
Cladding 587
Compressive strength 207, 225
equation 343
Concrete 1, 13, 37, 105, 115, 351
aerated 225
alkaline 309
appearance 13
applications 149
beams 473
blocks 161
fibres 443
heavyweight 96
innovative 31
lightweight 233
medium strength 463
mixes 613
performance 299, 329
prestressed 507
polypropylene fibre 443
product market 31
properties 149, 483
ready mixed 141
strength 463
supercover 411
versatility 149
Concreting
cold weather 213
hot weather 213
planning of 613
Condensed Silica Fume 233, 337, 421
Congested reinforcement 517
Consolidation 517
Construction 1, 23, 115, 181, 197, 483, 507
Subject index 644
Dampening 309
Deflection 401
Deformation 565
Density 245, 627
Design 1, 37, 49, 207, 483, 495
aids 75
conceptual 483
guidance 545
life 49
methods 375
practicality 37
process 75
srvice life 375
Differential settlement 207
Discrete choice logit model 31
Durability 1, 37, 49, 171, 207, 375
Fibre 351
aramid 453
carbon 453
composites 453
natural 431
non-ferrous 375
reinforced concrete 375
reinforced cement 431
pitch carbon 421
polymeric 431
Finish 13, 613
Finite element analysis 189
non-linear 553
Flat slab construction 545
Flexural 431, 463
toughness 421
Flowing concrete 517
Fluidity 329
Fly ash 171, 329, 337, 361, 369, 431
Foamcrete 233
Foamed 245
Force and energetic parameters 95
Foundation
caisson 537
failure 123
superficial 537
Fracture mechanics 95
Fresh concrete 607
heat of 281
of cement 361
number 343
Labour
intensive 189
law 115
Laminated solid 525
Lapped joint 401
Latex 309
Legal
requirements 123
liabilities 23
Limestone 207
Limit state 13
Linear regression 343
Load
arrangements 495
cyclic 565
repeated 565
Performance
concept 23
criteria 1, 23
quotient 463
thermal 171
Permeability 595
Personnel protective equipment 115
Pulversized-fuel Ash 171, 181, 233
Phase equilibria 289
Pipejacking 255
Placing 613
Polypropylene fibre concrete 443
Pore size distribution 361
Porosity 281, 309
introduced 225
Pozzolanic reaction 369
Pozzolans 361
Precast concrete 565, 587
Prefabrication 587
Prevention 411
Protein foam 233
Punching shear 545
Qualifications 131
Quality
control 587
improvements 197
procedures 123
Rebar 391
Recycling 265
Reinforced concrete 453, 473, 565
frames 495
beams 553
Reinforcement 391, 401
Relative toughness index 421
Residential development 123
Resins 453
Rheology 61, 607
Rice husk ash 315
modified 299
Rock pockets 517
Roofing tiles 351
Rubber pad 525
Subject index 648
Safety 115
factors 495
Segmental lining 255
Self-cure 595
Self-help projects 161
Service
holes 545
life 49
Serviceability 495
Setting kinetics 225
Shale Ash 329
Shear
bond 463
resistance 553
Shelter 171
Shrinkage 61, 595
Simply supported 507
Site workers 131
Slabs 473
Slag 309
Slipform 255
Slump 517
Soil 181
reaction modulus 537
pressures 537
Specific fracture energy 443
Specification 141, 613
Stabilisation 181
Standards 141
harmonisation 141
national 85
Statistical evaluation 31
Steel
fibre concrete 443
high tensile 463
stainless 391
Stiffening time 613
Strain 105
Stratlingite 289
Strength 1, 37, 95, 213, 245, 565
cube 189, 207
cylinder 627
high early 281
tensile 431
ultimate 105
Structural
analysis 495
design rules 85
steel shearheads 545
Subject index 649
Time dependent
effects 105
response 61
Tools 149
Training 131
Transmission lines 537
Transport 613
Tunnelling 255
Tunnels 453
immersed 225