BASICS

CONSTRUCTION
LOADBEARING
SYSTEMS
Alfred Meistermann
Alfred Meistermann

Loadbearing
Systems
Alfred Meistermann
Bert Bielefeld - Sebastian El Khouli

Entwurfsidee
Loadbearing
Systems

Birkhäuser
BIRKHÄUSER
Basel
BASEL
 Contents
Foreword _7

Loads and forces _9

Loadbearing structures and statics _9
Forces _9
Statical system _10
External forces _11
Internal forces _18
Dimensioning _28

Structural elements _35

Cantilever arm, simply supported beam,
simply supported beam with cantilever arm _35
Continuous beam _37
Articulated beam _38
Trussed beam _40
Lattice _41
Slab _44
Column _46
Cable _49
Arch _51
Frame _54

Loadbearing structures _59

Solid construction _59
Skeleton construction _62
Reinforcement _68
Halls _70
Plate structures _74
Foundations _77

In conclusion _81

Appendix _82
Pre-dimensioning formulae _82
Literature _84
Picture credits _85
 Foreword
When constructing a building, we need to know how its structural
properties function. Loadbearing elements can be the dominant features
of the design, or simply an invisible substructure – but a building is ­always
based on its loadbearing structure. It holds the building together, distrib-
utes loads into the ground, and guarantees stability. An understanding
of loadbearing – its structural principles and the specific qualities of
­individual loadbearing systems – is fundamental to applying these prin-
ciples sensibly in the design process and developing a solution that suits
the materials and the construction method.

It is often difficult, particularly at the beginning of a course when

there is so much new material to assimilate, to work one’s way into the
complexities of statics and loadbearing theory. Basics Loadbearing
­Systems bridges the fields of architecture and civil engineering and
­explains the fundamentals of loadbearing structure theory simply, com-
prehensibly and chronologically. To help general understanding, the
­author first explains the loads and forces occurring in a building using
examples and simple contexts. He introduces typical loadbearing struc-
tural elements and shows loadbearing systems and structures for the
­different building types that planners can use for their designs. The com-
pact knowledge conveyed here makes it possible for students to work
with loadbearing structures in an integrated way, and thus be able to
­design creatively.

Bert Bielefeld, Editor

7:18
7 AM
8 7:18 AM
 Foads and forces
loadbearing structures and statics
A great deal of philosophizing can be done about how design relates
to construction. Very different positions can be taken, but they are ­always
two sides of the same coin. Designing spaces means defining them, by
applying theory to structures that will need to be realized. Knowing about
structures is therefore one of the fundamentals of architectural theory.
It is very rare for the architect him- or herself to vouch for the stability of
constructions. But he or she should be in a position to select structural
elements correctly at the early design stages and to assess the dimen-
sions needed for them realistically. The next step is usually to develop
the loadbearing system with a structural engineer. To be able to work
­together effectively, fundamental knowledge is needed about loadbear-
ing systems and structures, their advantages and disadvantages and the
forces that come into play. These different forces seem complex at first,
but they are logically coherent.

It is easiest to explain how they fit together in the order in which they
are addressed for a statical calculation. A calculation of this type usually
follows these steps:

—— Analysing the overall structure and the function of the individual

structural elements in it – statical system
—— Determining all the forces working on the structural elements –
­assumed loads
—— Calculating the forces affecting a particular structural element
and the forces that it transmits to others – calculating the external
forces
—— Calculating the forces within the structural element itself –
­investigating internal or static forces
—— Determining the stability of the planned structural element
—— Proof that the planned structural element can withstand the forces
determined

forces
Force is defined as mass times acceleration. F = m · a

The unit used for measuring force is the newton; a newton corre- Newton
sponds roughly to the weight of 100 grammes. In building the newton is
complemented by the kilonewton and the meganewton.

7:19
9 AM
 line of application

r ce
f fo
fulcrum force
eo
siz

direction

lever arm

Fig. 1: Force Fig. 2: Torque

Kilonewton:  A force is determined by magnitude and direction. Its action is linear,

1 kN = 1 000 N, and is expressed by its line of application and the direction of this line.
­Meganewton: 
1 MN = 1,000,000 N >  Fig. 1

Moments, torque Forces can also work in a circle around a point. They are then called
torque or moments, and are defined by their size multiplied by the dis-
tance from the fulcrum (lever arm).

A simple example of torque is tightening a screw with a screwdriver.

This also demonstrates the link between force magnitude and lever arm.
The longer the lever arm, the greater the torque. > Fig. 2

Action =  Statics describes the distribution of forces in a system at rest. Build-

reaction ings or parts of buildings are usually motionless, and all the effective
forces balance each other out. This can be summed up in the law “action
= reaction”. It is used as a starting point in statical calculations, on the
basis that the sum of all forces in any one direction and its counter-­
direction is zero. If the action is known, the reaction can be determined
immediately. The chapter External forces, Support forces explains the
methods that apply this possible to loadbearing systems.

Statical System
A structural engineer first establishes the connections within the
construction in the statical system. A statical system is an abstract model
of the real, complex structure of the component parts. Supporting mem-
bers are considered as lines even if they have a wide cross section, and
their load is treated as a point. Walls are presented as disc structures
and their loads are applied in lines. Additional information the statical
system gives is how the structural elements are joined together, and how
their forces are distributed from one element to another. This is crucial

10 7:19 AM
Fig. 3: Load path

to the calculations. The symbols used in statical systems are explained

in the chapter External forces, Support forces > Fig. 8, page 16 and are used
subsequently in the text.

The next working stage involves identifying all the structural elements Positions
in sequence as positions and numbering them. Here it is also important
to establish which structural elements load which others.

For example, roof tiles are not just supported by the roof structure, Load path
but also affect the walls, right down to the foundations. It must be estab-
lished with absolute precision which structural elements absorb the loads
from the upper storeys. > Fig. 3 ◼

external forces
If we consider a building element such as a roof beam, we distinguish
between two types of force. First, there are the forces exerted on it by
the roof structure above it, and those that it transfers to the masonry
supporting it. If we do not consider its dead weight, it does not matter in

◼ Tip: For good cooperation with structural design phase, the main thrust of his or her
engineers, it is important that designers be work is to draw up the statics for planning
familiar with these specialists’ part of the permission and later to draw the plans for
work in a project and understand their constructing the shell. Here the interest is
working methods and aims. It therefore above all in the loadbearing parts of a
makes sense to look at their calculations building. All the non-loadbearing elements,
and positional and working plans and com- even non-loadbearing walls, for example,
pare them with the architect’s documents. are only significant as loads, and may not
After the structural engineer has devised feature in the plans at all.
the structure with the architect in the

7:19 AM
11
 the first place whether this beam is thick or thin, weak or strong, as we
are dealing with external forces that do not include the beam itself.

We must distinguish between external forces and the internal forces

operating in the beam itself. For example, how great is the bending force
in the roof beam exerted by the roof construction it supports? This bend-
ing moment is one of the internal forces that will be explained in the cor-
responding chapter.

Actions
Everything that can affect a structural element is called an action.
Actions are usually forces with different causes. Forces that affect struc-
tural elements mechanically are also called loads.

Loads Loads affect structural elements from the outside, and we must
­ istinguish between them and the reaction forces explained in the sub-
d
section Support forces. Loads are divided into various categories. We
distinguish between point, line and area loads, according to the degree
of abstraction of the statical system. > Fig. 4

In addition, we distinguish between constant, variable and extra­

ordinary actions, in relation to the duration of the action.

Permanent loads Constant action includes, above all, the weight forces of the struc-
tural elements, called permanent loads.

Working loads The working loads include the variable actions wind, snow and ice
loads. Working loads have to be planned in at standard levels for the build-
ing’s intended use. The most important are the vertical working loads
that must be worked out for floors. Whether the rooms are for homes, of-
fices, meeting rooms or some other purpose, they must be given an ap-
propriate working load value as an area load. Largely horizontally applied
loads also have to be taken into account, such as loads on railings and
parapets, braking, acceleration and collision loads for vehicles, dynamic
loads for machines, and earthquake loads. The size of these loads is fixed
in national standards, which give them in tables. > Appendix, Literature

Assumed loads
Vertical load/  After using the statical system to explain how the structure functions,
Horizontal load the next step is to determine the actions. All the acting forces must be
identified, assigned a value and added together. They are generally
­related to a metre or square metre of the structural element. Loads act-
ing obliquely are usually divided into a horizontal and a vertical element.

12 7:19 AM
 load type point load distributed load area load

example

symbol

units kN kN/m kN/m2

examples columns, walls, snow and wind load,

support loads beams panels

Fig. 4: Types of load: point, line and area load

For further calculations we distinguish between vertical loads, hori-

zontal loads and torque. ●

Load absorption area describes the particular reference area for Load absorption area
loads on a structural element. It is part of an overall surface whose load
is being dissipated to a certain structural element. It relates to the na-
ture and span of a structure.

Example: The beams of a timber beam floor are 80 cm apart. Which

part of the floor is acting on an individual beam? The load absorption area
extends from the middle of the space between the beams on the left-
hand side to the middle of the space on the right-hand side, twice 40 cm.
So overall it is again 80 cm wide. > Fig. 5 This is a simple example, but

● Important: Loads acting vertically per square metre

in a structural element: dead weight, working loads for
floors, stairs, balconies
Acting vertically per square floor plan metre: snow load
Acting at right angles to the area of the structural ele-
ment: wind load
Generally acting horizontally: loads on parapets and
railings, braking and acceleration loads, collision loads
from vehicles, earthquake loads

7:19 AM
13
 sp
ac
pressure

in
g
lo
ad

sp
ab

ac
tension
so

in
rp

g
ti
on
ar bending
ea

shearing

Fig. 5: Load absorption area Fig. 6: Forms of force action

determinin­g the load absorption area can be more complicated accord-

ing to the particular structural element.

Force action forms

So far we have considered loads and their magnitude, but how a load,
or more generally a force, acts on a structural element is also important.
Here we distinguish between the following action forms:

—— Compression: one stone lies on top of another, exerting pressure

on it.
—— Tension: tensile load is most clearly explained using the example of
a rope, which can absorb only tensile forces.
—— Bending: a beam is fixed at both ends and then loaded from above.
It sags, i.e. it is subject to a bending load.
—— Shearing: this load is explained by the way a pair of household
scissors loads paper to cut it. Two forces work on each other
slightly offset and transversely to the structural element. This load
often acts on connecting devices such as screws. > Fig. 6

Supports
Points of contact between structural elements at which forces are
transmitted are called supports. A simple example is a ceiling beam sup-
ported on masonry. The beam has its support on the crown of the wall.
In building the idea of the support is somewhat broader, and covers many
different points of contact between structural elements. For example,
when a flagpole is fixed into the ground or a steel beam is connected to
a steel column, this is also called a support. In terms of structural engi-
neering they differ primarily in the forces that they can dissipate.

14 7:19 AM
Fig. 7: Supports in steel construction

It is very easy to look at the different forms of supports in old steel

bridges. Large bridge girders are supported on very small points or nar-
row strips. This means that the girders can deflect without interference
from the supports, which are then known as articulated supports. These
are used on one side of the bridge, while those on the other side are
­additionally supported by steel rollers.

When the bridge girders expand with heat, the supports move on Expansion  
these rollers in order to compensate for the difference in length. Bear- bearings

ings of this kind can absorb the vertical forces affecting the bridge, but
do not resist horizontal forces such as those caused by expansion move-
ment as a result of temperature change, and they do not prevent the gird-
ers from deflecting either. For this reason they are called expansion
­bearings.

These supports are not on rollers and thus transfer horizontal as well Fixed, articulated
as vertical forces. They are known as fixed bearings or simply ­articulations. bearings

What happens to the above-mentioned flagpole fixed into the ground? Restraint
Its anchorage can transfer vertical and horizontal forces from the mast
into the ground, and thus also prevent the mast from Tiping over – a turn-
ing movement around the support. A support of this kind is called a
­restraint. > Fig. 8

We distinguish between three forms of support:

—— Simple supports can dissipate forces from one direction only. They
slide and are articulated.

7:19 AM
15
 simple support double support triple support
sliding, articulated fixed, articulated restraint

Fig. 8: The three types of support, the different ways of representing them in statical
­systems and examples

—— Double supports can absorb forces from several directions. They

are fixed and articulated.
—— Restraints are triple supports and can absorb forces from different
directions, as well as moments.

The correct choice of support is very important in construction, and

must therefore be represented in statical systems. > Chapter Loads and forces,
Statical system

Support forces
Let us assume that a beam is supported on a spiral spring rather than
masonry. The spring is compressed by the load from the beam, thus
­creating a counter-force to the load that the beam exerts.

Support reaction This force is called support reaction. > Fig. 9 If the beam does not move,
the reaction force of the spring is exactly the same as the force exerted by
the beam. Put simply: action equals reaction. > Fig. 10 It is not possible to
see this in the masonry that usually provides support, but it is compressed
just like the spring, so that it can generate the support reaction force.

When calculating a construction it is necessary to know the magni-

tude of the forces that the supports have to apply to support the structural
element above them. The support forces are therefore always ­calculated
immediately after investigating the loads. Applying the above-mentioned

16 7:19 AM
 action
load

beam

spiral spring

reaction
support force

Fig. 9: Support force Fig. 10: Action = reaction

law that action = reaction, it is possible to set up three ­theses for each
structural element that make it possible to calculate the support forces.
These three principles are the fundamental tools for statical ­calculations.

They are also known as the three conditions for equilibrium: > Fig. 11 Conditions for
equilibrium

∑V = 0

All vertical loads together are the same as all the vertical support
­reactions. This means: the sum of all vertical forces equals zero.

∑H = 0

All horizontal loads together are the same as all the horizontal sup-
port reactions. This means: the sum of all horizontal forces equals zero.

∑MP = 0

If a support is considered at a support point P, all the forces turning

clockwise around this point are the same as all the forces turning anticlock-
wise. This means: the sum of all moments around the given point equals
zero. Here it should be noted that any force or load can be seen as a tor-
sional force around a fixed point, so by definition the force times the lever
arm length gives the size of the torsional force. > Chapter Loads and forces, Forces

7:19 AM
17
 load

load P

BH-horizontal lever arm

Bv
load
Av-vertical Bv
lever arm support

V= 0 H= 0 MP = 0

Fig. 11: Conditions for equilibrium

It is only by working out the sum of moments around a support that

it becomes possible to calculate two different support forces. Because
the torque centre is in a support, the support force does not have a lever
there and equals zero in this equation. This means there is only one
­unknown in the calculation, and that is the other support force, which
can be calculated easily.

So for the beams shown in Figure 11 with a single central load the
sum of the torque around point P is as follows:

∑MA = 0 = A v • 0 + F • l/2 – B v• l → Bv = F l → Bv = F/2

 •

Both supports dissipate half the central single load. This conclusion
could have been reached without calculation in this case.

A rule of signs has to be decided for all calculations using the condi-
tions for equilibrium. Rules of signs are not defined and so the statement
must always be represented with an arrow. It shows the direction of the
forces that are treated as positive. In this case turning to the right was
treated as positive, so left-hand forces must be stated with a negative
rule of signs.

Internal forces
So far we have discussed only the forces impacting on a structural
element and the forces the support generates as a reaction to them.
These are called external forces, because the structural element itself
has not yet been considered. But what is happening in the beam itself, or
put another way, what forces are effective in the member?

18 7:19 AM
 V

M
N

N normal force
V shear force
M bending moment

Fig. 12: Internal forces

To understand this, imagine a beam on two supports is cut through

at a random point. What happens? It collapses and cannot support any-
thing, not even itself. The crucial question now is what forces have to be
effective at this cut face to prevent the beam from falling, or what forces
are needed in order to achieve an internal equilibrium of forces.

Here the above-mentioned conditions for equilibrium are useful, as

they apply equally to internal and external forces. It is assumed that the
external forces acting from the plane of the cut to the end of the beam
have to be as great as the internal forces that counteract the external
ones at the plane of the cut. > Fig. 12

Just as the external forces are identified as vertical forces, horizon- Internal focus
tal forces and torque, internal forces are identified as normal forces, shear
forces and bending moments, and their direction always relates to the
structural element itself.

Normal force
A normal force is a force working longitudinally or in the direction of
a structural element. As the first example illustrating a normal force we
will take a rope hanging on a hook, with a weight attached to the rope.
>  Fig. 13 The weight is the load and the hook provides the support reaction.

These would be the external forces.

Leaving out the weight of the rope itself, the same tensile force is Tensile force
e­ ffective at every point in the rope. Here it does not matter whether the
rope is short or long. Therefore the same normal force is effective at every
point in the rope, and its magnitude is the weight of the weight attached.

7:19 AM
19
 N

Fig. 13: Normal force with a rope as example Fig. 14: Normal force with a masonry pier as example

Two longitudinal directions of forces were explained in the section

Force action forms: pressure and tension. This is a tensile force.

Compression force We take a free-standing masonry pier for our next example. > Fig. 14
The pier’s dead weight is the only load identified: masonry is a heavy
­material. It is easy to calculate the support reaction of the foundation at
the base of the pier, as this must be the same as the weight of the pier.
But what happens in the pier itself? The topmost stone takes no load from
any other, so there is no normal force at this point. The second stone from
the top takes the load of the one above it. So at the point of the second
stone from the top there is a small normal force in the form of a compres-
sion force. This compression force becomes greater stone course by
stone course to the bottom of the pier. That is to say, the normal force
increases from the top to the bottom of the pier. > Chapter Loads and forces, Forces

The magnitude of the normal force can be demonstrated in diagrams,

similarly to the representation of loads. The two examples show differ-
ent flows of normal forces. In diagrams of this kind, tensile forces are
noted with positive signs and compression forces with negative signs.
>  Figs. 13 and 14

Sheer force
For external forces, a distinction is made between horizontal and ver-
tical forces. Internal forces have the same relationship, but their direc-
tion relates to the system axis of the member in each case. Just as the
longitudinally effective tensile and compression forces are defined as
normal forces, all the forces working transversely to them are known as
shear forces. They are not as easy to understand as normal forces, and
must not be confused with bending, which is explained in the next ­chapter.

20 7:19 AM
 Av Bv

V
V

Fig. 15: Flow of shear forces in a cantilever arm under Fig. 16: Flow of shear forces in a simply supported
uniform distributed load beam under uniform distributed load

The effect of shear force will be explained taking a cantilever arm as Cantilever arm
example. Figure 15 shows a beam that is fixed into a wall at one end. This
kind of beam is known as a cantilever arm. It could be part of a balcony,
for example, and is loaded by its dead weight as a uniform distributed
load. If this beam were cut through close to its end, the section cut off
would fall because of this distributed load. The load works transversely
to the axis of the member and thus produces the shear force. If a longer
piece is cut off, more of the uniform distributed load has to be absorbed
as a force transverse to the member axis at that point. Thus the shear
force is greater at this point than at the previous one. The force would
­increase with every additional cut. The shear force thus increases from
its free end towards the fixing point. So the support force at the point of
fixing must be able to react to this shear force equally.

Figure 16 considers a beam on two supports, called a simply sup- Simply supported
ported beam, with a uniform distributed load. The simplest thing to beam

­understand is the flow of shear forces when imagining one section after
another cut off from left to right and considering what external forces
are at work to the left of the cutting face.

The first interesting cutting face is just to the right of the support on
the left. What happens in this section? The support force from the sup-
port is exerted upwards transversely to the member axis. Thus the shear
force corresponds to the support force. But if a further cut is made to the
right, part of the line force works in the other direction. This reduces the
shear force in relation to the previous result.

Now a cut is made precisely in the middle. What forces are working
transversely to the member from its left-hand end to the cutting plane?
They are, first, the support force towards the top, and then the distributed

7:19 AM
21
 load of the member section from the left-hand end to the centre. So half
the distributed load of the entire member is effective. In a symmetrical
system like this one it is easy to establish that each support dissipates
half the distributed load. In this case the shear force in the middle of the
beam equals zero.

If we now consider another cutting face to the right of this, an even

greater proportion of the line force is effective. This means that the shear
force becomes negative. At the cutting face just in front of the left-
hand support almost the whole distributed load is working against the
­unchanged support force of the left-hand support. It is only the support
force of the right-hand support that makes the result add up to zero again.

If the right-hand section of the beam had been considered instead

of the left, the result would have been the same. So it does not matter
which subsystem is considered, as the internal forces have to be in equi-
librium at every point in the beam. This applies to all internal forces.

Bending moment
The effect of moments has already been discussed in the chapter
­External forces. Here all the effective forces were seen as turning around
a fixed point. Their magnitude is defined as force times lever arm. > Chap-
ter Loads and forces, Forces and Chapter External forces While the support forces were
interesting for the external forces, the forces working in the run of the
beams are important for determining the internal moments.

Bending The internal moments cause the beam to bend. Bending is the key
load for which many structural elements have to be dimensioned. When
making statical calculations it is therefore necessary to know how great
the bending moments need to be at any given point in the beam. This is
shown in the moment gradient, which is thus an important aid for con-
structing members under bending load.

The direct link between internal moment and bending will again be
explained below using a cantilever arm. How does the cantilever arm
­deform under a uniform distributed load? The load causes the beam to
bend downwards. >Fig. 17 Here, deformation through bending means that
the beam has to become longer at the top and short at the bottom. This
creates a tensile force at the stretched top side and a compression force
at the squashed bottom side. These tensions counteract the load as an
internal force.

Thus the bending itself creates the internal moments whose magni-
tude depends on the magnitude of the external forces and the length of
their lever arm. For the cantilever beam there is low distributed load with

22 7:19 AM
 tension single load

pressure

M M

Fig. 17: Bending moment of a cantilever arm under Fig. 18: Bending moment of a simply supported beam
equal line force under single load

a low lever arm force at the free end, and the moment is consequently
small. But at the fixing point the full distributed load is effective, with a
great lever arm force, so the moment is large. > Fig. 17

A simply supported beam under a single load will deflect downwards.

Thus, in contrast to the cantilever arm described above, it will be stretched
at the bottom and squashed at the top.

Here the bending is in a different direction from the previous exam-

ple. How does this affect the flow of forces? Let us examine the beam
from left to right.

The support force is effective at the right side of the left support, but
it has no lever arm force, so the bending moment is zero. As the lever
arm force increases with distance from the support, the moment increase
is linear. This is the case up to the point where the single load is exerted.
To the right of this, the single load works against the support force with
increasing lever arm force, and the bending force diminishes until it be-
comes zero again at the second support. This test can be carried out as
from the left-hand or the right-hand side as desired; the result is the same
in each case. > Fig. 18

How does the flow of forces change if a uniform line force q is being
exerted, rather than an individual load? A distributed load can be summed
up as a resulting single load whose line of action lies at the centre of grav-
ity of the distributed load. The magnitude of this resulting individual load
is force per unit of length times its effective length.

To calculate the bending moments, these resultant individual loads R = q · l [kN· m] 

and their lever arm lengths have to be established at the various cutting m

7:19 AM
23
 A

l lever arm
l a l distributed load
B M

l lever arm
lb l distributed load parabolic path

Fig. 19: Sectional results for a distributed load Fig. 20: Bending moment for a simply supported beam
under equal distributed load

planes. > Fig. 19 These act against the support force with increasing mag-
nitude. The resultant moment gradient is a parabola, because the length
as that of the distributed load and that of the lever arm goes into the
­calculation twice.
q • l2
Moment of a distributed load: MA = q • l • l/2 → MA =
2

The supports are important points for the moment gradient, and the
bending force is again zero at both. How can this be explained? If we
make a cut at the support and look in the direction of the support > Fig. 19
no force has a lever arm, because the member has no measurable length
here; we are actually looking at a point. In general it can be said that
bending requires a fixed beam cross section that can resist moments.
But this does not apply if a cut is made in an articulated joint, which a
hinged support is. A chain, for example, is an accumulation of articulated
joints, and therefore cannot absorb bending. Hence we have an impor-
tant principle: the bending moment in an articulated joint is zero. > Fig. 20

Maximum moment In this example, the greatest moment is in the centre of the span. To
bear loads, the beam must be able to resist this greatest moment. It is
true in general that when dimensioning a structural element, the bend-
ing, the location and value of the maximum moment must be established.

When planning elaborate beams over large spans, the maximum

­ oment is not the only important factor. It can be economical in terms
m
of material to adapt the cross section of the beam to the moment gradi-
ent, in other words to shape the beam so that it is dimensioned precisely
for the bending moment effective at every point. For this reason an

24 7:19 AM
a­ rchitect should be able to determine the moment gradient in a beam
qualitatively appropriately to the load.

Correspondence between the internal forces

The three different internal forces are introduced in the above para- ◯
graphs. When calculating loadbearing structures it is generally neces-
sary to establish all three internal forces, so that a structural element
can be dimensioned to deal with all three working together.

Shear force and moment correspond closely above and beyond this.
The two forces, which result from the same load, permit inferences to be
drawn from each other. For example, if no force is acting around a rod,
the value of the shear force cannot change either, i.e. it is constant. But
because moments are defined as force times lever arm, their changes of
magnitude in an unloaded area are linear. If a force is acting at the place
in question, the values of the resultant moment change in proportion to
its distance from that place. Such interrelationships between shear force
path and moment gradient are inevitable. > Fig. 21

The following correspondence in particular is important for statical

calculations: if we compare the force flows, it becomes clear that the
shear force has a zero crossover at the locations of the maximum mo-
ments. This turns out to be useful, because the location of the maximum
moment can be ascertained from the flow of shear forces, and then has
to be calculated at this point only. > Figs. 21 and 22

And with a little experience, it is also possible to determine the an-

ticipated shear force flow and moment gradient qualitatively. Figure 23
shows the appropriate force flows for some common load types.

1. If no force is effective in the member path, the flow of shear forces

is constant and the moment gradient linear.
2. A single line creates a break in the flow of shear forces and a kink
in the moment gradient.
3. For uniform distributed loads the flow of shear forces is a sloping
straight line and the moment gradient a parabola.

◯ Note: Signs for representing internal Bending moment: The moments are drawn
forces are fixed as follows: in the direction of their deflection, the posi-
Normal force: Pressure (–) is represented tive moments downwards and the negative
upwards, pressure (+) downwards. ones upwards.
Shear force: The positive shear force is But these conventions should not be seen
drawn above, the negative below the sys- as definitive. For example, some countries
tem line. represent the bending moments the other
way round.

7:19 AM
25
 equally distributed load
single load

V V

M M

Fig. 21: Shear force and moment of a simply supported Fig. 22: Shear force and moment of a simply suppor­ted
beam subject to single load beam subject to a distributed load

1. 2. 3. 4.

loading no load P single load distributed load g g1 g2

V=0
V

shear force

M max. M max. M

moment

Fig. 23: Correspondence between load, shear force and moment

4. A break in the uniform distributed loads creates a kink in the flow

of horizontal forces; in the moment gradient two parabolas with
different inclines fit together with the result that they have the
same tangents. Columns 1 and 2 in the table could be details from
a system as in Figure 21, while Figure 22 can serve as an example
for column 3.

26 7:19 AM
 s load type 1
g snow on cantilever arm

section s load type 2

g snow on roof alone

moment gradients
LF 1
LF 2 s load type 3
LF 3 g snow overall

s snow load
envelope curve g dead weight

Fig. 24: Loading types and envelope curve

Loading types
In practice, many different loads overlap. They have to be added
­together in the calculation in order to dimension the structural elements
for the maximum loading. But there are also cases where danger does not
derive from the maximum load. Maximum values for internal forces, which
are crucial to the dimensioning, are also possible for other load combina-
tions. These different combinations of loads are called loading types.

Take the following example: a small workshop has a flat timber beam
roof with a canopy protruding extensively on one side for the storage of
material. A great deal of snow then falls in the winter. As the workshop
is well heated and the roof poorly insulated, the snow on the roof melts,
but not on the canopy, which has unheated space beneath it. The snow
remains on the canopy alone. This load presents the risk of the workshop
roof rising, and the roof beams above the wall next to the protruding sec-
tion breaking. > Fig. 24

To avert these dangers, the structural engineer must calculate not

only the snow load for the building as a whole, but the snow load on the
canopy as well, as they each introduce different risks. The first step must
be to establish what combinations or types of load are possible, and put
them together. If the moment gradients of the different load types are
then mapped onto each other, the possible extreme value for each point
can then be read off from this diagram.

This figure is also called an envelope curve. Its extremity shows the Envelope curve
load type that is crucial for each point. Figure 24 shows that the great-
est positive moment in the span occurs in load case 2, but the greatest
negative moment in load types 1 and 3.

7:19 AM
27
 dimensioning
A statical calculation runs similarly to the sequence in the above
chapters. After the statical system has been established, the assumed
loads are calculated and then the external forces, and after that the
­internal forces determined for the component parts of the building.

It would be wonderful if we could now simply work out the required

cross section for the element. Unfortunately this is not as simple as it
sounds, as all the parts of the building become part of the calculation
themselves, as loads; in other words, all the structural elements have to
be known in order to work out the assumed loads, so that their weight
can be included. If the calculation then reveals that one of the structural
elements assessed will not bear sufficient load, we have to start again
from the beginning.

Even if this does not mean that all the work is invalidated, careful
planning is clearly advantageous at this point: dimensions should be
­estimated in advance. This can be done with the aid of rough formulae.
>  Appendix, Pre-dimensioning formulae

Strength
After the forces coming into play have been determined, the load-
bearing capacity of the structural elements is now of interest. This
­depends mainly on two aspects, the material and the cross section.

One of the first steps in designing a construction is to decide on the

materials. Every building material has its advantages and disadvantages.
The strength or resistance offered by different materials is particularly
important for construction. Cable constructions, for example, resist ten-
sion but not compression, while masonry resists compression but not
tension. Timber, steel and reinforced concrete constructions are com-
pression- and tension-resistant, and also resist internal forces. > Chapter
Loads and forces, External Forces

It has already been established that compression and tension are

both produced by bending. Consequently, only materials that resist both
loads can be used when bending loads come into play (e.g. timber and
steel members).

Tension Materials also differ in their capacity to absorb forces. This capacity
is expressed as the amount of force a material can absorb over a given
F [kN] 
s =     area. The strength per area is expressed as tension s.
A [m2]

Robert Hooke, To understand the concept of resistance we must cite Hooke’s Law,
­1635–1703 which states that tensions and extensions are proportional in the elastic

28 7:19 AM
field. What does this mean for building materials? Every material, whether
it is wood, steel, reinforced concrete or masonry, is essentially elastic. If
a structural element is loaded, tensions are created, causing the mate-
rial to extend proportionately. So if a beam is loaded, it deflects, or sags.
If the load is doubled, it deflects to twice the previous extent. If the load
is reduced again the deflection also decreases.

This simple pattern applies only up to a certain point. If the tension

becomes too great, the material no longer responds elastically, but plas-
tically, i.e. permanent deformations occur. This is the point at which the
structural element starts to be damaged. If it has to take a further load,
it fails completely, although the failure differs in kind from material to
material. The value indicating how much tension a material can absorb
before it deforms plastically and fails is a purely material characteristic
and has nothing to do with the geometry of the structural element. It is
important for construction that the maximum admissible value is not
reached at the highest possible load. The tensions a particular material
can absorb are established under laboratory conditions, taking variations
in material quality into consideration.

The value established in this way is known as admissible tension and Admissible tension
can be ascertained from tables. > Appendix, Literature

In addition, every material is available in different qualities with a va- Strength classes
riety of admissible tensions, and is assigned to a “strength class”. For ex-
ample, normal and high-strength concretes are distinguished by their
strength class. The actual verification of a structural element’s loadbear-
ing capacity always works on the principle that the actual tensions must
be less than the admissible tensions. The admissible tensions can be es-
tablished from tables, but the main part of the work is in working out the
actual tensions. If structural elements are loaded normally, working these
tensions out is simple. The existing tension corresponds to the normal
force per sectional area of the structural element. If the result shows that
the existing tension is lower than the admissible tension, the structural
element is correctly dimensioned. Unfortunately this simple verification
is only rarely the deciding factor for dimensioning. Cables, which in fact
can absorb only tensile forces, are dimensioned in this way, but in most
cases the bending load is the key factor for dimensioning.

Moment resistance
Every proof of suitability for a structural element is based on the
­actual tensions being lower than the admissible tensions. This also
­applies to members subject to bending loads.

7:19 AM
29
 1 /2 h
zero line

unloaded compression

h
tension

1 /2 h
loaded

2 /3 h
compression
M

tension

Fig. 25: Deformation by bending Fig. 26: Tension distribution in the member subject to

bending

Stress distribution When explaining the bending moment we stated that a bending mem-
ber is subject to tensile stresses on one side and compression stresses
on the other, but how great are these stresses, and how exactly are they
distributed?

To find this out, consider an unloaded member marked transversely

with straight lines. If it deflects when a load is applied, the marks incline
towards each other in a trapezoid shape, but the lines remain straight.
>  Fig. 25 If the extensions and tensions are proportional, this results in a

stress distribution that is also in straight lines from the tensile stress at
the lower edge via the middle level, which is tension-free, to the compres-
sion stress at the upper edge.

Neutral tension plane As can be seen in Figure 26, the compression and tension distribu-
tions each form a triangle. These triangular tensions can each be summed
up in a resulting tension at the centre of gravity of the triangle, with a dis-
tance from each other of 2/3 of the height of the cross section. This
length represents the lever arm of the internal moments that counteract
the loads and are thus responsible for loadbearing capacity. So the
greater the height of the member, the longer the lever arm of the inter-
nal tensions, and the greater the stability of the member.

Thus, the length of this lever arm is the key to resistance to bending,
but the section width is also important. This section resistance is ex-
pressed as moment resistance. Moment resistance is a value relating to
the geometry of a member and not to its material.

For example, the rectangular cross section customary in timber con-

struction produces the moment resistance W with a magnitude of W =
w · h2 / 6.

30 7:19 AM
 It is worth looking more closely at this formula: the height h is
squared, while the width w is simply entered as a factor. An upright rec-
tangular section has a higher loadbearing capacity than a square one, or
a horizontal rectangle. Expressed more precisely, doubling the width of
a profile doubles the loadbearing capacity, but doubling its height multi-
plies it by four.

The moment resistance for dimensioning a simply rectangular cross

section can be established by using the formula above. For other cross
sections, for all steel profiles, for example, the dimensioning is more com-
plicated. For this reason, moment resistance values are always given in
sets of tables. > Appendix, Literature

The term moment resistance contains the word moment. In contrast

with the concept of moment or torque explained in the chapter Forces,
moment resistance refers not to an individual force with a particular ­lever
arm, but to area elements and their lever arm around the tension zero
line, as we are considering a cross section area. > Fig. 26 Like moment of
inertia, explained in the following chapter, moment resistance is also
­defined as an area moment.

Moment of inertia
Moment of inertia is best explained by its effect. Moment resistance
expresses a member profile’s resistance to bending moments, while the
moment of inertia relates to its deflection. It describes the rigidity of a
cross section.

Like moment resistance, moment of inertia is based on the distribu-

tion of tension in a cross section subject to bending. Here the severely com-
pressed and extended areas at the edge are more effective than those in
the zero tension line area. But the distance of the area elements from the
zero line is more important for moment of inertia than moment resistance.

The moment of inertia is the sum of all area elements in the cross
section multiplied by the square of their distance from the zero line.

The moment of inertia for rectangular cross sections is calculated

using the formula I = w · h3 / 12. So here the cross section height is raised
to the power of three, which means that deflection is reduced by one
eighth if the member height is doubled and the width remains the same.

Moment of inertia can be used to calculate the expected deflection Deflection

of a member. Even though the priority is to dimension structural elements
for loadbearing capacity through moment resistance, evidence is addition-
ally needed that an admissible maximum deflection is not being exceeded.

7:19 AM
31
 compression force
shear force shear force
tensile force

deformation of the individual planks

pinning the planks M

Fig. 27: Example: effect of shear stresses Fig. 28: Stress fields in a member subject to bending

Shear stress
Let us take the following example to explain shear stress: two planks
are laid one on top of the other as a simply supported beam, and a load
is then exerted on them. Both planks will deflect under the load and shift
in relation to each other. > Figs. 27 and 28 They should be fastened together
to increase their loadbearing capacity, as a tall cross section has a higher
loadbearing capacity than two planks on top of one another with the same
height. > Chapter Loads and forces, Dimensioning What is the best thing to do?

One possibility could be to drill through the unloaded planks and fas-
ten them together with bolts and dowel pins.

But now we have to ask what stresses these dowel pins are actually
expected to absorb and how they come into play. The answer to the first
part of the question is simple. Shear stresses are responsible for the
planks’ shift in position. The simplest way to explain where these stresses
come from is by means of the members shown in Figure 28.

In this member subject to a uniformly distributed load the greatest

tensile stress is in the middle of the span on its lower side and the great-
est compression stress on its upper side. These stresses reduce as the
bending moment decreases towards the supports, but what happens to
the stresses that cannot just simply disappear? The compression and
tensile stresses cancel each other out towards the supports and this hap-
pens as a result of the shear stresses that increase as the bending
stresses decrease.

32 7:19 AM
 stem
flange

Fig. 29: Spaces for service ducts in a member subject to bending stress

The bending stresses can be discerned in the moment gradient, while

the shear stresses are proportional to the shear forces and in the case
of a member subject to a simple and uniform load the shear forces in-
crease towards the supports. > Chapter Loads and forces, Internal forces

It is essentially true of members of this kind that the tensions result-

ing from bending in the centre of the span are greatest at the top and
bottom of the section > Chapter Loads and forces, Internal forces and the maximum
shear stresses are at the supports. Timber, for example, is a material that
is sensitive to shear stresses. In timber structures it is often necessary
to reinforce members to absorb shear stresses at the supports.

Another example is offered here to clarify shear stresses further. The I-beam section
usual steel sections, such as I-beam sections, are designed so that the
flanges can absorb the bending compression and bending tension, while
the stem absorbs the shear stresses.

For example, if an architect wants holes for wiring or pipework to be

cut in a simply supported beam, this does not present a problem in the
middle of the beam, as the forces in play tend to be small there, while
the two flanges are fully loaded with bending compression and bending
tension. But holes should not be drilled near the supports, because here
the stem is heavily loaded with shear stresses. > Fig. 29 ◼

◼ Tip: The cabling and ducts for power, water, sewage

disposal and, above all, ventilation can have a crucial
effect on the design of a loadbearing structure. They
should be fixed at an early stage and agreed with the
structural engineer. Essentially, loadbearing structure
and service pipes and ducts should be planned so that
as few crossing points are created as possible.

7:19 AM
33
34 7:19 AM
 Structural elements
Cantilever arm, simply supported beam, simply supported beam
with cantilever arm
Loads and forces were explained in the first chapter using the can-
tilever arm and the simply supported beam as examples. These two load-
bearing systems form the basis for most of the more highly developed
and more complex systems. It is worth recapitulating their advantages
and disadvantages.

A cantilever arm can be compared with a long lever used to lift heavy Cantilever arm
loads. Consequently the leverage acting at the anchor point is the big-
gest problem. As can be seen in Figure 30, this is the point of maximum
torque and maximum shear force, and the anchor point has to absorb
both. This is scarcely feasible in timber construction, as no nailed or
screwed joint could do the job unless the anchor point were long enough.
But an anchor point in masonry is easily possible, although the danger
remains that the long lever could lift masonry even if there is not enough
of it above the anchor point. If we look at the moment and shear force
gradients, it is clear that a cantilever arm subject to a uniformly distrib-
uted load has to be dimensioned for the area around its anchor point, but
is thus overdimensioned for the rest of its length. It therefore makes
sense, saves material and is customary for a cantilever member to have
its height reduced from the anchor point to the free end to correspond
with the internal force gradients.

The simply supported beam is probably the most common loadbear- Simply  
ing system, and it is worth looking at carefully again here. A simple ­timber, supported beam

statical
lever action
system

M
example

Fig. 30: Cantilever arm: statical system, internal force gradients and example

7:20 AM
35
 statical
system statical
system
V
bending line

M V

example
concrete
prefabricated
M
element

Fig. 31: Simply supported beam Fig. 32: Simply supported beam with cantilever arm

steel or even concrete section with a consistent cross section is normally

used, as these are easy to produce and cheap, and also have the advan-
tage of offering a flat surface at the top and the bottom. But actually most
simply supported beams are only fully exploited at one point, in the mid-
dle, the point of maximum moment. So it can make sense to adapt the
beam to the moment gradient and make it higher in the middle than at
the supports. > Fig. 31 and Chapter Loads and forces, Internal forces

Wood is a natural product that can absorb considerably fewer forces

transversely than it can longitudinally to its grain, and it is thus sensitive
to shear forces. In the best cases a timber beam is thus fully exploited at
three points, in the middle because of the bending moment and at both
◼ ends because of the shear forces. > Chapter Loads and forces, Dimensioning

Simply supported The simply supported beam with cantilever arm is a very useful sys-
beam with cantilever tem from the point of view of loadbearing theory. It could be said that as
arm
a combination of the two previous ones it compensates for the disadvan-
tages in each case. The problem with the cantilever arm is its anchor
point. But in this system the length of the anchor point, i.e. the span
length, is usually greater than the protruding section itself, and thus
­unproblematic. The key factor for this beam is what happens above the
support in the cantilever arm. Here, the cantilever section has its maxi-
mum moment, with a negative value. > Fig. 32

A simply supported beam has its maximum positive moment in the

middle of the span while at the support it is approaching zero. How do
these two lines fit together? This becomes clear if we imagine a member
of this kind deflecting when loaded. The cantilever arm would hang down

36 7:20 AM
and the member would also sag in the span. It hangs in at the articulated
end support, but curves over the other one and lies horizontally over the
support.

This means that the inflection point of the “bending line” shifts from Bending line
the support into the span. > Fig. 32

This is also reflected by the moment gradient. Corresponding with

the protruding section, the negative maximum lies above the support.

A negative above a support is called a moment at support. It is not Moment at support

relieved in the span before a “midspan moment” comes into being. This
is the term for the positive moment in the area between the supports. Be- Midspan moment
cause of the moment at support it is smaller than in the case of a pure
simply supported member. The member in the span is thus relieved by
the cantilever arm. This means that a member of this kind can be smaller
than a simply supported beam over the same span width. ◯

Continuous beam
Continuous beams extend over several spans. They are defined pre-
cisely according to the number of these spans. A two-span member has
three supports, a three-span member four, and so on. Such systems are
a logical extension of the situation explained above. As in the simply sup-
ported beam with cantilever arm, a moment at support is created above
a central support, and this reduces the bending moments in the spans.
Here, the inflection points in the bending line correspond with the zero
points in the moment line, although the bending line and the moment line
do not have the same form. But the form of the bending line indicates the
moment gradient. > Fig. 33 and Chapter Loads and forces, Internal forces

So the advantage of continuous members is that they reduce the

span moments through the moments at support over the supports. Lower
span moments means that the members can be smaller.

◼ Tip: Squared timber sections used as beams should ◯ Note: A positive moment or a midspan moment
be neither too slender nor too wide. Sizes with side ­signifies tensile stress on the bottom side and com-
ratios between 2/3 and 1/3 make sense. pression on the top side of the member section. A
Building tables give timber sections. The section sizes ­negative moment or a moment at support creates com-
quoted in these tables are usually available from stock pression on the top side and pressure on the bottom
in the timber trade, so they do not need to be specially side (see Chapter Loads and forces, Internal forces).
cut to size, which would make additional work for the
carpenter.

7:20 AM
37
 three-span member two-span member

bending line
V

load source area

M M

Fig. 33: Three-span member Fig. 34: Load source area for a centre support

Effect of continuity The effect of continuity thus makes significant material savings pos-
sible.

Without closer consideration it might well seem that a central sup-

port has to bear exactly twice as much load as the peripheral supports.
But this is not the case. The threshold of the load source area is not in
the middle of the span, but at the point where the shear force is zero and
the span moment at its maximum. Thus, a central support takes more
than twice the load at a peripheral support. > Fig. 34

Articulated beam
Another possibility emerges when looking at the moment gradient
of the continuous member: the individual beams can be fitted together
at the points of zero moment. This sustains the effect of continuity and,
Point of zero moment above all, a point of zero moment means that there is no bending at this
point. So if a beam joint is planned for this point, the moment gradient
does not change when compared to the continuous beam. In timber con-
struction a beam joint is an articulation, as is the case with almost every
nodal point, and the bending moment at an articulated point is inevita-
bly zero. > Figs. 35 and 36

Statical determination An added articulated joint affects the system in a further way. Con-
tinuous beams and articulated beams differ in one essential quality. What
happens to a continuous beam if one of the supports is lowered for some
reason? The beam will have to bend in order to remain supported by all
of them. This creates stresses in the structural member. If this were to
happen to an articulated beam, there would be no stresses in the sec-
tion, because of the articulated, sliding nature of the support system.

38 7:20 AM
 example from timber construction:
two-span member purlin joint
lapping with bolt

example from steel construction:

girder joint
butt plate welded on

Fig. 35: Articulated beam Fig. 36: Examples of articulated connections for

s­ tructural elements

Loadbearing structures in which stresses would be created if a sup- Statically

port were to be moved are called statically undetermined, and if this does undetermined

not happen, they are called statically determined. > Fig. 37

For example, cantilever arms and simply supported beams, to which Statically determined
this distinction applies as well, also prove to be statically determined sys-
tems. More loadbearing systems will be explained in the following chap-
ters that can be statically determined or statically undetermined. Stati-
cal determination always depends on the number and nature of the
supports, and the number of articulation points. Adding articulations can
turn a statically undetermined system into a statically determined one.
But care is needed here, as a superfluous addition would render the sys-
tem unstable.

Closer consideration of a three-span member with a uniformly dis- Three-span member

tributed load shows that it would need two articulated joints in order to
raise or lower each joint without causing stresses. In other words, two
articulated joints are needed to make it into a statically determined sys-
tem. Four points of zero moment show in the moment gradient. There
are thus several possibilities for arranging the joints within these points.
>  Fig. 38

What is the difference between statically determined and undeter-

mined systems in practice? Statically undetermined systems offer a
somewhat higher degree of safety based on the distinctions identified
above. If, for example, one support for a continuous member should fail,
there is a chance that the structural element will not collapse because
the member is still supported by the remaining ones. In a statically

7:20 AM
39
 three-span member, statically undetermined

statically undetermined member M

stresses when support lowered

statically determined, two variants

statically determined member unstable system, one articulated joint too many
no stresses when support lowered

Fig. 37: Statical determination Fig. 38: Arrangement of articulation points

­determined system, such as a simply supported beam, this would not be

the case. In addition, statically undetermined systems cannot be calcu-
lated using the three conditions for equilibrium. More elaborate calcula-
tion methods are needed here.

Trussed beam
The span width is the most important criterion for choosing a par-
ticular loadbearing system. In any construction, it is possible to assign a
sensible point for a width at which the span can still function, but will
­become inefficient if that point is exceeded. For example, this point is
reached at approx. 5–6 m for the efficient use of single timber beams.
Further measures are needed for wider spans: for example, if it is not
possible to place a support underneath, a brace can be inserted instead,
> Fig. 39 which will dissipate its load to the supports via a truss. The truss

pushes the brace upwards, like a drawn bow, and thus works like a sup-
port, even though it does not touch the ground. This system is called a
trussed member or beam.

It is also possible to truss a beam doubly or triply. > Fig. 40 This in-
creases the span even further, though the forces in the structural mem-
bers increase correspondingly. How are the individual parts of the trussed
member loaded? The brace is compression loaded, as it is supporting the
beam. The truss, which is usually made of steel rods, is subject to a ten-
sile force and the beam, which was originally subject only to a bending

40 7:20 AM
 single truss with sag rods
brace

double truss with sag rods

tie member

M triple truss with sag rods

Polonceau truss

Fig. 39: Trussed member Fig. 40: Multiple trussed members

load additionally acquires compression as a counterforce to this tensile

force.

More complicated systems can be constructed with the aid of braces Jean B.-C. Polonceau,
and trusses. One example is the Polonceau truss, named after its inven- 1813–1859

tor. > Fig. 40

It is possible to sum up what was achieved by the trussed member

as follows: the simple beam becomes a complex system that deals with
the loads not just by absorbing bending moments, but by dissipating them
as compression and tension in various structural elements set at a great
distance from each other. The upper member is no longer subject mainly
to bending forces, but also to compression, and the truss dissipates the
tension forces. When explaining the bending moment, we talked about
the lever arm of the compression-loaded cross section parts as opposed
to the compression-loaded ones. This lever arm is clearly enlarged here.
Hence, such systems are considerably more efficient. > Chapter Loads and
forces, Dimensioning

lattice
A trussed girder with more than three struts makes less sense for a
number of reasons. But if the struts are supported individually in every
section, this produces a new system that can cope with considerably
larger spans. It is known as a trussed, lattice or skeleton girder. In these

7:20 AM
41
Fig. 41: Steel lattice constructions

lattice girders the tension-loaded sections are usually made up not of ca-
bles or bars, but of timber or steel sections. Lattices are efficient sys-
tems that are very common, and can be adapted to fit the requirements
of a particular situation. They can be realized in almost any material, and
the bars can be arranged in very varied ways. > Fig. 42

Tension diagonal In the examples discussed so far, the diagonals have been realized
as tension-loaded bars, corresponding to the truss below them. But it is
equally possible to install the bars exactly the opposite way round.

Compression diagonal They are then compression loaded. To identify the direction of forces
in the diagonals, it helps to ask whether they are loaded with compres-
sion forces, like an arch, or with tension forces, like a sagging cable.

Alternate diagonals It is also possible to construct lattice trusses with diagonal members
alone, in alternate directions. The look of the truss does not change in

lattice with tension diagonals

lattice with compression diagonals

lattice with alternate diagonals

Fig. 42: Trussed girder – lattice girder Fig. 43: Diagonal members in a lattice girder

42 7:20 AM
 purlins

Fig. 44: Supporting the top boom, which is subject to buckling risk, with purlins

the middle. The bars in one direction are compression loaded and the
others tension loaded, and in the middle of the lattice structure the sec-
tions change their load, while their arrangement remains the same.

There are individual bars in lattice girders that on closer considera- Unstrained members
tion are not directly involved in dissipating the load. They are neither
­tension nor compression diagonals, and are thus called unstrained mem-
bers. Nevertheless, they cannot usually be omitted, because they are
needed for structural reasons. This can mean that they complete the
­outlines of the system or hold it in position. In the figures, compression
members are drawn as thick lines, tension members as thin lines and
­unstrained members as dashed lines. > Fig. 43

The height and length of a lattice girder is calculated according to Top boom
the span. But its width depends only on the girder section selected in
each case, and is usually very narrow in comparison with the overall
length. For this reason the compression loaded upper section, called the
top boom, is at risk of buckling. > Fig. 44

This problem can be solved in a variety of ways.

The top boom can be fixed to a ceiling above it or to longitudinal ceil- Purlins
ing beams, and thus prevented from moving out of place; or it can be con-
structed as a buckleproof girder in its own right.

If a second boom is added to the top boom, which is in danger of Three boom truss
buckling, and a kind of lattice is constructed with diagonal braces be-
tween the two, this produces a rigid support element in both directions,
called a three-boom truss. > Fig. 45 ◼

7:20 AM
43
 axonometric drawing section

Fig. 45: Three-boom truss

Slab
Timber or steel constructions are almost always directional systems,
i.e. the bar-shaped sections means that loads are always dissipated in a
particular direction. Concrete, however, makes it possible to created stat-
ically non-directional, flat structural components.

Reinforced concrete Fundamentally, the following applies to the loadbearing properties

of reinforced concrete: as artificial stone made of cement, water and
­aggregates such as gravel or chippings, concrete is very good at absorb-
ing compression forces, but like masonry does not absorb tension forces
well. It is therefore usually combined with steel.

Reinforcement In this material structure, the concrete absorbs the compression

forces and the steel the tension forces. The previous chapters on the var-
ious support types have already explained where tension forces occur in
structural components. This is precisely where the reinforcing elements
are placed in reinforced concrete, i.e. bar steel is cast in. For slabs, this
is usually needed on the undersides and in peripheral areas. If a rein-
forced concrete slab is used like a continuous girder, steel reinforcement
is also built in on the top side. When concreting a floor slab, steel bars
are usually inserted in the form of mats welded crosswise, enclosed by
concrete on all sides to be able to support the load compositely. The
thickness needed for a concrete floor depends on the span, and is usu-
ally 15–25 cm. > Fig. 46

◼ Tip: The nodes in lattice girders should be con-

structed so that the sections – or more precisely, their
centre lines – meet precisely at one point. This avoids
generating forces that twist the node and would sub-
ject it to additional loading.

44 7:20 AM
 compression zone
approx. 15–25 cm
statically effective height
steel reinforcement (tension)

concrete covering

Fig. 46: Section through a reinforced concrete slab with a lower reinforcement layer

Concrete slabs are almost the only structural components that can
be non-directional. Over a square space, a concrete floor can dissipate
its loads to all four walls at the same time. But if the floor is rectangular
the loads will be dissipated via the short span in the first place, because
if the deflection is even it will be more heavily loaded than the long span,
thus producing greater tensions. For a concrete floor that is twice as long
as it is broad, the proportion of loading absorbed by the long span is
scarcely significant. But the reinforcement is not simply fitted to relate
to the main direction in which the load is borne. A slab is always rein-
forced transversely as well, as the flat effect brings its own advantages.
This means, for example, that point loads are better distributed, and the
forces in the floor remain lower.

The longer the span, the thicker the floor will need to be. But if a floor
is thicker than 25 cm, its dead weight becomes so large that it is scarcely
viable any longer as a solid flat floor. Strictly speaking, only the upper
edge of the floor is effectively involved in dissipating the compression
load, and the steel bars dissipate the tension stresses. The rest of the
structure is actually just a link, or a filler.

If floors are very thick, it makes sense to reduce their dead weight Ribbed floor
by omitting areas from the lower edge to the effective upper zone. The
reinforcement is then placed mainly in ribs, which lie very close to each
other. A ribbed floor can thus accommodate much wider spans than a
flat one.

Another approach to bridging large spans involves using binding Binding joists
joists. Unlike ribs, binding joists are not seen as part of the floor area, but
as beams on which the flat floors will be laid. > Fig. 47 and Fig. 69, page 63

For concrete structures cast on the building site (in-situ concrete Slab beams
structures), binding joists best use the advantages of the monolithic

7:20 AM
45
 binding joist
flat floor
(prefabricated
element)
zero line

ribbed floor

effective width slab beam

floor with zero line

binding joists

Fig. 47: Reinforced concrete floors Fig. 48: Slab beam effect

constructio­n method. Here, “monolithic” means that all the in-situ con-
crete elements, even if they have been concreted at different stages, work
as a continuous structure. So binding joists exploit not only the static
height to the bottom edge of the slab, but also its thickness. Furthermore,
the part of the slab by the beam on both sides enlarges the compression
zone. In such cases, the term slab beam is used. > Fig. 48

Column
Unlike horizontal loadbearing elements, columns are subject to
hardly any bending load, but primarily to normal forces. A very narrow
column cross section would suffice for dissipating normal forces if it were
not for the danger that the column might sag sideways and fail.

Buckling Slender columns run the risk of buckling, but the magnitude of this
risk depends on various factors. The important features for a column are
the loads, the material, and how slender it is. The Swiss mathematician
Leonhard Euler (1707–1783) established how the way columns are fixed
at the top and bottom affects their buckling properties, and identified
four different cases, which are named after him.

Euler cases The Euler cases set out four ways in which columns can be braced or
provided with articulated joints. When buckling, columns adopt the form
of a sinus curve. The way the columns are attached affects the length of
this sinus curve, or the distance between their inflection points, which is
important in its turn for the stability of the column. The length of the col-

46 7:20 AM
 F F F F

SK

SK
SK
SK

Euler case 1 Euler case 2 Euler case 3 Euler case 4

SK = 2 l S K= l S K = 0.7 l S K = 0.5 l

Fig. 49: Column buckling lengths according to Euler

umn in relation to the deformation curve is known as the effective or

buckling length.

Figure 49 shows the four cases with the same column length. Euler
case 1 works on the flagpole principle: the deformation curve is very long,
which is unfavourable in terms of stability. Euler case 2 relates to a col-
umn that is attached by an articulated joint at the top and the bottom.
This case is very common, and the deformation curve or buckling length
is shorter, which makes the column more stable. In Euler case 3, the col-
umn is braced on one side. This bracing stops the column from twisting
at that point and thus reduces the length of the sinus curve, i.e. the buck-
ling length. Euler case 4 with bracing at the top and bottom produces the
shortest buckling length for the column, and is consequently the most
stable variant. ◯

◯ Note: Euler’s buckling behaviour for columns

assumes compression- and tensionproof material such
as steel or wood. Euler’s scheme is not suitable for
dimensioning columns in masonry or concrete.

7:20 AM
47
 z axis
weak axis

y axis
strong
axis

rectangular square circular circular

section section column tube

Fig. 50: Column sections

Slenderness Another major factor for dimensioning a column is its slenderness.

It is easy to assume that slenderness implies a ratio of column length to
its thickness. But this is not the case: thickness is not part of the equa-
tion, but its stability as a ratio of moment of inertia and cross section
area are; and then again it is not the length of the column, but Euler’s
buckling length, as explained above, that is crucial. So the slenderness
of a column is the ratio of its buckling length to its bending strength.

It is possible to use these factors involved in calculating columns to

make a theoretical statement about the best possible shape for them:
columns that are loaded only vertically can buckle in any direction.
­However, they will actually buckle in the direction in which they have the
lowest bending strength. So columns should be equally stable in every
­direction, as would be the case for a square or – even better – a circular
column.

In addition, bending strength in relation to the moment of inertia

makes it possible to draw further conclusions about the ideal cross sec-
tion. With respect to the distribution of tensions in a buckling column it
is clear that the areas some distance away from the plane of zero ten-
sion or the centre are the most effective. In tubes, the less effective cen-
tral areas are omitted. Here the material is placed as far away as possi-
ble from the central point. This suggests that a tube, and ideally a round
tube, is the best possible shape for a column. This deduction is highly
theoretical, and is intended only to explain the loading on a column, as
ultimately a lot of other factors influence their structure, and all these
have to be taken into consideration. > Figs. 50 and 51

48 7:20 AM
Fig. 51: Columns

cable
Cables do not obey any of the rules explained in previous chapters.
If a cable is part of a loadbearing structure, it sags according to the load
suspended on it or its own weight, and changes its form with every change
of load. It cannot resist bending moments, and always takes up the form
in which no moment occurs anywhere. This form corresponds precisely
with the bending moment of a girder, rather than a cable.

So the “funicular line” corresponds with the cable’s moment curva- Funicular line
ture. > Fig. 52

A second important difference from the loadbearing structures dis- Sag

cussed previously is the fact that cable loadbearing structures always
have horizontal reactions at the support as well. Cables dissipate all loads
as normal forces, i.e. the funicular force, and correspondingly the hori-
zontal reaction force, follow the direction of the cable at the support
­exactly. Only if the cable were hanging vertically would the reaction at
support be vertical alone. > Fig. 13, page 20 In Figure 53, comparing the two
­cables shows that the vertical proportion of the funicular force corre-
sponding to the magnitude of the loads remains the same, while the hor-
izontal proportion changes with the angle of the cable, i.e. with its slack.

Something we have all experienced with a tight or slack washing line Funicular force
is an important factor for all cable loadbearing structures: a low sag level
means a high funicular force, and a high sag level a low funicular force.

So why are they not used much more often? Cable-stayed struc-
tures have their vagaries in practical applications. The magnitude of the
de­formations they admit causes great difficulties in built structures.

7:20 AM
49
 loaded cable

S1V
S1
S1H

f1
single load

S2V
loaded simply supported beam S2
S2H

f2
M S funicular force
f sag

Fig. 52: Funicular line – moment curvature Fig. 53: Funicular force dependent on sag

­ ncontrolled movements – flapping in the wind, for example – have to

U
be completely suppressed in order to avoid large dynamic stresses. Ca-
ble-stayed structures therefore have to be stable in form in every case,
which few methods can achieve. One possibility is to weight the cable
structure so that possible changes of load or wind loads remain low in
comparison with the dead weight of the structure. This solution is avail-
◯ able for suspended roofs, for example.

The disadvantage here is that the additional loads, usually applied in

the form of prefabricated concrete parts, actually mean losing the advan-
tages of the cable structure, and also that correspondingly large funicu-
lar forces have to be dissipated.

◯ Note: The cables in cable-stayed structures are

made of high-strength steel. Many thin steel wires,
with a diameter that varies according to cable type,
are twisted around each other to form strands.
These strands are then made into cables.

50 7:20 AM
 formal stability formal stability through formal stability through
through weight reinforced plane counter-bracing cable
suspended roof suspension bridge Jawerth truss

Fig. 54: Reinforcing cable-stayed structures

Another solution is reinforcement with flexurally rigid structural ele-

ments. In suspension bridges, for example, the suspended carriageway
is so flexurally rigid that it reinforces the whole bridge.

A cable-stayed structure can also be reinforced by counterbracing

with additional cables. This can be done in a variety of ways. Two-dimen-
sional beams can be manufactured, e.g. the Jawerth truss, which is rem-
iniscent of lattice structures, but actually has nothing in common with
them. All the cables in systems of this kind are so highly pretensioned
that no cable slackens, even under the highest possible load. This means
that the structure remains stable in form and capable of loadbearing.
>  Fig. 54 In two-dimensional structural elements made up of cablenets,

­rigidity in the system is achieved by prestressing areas that curve in

­opposite directions to each other. > Chapter Loadbearing structures, Plate structures

arch
If a loaded cable is fixed and turned over, we have a form that dissi-
pates loads as compression forces, and not tensile forces. This is the
ideal arch form, since, like a cable, it dissipates load only as normal forces.

This ideal form, which can be established by calculation or by a draw- Resistance line
ing method, is called a resistance line.

Arch and cable also have other things in common.

The arch also dissipates vertical and horizontal forces in both sup- Arch height
ports, and as with the cable, the height of the arch, measured from floor
or base level to apex, is linked with the magnitude of the horizontal forces:
the shallower the arch, the greater the proportion of horizontal forces
working as compression forces, known as arch thrust. > Fig. 56

7:20 AM
51
Fig. 55: Arcuated loadbearing systems

The crucial difference between cable and arch lies in the fact that as
solid arch, unlike a cable, cannot follow a change of load by changing its
shape. A resistance line as an exact arch form applies only to an individ-
ual load position. If the load changes, the resistance line changes as well.
This means that both normal force and bending moments are created in
an arch. There are various ways in which arch structures can deal with
◼◯ these problems.

Masonry arches usually have a very large dead weight. Because the
working load is small in comparison with the dead load, there are few
consequences for the resistance line if it changes. The arch remains sta-
ble. Arches can also be reinforced with additional structural elements.
For example, if masonry is raised round an arch in the form of a wall, it
prevents the arch from deforming or losing its loadbearing capacity. It is
also possible to make arches of rigid materials such as laminated timber
or steel. Here, the static height of the arch support must be large enough
to be able to absorb the moments as well as the normal forces. > Fig. 57

◼ Tip: Genuine arcuated loadbearing systems should ◯ Note: Arcuated loadbearing systems are derived
not be confused with arcuated bending beams. An arch from masonry construction. Since masonry can absorb
whose horizontal forces are not absorbed by both sup- only compression forces, all the apertures have to be
ports can dissipate its forces only by bending. spanned by arches. Old masonry structures present an
opportunity to study many sophisticated arcuated load-
bearing systems and the skilful treatment of arch
thrust.
Further information about masonry arches can be
found in Basics Masonry Construction by Nils Kummer,
Birkhäuser, Basel 2007.

52 7:20 AM
 from base to apex
load dissipated through
normal force
high dead weight

height
resistance line
AH

Av B

surrounding masonry

resistance line altered

by load change

flexurally rigid arch section

Fig. 56: Load dissipated by arches Fig. 57: Arch reinforcement

In arches, we distinguish between three different statical systems:

two-articulated arches, three-articulated arches, and arches without
­articulation.

A two-articulated arch has articulated supports. They absorb hori- Two-articulated arch
zontal and vertical forces, but no moments. The question of what would
happen if a support were lowered shows that this is a statically undeter-
mined system.

Adding one more articulation, usually at the apex of the arch, turns Three-articulated arch
this statically undetermined into a statically determined system. This
makes hardly any difference to the loadbearing properties, but the
­advantage in terms of construction engineering is that an arch is easier
to transport in two parts. The articulation is created by the fact that the
two parts of the arch are then simply leaned against each other at the
apex and screwed together.

Bracing the supports makes the arch more rigid because the braces Arch without
prevent any distortion caused by bending moments. The effect can be articulation
(braced arch)
compared with columns supported as in Euler cases 2 or 4, and the brac-
ing makes the structure more rigid here as well. Braced arches are stat-
ically undetermined. They are very rare because effective bracing re-
quires a very elaborate construction. > Fig. 58

7:20 AM
53
 forces at support

two-articulated arch

tension rod

three-articulated arch

braced arch chain of arches

Fig. 58: Statical systems in loadbearing arches Fig. 59: Horizontal force dissipated in load­bearing
arches

Arch thrust There are various ways of handling the horizontal forces produced.
Either the supports can be constructed so that they make it possible to
dissipate the arch thrust, or a tension rod can be inserted between the
supports to balance the horizontal forces in one with those in the other.
If several arches are built adjacent to each other, the horizontal forces
acting on the connected supports cancel each other out, so that only ver-
tical forces have to be dissipated. > Fig. 59

Frame
A simple loadbearing system consists of two columns with a beam
or truss above them. But this system is not stable until the columns have
articulated support at the top and bottom. Stability can be achieved by
connecting the horizontal beam to the columns in a way that is flexurally
rigid. This produces an efficient system, a frame.

Rail In a frame, the horizontal members are called rails and the supports
posts.

Post When the rails and posts are joined rigidly, they behave as though
the beam is running “round the corner”. So if the rail bends under a load,
it also transfers the bending force into the posts. These would deflect
outwards if they were not supported. The supports thus resist deforma-
tion and the stresses are addressed by the structure as a whole. The

54 7:20 AM
Fig. 60: Corners of steel frames

posts also limit sagging in the rails. So each rail does not function like a
simply supported beam, but is partially restrained.

This is also clear from the moment gradient. A characteristic feature

of a frame is the moment at support created by the restraining effect of
the posts in the corners of the frame. This reduces the moment of span
of the frame rail. The advantage for the loadbearing capacity here is that
of a continuous, as opposed to a simple, support. The moment at sup-
port reduces the moment of span, which means that the dimensions of
the beam can be smaller. > Fig. 61

It also becomes clear that the corners of the frame are subject to a
high load by the moment at support. They have to be constructed care-
fully, in order to have the required flexural strength. So that the simplest
possible structural elements can be prefabricated, it makes sense to man-
ufacture the rails and posts separately and join them together only at the
building site. But this exacerbates the problem with the flexurally rigid
corners, which nevertheless create another essential advantage for the
system. We said at the outset that it is only the flexurally rigid corners
that make the frame into a stable system. They reinforce it longitudinally,
which is important for skeleton structures. A frame in a statical system
has a similar function to a complete shear wall, and can be used to rein-
force built structures. > Fig. 61 and Chapter Loadbearing structures, Reinforcement

The frames in Figure 59 are shown with two articulated supports. Two-articulated frame
They are called two-articulated frames and, like two-articulated arches,
are statically undetermined systems.

A further articulation can be added to frames as well as to arches, to Three-articulated

make the system statically determined. This makes very little difference frame

7:20 AM
55
 vertical distributed load horizontal distributed load

support deflection disc action

AH BH BH

Av Bv Av AH Bv
bending line bending line

M M

Fig. 61: Frame subject to horizontal and vertical distributed load

two-articulated frame three-articulated frame restrained frame

other three-articulated frame forms couple roof

Fig. 62: Various frame types

56 7:20 AM
to the loadbearing capacity, but the construction can benefit from it ­under
certain circumstances, especially as this third articulation can be placed
in a variety of ways. It can be in the middle, in the ridge or even in a cor-
ner of the frame. Because the bending moments are zero at the articula-
tion point, the construction can be more filigree here than in the areas
with greater bending moments.

The rigidity of frames can be further increased by bracing the posts Restrained frame
into the supports. They are then called restrained frames, but are seldom
used, because restraining the posts is a very elaborate process. > Fig. 62

7:20 AM
57
58 7:20 AM
 Loadbearing structures
Buildings are complex three-dimensional structures, and at first their
loadbearing systems seem immensely elaborate and difficult to analyse.
But fundamentally all construction types are derived from two principles:
solid construction and skeleton construction. These two principles have
been applied since the earliest days of building, and all the techniques
so far invented follow them; the same rules apply to ancient clay huts or
pile dwellings as to the modern building industry’s complex systems.
­Figure 63 shows a specimen ground plan as a solid structure, a skeleton
structure and in some hybrid constructions.

solid construction
Solid structures are made up of flat elements that dissipate vertical Disc
and horizontal loads. Wall-like discs can be loaded vertically as well as
horizontally in their longitudinal direction. But conversely, they have
barely any transverse loadbearing capacity, i.e. via their surface. > Fig. 64
Discs or walls can fail in a variety of ways; they can buckle or fall over.
When building using solid techniques they are protected from this through
reinforcement by other walls, placed at certain intervals adjacent to or
intersecting them. The walls support each other mutually and this makes
a sold structure stable.

A structure of this type is also called modular. We distinguish ­between Modular construction
loadbearing, reinforcing and non-loadbearing walls. Non-loadbearing method

walls can be removed without affecting structural stability. Reinforcing

walls are also deemed to be loadbearing in standards. As a rule, load-
bearing walls are thicker, which enables them to dissipate the ceiling
loads.

solid construction hybrid construction skeleton construction

reinforcement needed

Fig. 63: Specimen ground plan as a solid, skeleton or hybrid construction

7:21 AM
59
 vertical loads
can take little
horizontal load

disc action longitudinal transverse

longitudinally wall type wall type

Fig. 64: Load directions for disc walls Fig. 65: Load directions for loadbearing walls

Solid structures are divided into longitudinal and transverse wall

types.

Longitudinal wall type If one or two loadbearing central walls run parallel to the long sides
of the building, this is known as a longitudinal wall type; most simple, ur-
ban homes are built on this principle.

Transverse wall type The transverse wall type, also known as crosswall construction, is
suitable for buildings such as hotels and terraced houses, where small
rooms are the principal requirement. It is possible to make a distinction
between these types when using floors with timber beams or prefabri-
cated concrete parts with uniaxially directed stresses. When using con-
crete floors that dissipate their loads in several directions, longitudinal
and transverse walls are usually loadbearing. > Fig. 65

◯ Note: The terms solid or massive construction and

skeleton construction do not mean the same thing for
architects as they do for structural engineers. The
above explanation is couched in architectural language,
which is based on geometry and structure. For struc-
tural engineers, solid construction is a subject in its
own right dealing with masonry and reinforced con-
crete. So structural engineers tend to link the term
solid construction with the material.

60 7:21 AM
 masonry construction reinforced concrete log construction
panel construction

Fig. 66: Solid constructions using different materials

The original solid construction is the masonry building. Masonry walls Masonry
cannot absorb tensile stresses, and have to be reinforced appropriately
to their height, length and thickness. Tensile strains are best avoided by
clear load dissipation without protrusions, shoring or wide apertures. ◯

The chapter on slabs explained that reinforced concrete can also ab- Concrete
sorb tension forces. This means that concrete walls are considerably more
stable than masonry walls, and that solid structures in concrete can be
designed with a much greater degree of freedom in relation to room sizes,
spans, apertures and structural complexity. They can be cast in situ or
constructed from prefabricated parts, which are made up either of small
slabs or of wall elements the size of the room, known as large panels.

Construction using large panels is the popular industrialized build- Panel construction
ing method, and is usually known as large-panel or slab construction. The method

components are fixed together with steel structural elements and con-
crete to create a continuous, monolithic structure.

Although timber construction usually employs the skeleton methods, Timber

there are some structures that are better classified as solid construction.

The first is log construction, where timber sections are piled horizon- Log construction
tally to construct walls. Walls of this kind are stabilized by the halving method

joints used for the timbers at the corners of rooms or the whole building.
The timber industry has progressed in recent years to the extent that for
a few years now there have been panel materials on the market that make
panel construction possible. Some of these are panels glued together
from planks, like laminated timber, and some are plywood panels made

7:21 AM
61
Fig. 67: Skeleton construction

from boards layered crosswise. These panel materials make construc-

tion methods viable that are very different from traditional timber con-
struction methods. These methods are still in development at the time
of writing.

skeleton construction
Skeleton constructions are made up of bar-shaped elements form-
ing a structure like scaffolding. Panel and wall elements are then added
to this structure. The loadbearing structure and the elements that create
the interior spaces are, in principle, two separate systems. > Fig. 68

Fundamentally, skeleton structures are made up of three kinds of

structural elements: the columns, the floor beams including the floor
structure, and the reinforcement structures that absorb horizontal forces.
These structural elements are fitted appropriately to the material at nodal
points, almost always with articulated joints. Joints are articulated if they
are not rigid enough to act as a restraint. They do not have to take the
form of a hinge or similar. In principle, any material that is both compres-
sion- and tension-resistant can be used for skeleton structures, for
­example, timber, steel or concrete. Each of these has its own construc-
tion methods, with a particular set of problems arising from the material
and the methods used in jointing it.

Concrete Probably the commonest material for skeleton structures is rein-

forced concrete, with both in-situ cast concrete and reinforced elements
as viable possibilities. A solid reinforced concrete floor slab is normally
used for in-situ concrete structures, and then only reinforcements
and columns are needed. This simplicity also explains the flexibility and
economic viability of the system. > Fig. 69a But floors, as point-supported
flat structures, admit only limited spans. All the forces from the floor slabs
have to be transferred into the columns, which means that the points of

62 7:21 AM
 building envelope skeleton structure reinforcement and
without reinforcement floor areas

Fig. 68: Structure for skeleton method

a. b. c.
point-supported flat ceiling splayed-head column joists joists (one direction)
(variants)

d. e. f.
joists (both directions) main and subsidiary prefabricated component
joist systems system

Fig. 69: Reinforced concrete skeleton construction methods

7:21 AM
63
 transition from column to floor are very heavily loaded. There is a risk of
the column punching through the floor slab.

Splayed-head To avoid this, the edge can be reinforced in a different way. One
­columns method is to use “splayed-head columns”. > Fig. 69 b

Joists If the spans are too great for this system, joists are used. These run
from column to column like beams, and support the floor slabs on a lin-
ear pattern. Joists can be arranged in a number of different ways. Accord-
ing to the span, they are planned to run in one direction, in both direc-
◯ tions, or as a system of main and subsidiary joists. > Fig. 69 c, d, e

Skeleton structures can also be erected using prefabricated ele-

ments. There are various prefabricated systems containing components
for ceilings, joists, columns and foundations. Transport to the building
site is a crucial factor in terms of size. The structural elements should
ideally not exceed the stipulated dimensions for a lorry-load in order to
be financially viable. All that usually happens at the building site is that
the parts are placed on top of one another and fixed securely into posi-
tion. This means that the joints are articulated in principle.

Pi plates Pi plates are often used as prefabricated floor components instead

of flat floors. They are narrow plates with two ribs, and can be fitted
­together to make floor slabs. They work on the T-beam principle and thus
make large spans possible. Floor slabs of this kind are laid on concrete
beams that are supported on the column brackets in their turn. > Fig. 69 f
and Chapter Structural elements, Slab

Steel Steel structures are almost always skeleton structures. They are usu-
ally built up of standard steel construction sections, called “rolled sec-
tions”, in different profile series. > Fig. 70 The size needed is established
by statical calculation.

Rolled sections Rolled sections are produced to a height of up to 60 cm. If higher

units are needed, they have to be welded from sheet metal; in steel con-
struction, elements up to several centimetres thick are still known as
sheets.

◯ Note: The nature of the floor structure greatly

affects the clear height of each storey, so they should
be included in the structural considerations as early as
when the first sections are drawn. The greater the
width of the floor spans, the greater their structural
height.

64 7:21 AM
 IPE HEB

double-T sections U section

rectangular round tube angle

tube

Fig. 70: The most important steel sections Fig. 71: Corrugated sheet

Corrugated sheets are generally used if large areas have to be cov- Corrugated sheets
ered. They acquire their loadbearing capacity from their trapezoid folds,
and are able to function over large spans and serve as floor or roof struc-
tures. > Fig. 71 ◼

As a rule, steel structural elements are made in a steel construction

workshop in transportable sizes and are then assembled on the building
site. In the workshop, welding is the simplest and best method for man-
ufacturing steel structural elements, but as welding is difficult on the
building site, screws should be used for assembly connections.

It is possible to create flexurally rigid corners with an acceptable

­ egree of effort and expense in steel construction. This means that col-
d
umns and beams can be fitted together to form loadbearing frameworks
to exploit their reinforcing capacities. > Chapter Structural elements, Frame To han-
dle the very strong forces at the frame corners, the connections have to
be much more powerful at these points. For example, for double-T sections,

◼ Tip: All the current construction reference works

give steel construction section tables with precise
dimensions and statical values. Generally speaking,
sections from these series should be used in construc-
tion, as they are obtainable for any steel construction
firm and economical in use.

7:21 AM
65
 joint piece flange plate

articulated joint flexurally rigid joint

(frame corner)

Fig. 72: Articulated and flexurally rigid beam jointing in steel construction

both flanges on a horizontal member have to be fastened to the post with

a flange plate, and the connecting screws should be as far away from each
other as possible. Conversely, for an articulated joint, the rib can be
screwed with a simple sheet metal joint piece. > Fig. 72

Fire protection Although it seems remarkable at first, steel structures are at greater
risk in fire than timber structures. Steel softens when heated to high tem-
peratures, and quite quickly loses its entire loadbearing capacity. Steel
must therefore always be protected from fire in high-rise buildings, for
example by cladding loadbearing members with plaster or with a foam-
ing paint.

Composite The rate at which the steel heats up in a fire can also be lowered by
­constructions installing it in combination with concrete. For example, tubular steel pro-
files can be set in concrete in these composite constructions, or double-
T sections filled with concrete. As well as slowing down the heating pro-
cess, the concrete will ensure a certain residual loadbearing capacity in
the event of fire. > Fig. 73

Timber Timber is the earliest timber construction material. Various cultures

have some very old, but very sophisticated timber construction tech-
niques. This is a complex matter, because there are a number of con-
struction methods and an infinite number of variants and mixtures be-
◯ tween them. Here are the most important categories:

Traditional Traditional timber-frame construction is a pure form of skeleton con-

­timber-frame struction filled with clay or brick. As a craft construction method, it is
c­ onst­r uction
characterized by joints using skilfully created forms, without any metal
connecting devices. Timber-frame buildings are seldom constructed like
this any more, but they are often found in the field of heritage.

66 7:21 AM
 reinforcing steel
welded on

concrete-filled composite section with cavity concrete

tubular steel section

the composite effect is

created by bolts welded on
at certain points

floor with joists

Fig. 73: Composite steel structures

American timber construction using the balloon and platform frame Balloon and platform
methods differs from traditional timber-frame construction in that it uses frame construction

thin log- or plank-like timbers that would have no loadbearing capacity in

their own right and would buckle if the timber cladding that forms the
wall surface did not hold them in position. They work like ribs; they are
stable with the cladding. The method is therefore sometimes called rib
construction. Nailing is the principal jointing device. Constructions of
this type are highly economical and flexible.

Modern engineering timber skeleton constructions have an ideal Engineering timber

loadbearing system from a statical point of view, and can be constructed skeleton
constructions
differently according to use. Materials such as laminated timber or vari-
ous sheet products are used.

Prefabrication is becoming increasingly accepted in timber skeleton Modern timber-frame  

construction. Here, wall and floor elements, in dimensions that can be construction

transported on lorries, are preferred as component sizes. Timber-frame

◯ Note: More information on timber construction can

be found in Basics Timber Construction by Ludwig
­Steiger, Birkhäuser, Basel 2007.

7:21 AM
67
 traditional timber frame balloon frame modern timber frame

Fig. 74: Timber skeleton construction methods

construction seems best suited to this. Prefabricated parts are manufac-

tured consisting of derived timber sheets, onto which loadbearing sec-
tions are screwed. The parts can also be supplied with built-in insulation,
cladding, windows or doors, and then fitted together. Similarly to the
American methods, the frame sections work with the laminated wood
­areas to form a loadbearing framework. > Fig. 74

Reinforcement
When planning skeleton constructions the key aim is usually to dis-
sipate dead weight and vertical working loads, for which floors and
­columns are constructed. But attention must be paid to horizontal loads
as well. The most important horizontal load is wind load, which can act
on the building in any direction. Because the component joints are gen-
erally articulated, skeleton constructions have almost nothing to resist
horizontal loads. They therefore need effective reinforcement, i.e. a con-
struction that can transfer the horizontal loads from the facades into the
foundations.

Reinforcing constructions function as a disc. They can accept hori-

zontal forces longitudinally and dissipate them downwards. In tall build-
ings they work as vertical loadbearing members that can dissipate wind
loads into the foundations from all floors.

Disc action A disc can be solid, usually made of masonry or concrete. Disc ­action
can also be created by a diagonal brace in one compartment of the skel-
eton structure. This brace reacts to compression for loads in one direc-

68 7:21 AM
 disc brace cross frame
compression- and tensionproof
tensionproof

Fig. 75: Reinforcement

tion, and to tension for loads in the other. The same effect is achieved
from two tensionproof crossing diagonals. > Fig. 75 The reinforcing action
of frame systems was also pointed out in the chapter Frames.

Skeleton constructions have to be reinforced transversely and lon-

gitudinally. Reinforcement in each direction is not sufficient because,
considering the ground plan, two disc elements always intersect at one
point. This intersection would then be the point around which the load-
bearing structure could twist and would collapse. To prevent this, we
need a new plane of reinforcement, which can be positioned as wished,
but it must not intersect with the other two at the same point. > Fig. 76 a

Reinforcing structures can be arranged on the ground plan in differ-

ent ways. But they should be placed near the centre, because otherwise
the long section of the building would acquire a long lever arm around
this reinforcement, thus creating powerful forces that would place an
­unnecessary strain on it.

If a skeleton construction is loaded horizontally, all the forces from Floor disc
one direction must be transferred into the wall disc provided for the pur-
pose. This needs a rigid floor disc, as assumed in Figure 76. A floor can
also consist of joists with a covering on top of them. A floor of this kind
is not a disc, because the joists can shift in relation to each other. Not all
the horizontal forces can be transferred into the reinforcing structure,
but intermediate floors can easily be made into rigid discs by adding
braces or cross-braces. > Fig. 77

7:21 AM
69
 a. mistake: b. bad:
common fulcrum badly off centre

c. reinforcement with 3 d. reinforcement with 3

discs and floor slab discs and floor slab
(building core) (façade)

Fig. 76: Arranging reinforcing walls in buildings with floor discs

Building core In high-rise buildings, building cores containing fire escapes and lift
shafts are often used as reinforcing structures. They consist of mostly
closed walls and run from the roof to the foundations, and can act as ver-
tical loadbearing members. In high-rise building, it can be more problem-
atic to dissipate the horizontal loads than the vertical ones, because wind
speeds increase with the height of the building and the effect of the wind
loads is much greater. Although the building core provides reinforcement
in most high-rise buildings, one possibility is to make the whole facade
of the building function into a vertical trussed girder, thus working with
the maximum girder dimensions, i.e. the whole width of the building.

For architects, the key question is whether their design will be ade-
quately reinforced or not, or put it another way, whether it is stable or
not. In addition, a distinction is made between less rigid or more rigid
loadbearing systems. This depends on how generously or how centrally
the reinforcements are arranged. The different reinforcement methods
do not have identical effects, and here too a distinction can be made be-
tween more or less rigid ones.

Halls
The hall concept ultimately means little more than the fact that large
spaces are enclosed, as they can be built using both solid and skeleton
methods, and take any conceivable form. What they have in common is
the large span for the roof loadbearing structure. The roof geometry can

70 7:21 AM
 P

no reinforcement cross-bracing with compression- and

no disc action tensionproof sections tensionproof braces

Fig. 77: Creating rigid floor discs in skeleton structures

also be designed in a variety of ways. It can be focused on draining the

surface of the roof, on an advantageous shape for the roof girders, or on
the way in which the skylights are built into the roof area.

Skylights can be placed longitudinally, or be constructed in the Shed roof

­ irection of the loadbearing structure and as an integral part of it, as in
d
shed roofs. > Fig. 78

Halls thus need a roof loadbearing structure that can handle a large
span. It is an advantage here to the make the roof area a lightweight struc-
ture, as the dead weight is an additional load on the structure as a whole.
A large number of statical systems are available for hall construction.
The most common are described briefly below.

Long girders resting on columns or walls are also known as roof Truss
frames or trusses. Because their support points are articulated, a con-
struction of this kind must be reinforced either by creating a rigid roof
disc and the facades, or by bracing the columns. Roof trusses can be
made of wood, steel or concrete. > Fig. 79

Arches make appropriate loadbearing structures for large spans, and Arch
thus for halls, because the loads are dissipated mainly as normal forces
and not as bending forces. A solution must however be found for the great
horizontal forces at support. The arches either run to the floor, so that the
arch thrust can be transferred into the foundations directly, or they sit on
columns or walls, which then have to be reinforced with structures such
as buttresses. It is possible to use tie members between the ­supports to
balance the horizontal forces on both sides. Then only the vertical forces
have to be transferred into the walls. > Fig. 80 and Chapter Structural elements, Arch

7:21 AM
71
 shed roof
roof lights

Fig. 78: Examples of skylights for halls

Frame Frames are well suited to hall construction. They can be used to
c­ reate all kinds of roof geometries, unlike arch constructions. Asymmet-
rical forms can also be implemented very well with two- and three-artic-
ulated frames. The section dimensions must however always match the
moment gradient, which must be established for the particular geo­metry
and load pattern. > Fig. 81 and Chapter Structural elements, Frame

Beam grid The systems named so far consist of girders, spanning the space in
one direction. These are directed systems. But it is also possible to ­design
loadbearing structures that dissipate their loads on all sides. Load dissi-
pation on several sides makes sense primarily for spaces with approxi-
mately equal spans in both directions. Here, the girders cross over each
other, thus forming a grid. Girder grids of this kind can be made of vari-
ous different materials. They can run through joists in conjunction with
an in-situ cast concrete ceiling, and thus form a monolithic bond. Flexur-
ally rigid connection of each intersection point is more laborious at the
assembly stage for steel and timber.

Three-dimen­sional Trussed girders can also be extended to form a three-dimensional

frameworks loadbearing system. These are then called three-dimensional frameworks,
and are defined by the design of the bar and node components. Three-
dimensional frameworks are almost always made of prefabricated steel
elements. > Fig. 82

Reinforcement Reinforcement or stiffening for halls obeys the laws explained in the
previous chapter. But other points must also be borne in mind. For exam-
ple, reinforcing just one loadbearing axis is not sufficient for halls above
a certain size, because the loads within the loadbearing structure travel
a great distance before they are transferred, and so the structure as a
whole would not be rigid enough.

72 7:21 AM
 lattice truss truss using prefabricated truss using laminated
concrete truss elements timber elements
(braced columns)

Fig. 79: Examples of roof truss structures

two-articulated arch symmetrical two-articulated frame

two-articulated arch with tie member two-articulated frame as lattice

arch on buttresses asymmetrical three-articulated frame

Fig. 80: Arcuated loadbearing systems Fig. 81: Frame loadbearing systems

Long girders dimensioned appropriately for their span are at risk of

buckling in their transverse direction. > Fig. 44, page 43 There is a risk of fail-
ure through great imposed loads, or through wind load against the gable.
To prevent this, a joining construction is usually added at roof level to re-
inforce the gables and transfer the wind loads to the eaves. Adding mem-
bers that run transversely to the girders, the purlins, means that the ad-
ditional trusses are attached to the reinforced areas and thus prevented
from buckling. > Fig. 83 and Chapter Loadbearing structures, Reinforcement

7:21 AM
73
 beam grid three-dimensional framework three-dimensional framework
on rectangle on equilateral triangle

Fig. 82: Beam grid and three-dimensional frameworks

Steel is an ideal material for hall construction. Steel loadbearing

structures are light, have a high loadbearing capacity, and can be used
to implement all conceivable statical systems economically. One great
advantage here is that flexurally rigid joints can also be created easily.

Wood is also a very efficient material for hall construction. Arches,

trusses or frames in laminated timber, or lattice trusses made of solid
timber sections can be used. Flexurally rigid joints for frame systems can
be created with laminated timber.

Concrete halls are always made from prefabricated parts. Their load-
bearing systems differ from those in other halls in that the columns are
usually clamped into the foundations, but the trusses are articulated at
the support points. Reinforced concrete trusses are usually manufac-
tured in adjustable steel formwork, so the system allows little flexibility
in choosing the girder geometry.

Plate structures
The chapter Structural elements discussed arches and cables, which
dissipate their loads as compression or tension forces, unlike girders sub-
ject to bending loads. This load dissipation principle can also be imple-
mented in three dimensions by using plate structures. A large number of
different concepts and many variants occur in their design. The most im-
portant groups are named below, to give a general idea.

Folded plates/shells Folded plates are made up of flat surfaces and acquire their loadbear-
ing capacity from the disc action of these areas, while shells are curved
loadbearing systems that can differ considerably in their forms. > Fig. 84

Beam-like plate Shells or folded plates can span from support to support as long sec-
­structures tions, like beams. Fitting them together then forms a roof. When handling

74 7:21 AM
Fig. 83: Example of reinforcement for a hall

large spans, it is important to construct a great statical height with the

lowest possible dead weight. Beam-like plate structures are very suit-
able for this, because of their curved or folded girder sections. Their load-
bearing action is most closely allied to that of corrugated sheets. > Fig. 71,
page 65 Beam-like plate structures must be supported at the edges, to pre-
vent them from collapsing as a result of lateral deflection. > Fig. 85

As with loadbearing systems based on arches and cables, plate struc- Tension-/
tures are distinguished according to their loading type. compression-loaded
plate structures

Domes, shells and similar loadbearing structures are compression- Domes and shells
loaded in some areas and tension-loaded in others. The more continu-
ously their periphery can be supported, the better they will dissipate loads.

Conversely, all suspended constructions, such as cablenets and Cablenets and

membrane constructions, are only tension-loaded. Concrete structures membrane
constructions
can be tension-loaded as well. They are supported by flexurally rigid pe-
ripheral beams or cables. Peripheral cables of this kind then dissipate
the powerful tensile forces through guys that run into foundations with
tensionproof anchorage.

Single-curved surface curve in one direction, but are linear in the Single-/double-
other. All curved surfaces that can be made from a flat surface, such as curved surfaces

a sheet of paper, are single-curved. They are always sections of cylinders

or cones. They can be supported at their ends, like beams, or on their
long sides. Unlike beam support, a longitudinally supported, single-curve
shell dissipates its loads according to form by the same action as an arch.

Double-curved means that the shells cannot be formed from flat sur-
faces. Figure 86 shows some examples of this. A double curve makes the
surfaces rigid in three dimensions. It ensures that tension-loaded

7:21 AM
75
Fig. 84: Folded plates and shells

Fig. 85: Beam-like folded plates and shells

surfaces such as cablenets and membranes will not deform, provided

there has been adequate pretensioning. Compression-loaded brick or
concrete shells thus form loadbearing surfaces even when the material
is not very thick.

Single- or counter-­ Shells or domes are double-curved plate systems operating in one
directional curved direction. Both curves point in the same direction. Counter-directional
surfaces
curved surfaces are also called saddle surfaces and usually occur in
cable­nets or membrane constructions.

Shells or domes are double-curved plate systems operating in one

direction. Both curves point in the same direction. Counter-directional
curved surfaces are also called saddle surfaces and usually occur in
cable­nets or membrane constructions.

76 7:21 AM
 single-directional curved surfaces

counter-directional curved surfaces (saddle surfaces)

Fig. 86: Double-curved surfaces

foundations
The subsoil is part of the loadbearing structure, as well as the foun- Subsoil
dations, and like all the other structural elements it must be able to han-
dle the forces it has to accept. Like every other building material, it
­responds to loads with deformations, which can involve sinking several
centimetres. Sinking is thus a normal facet of loadbearing behaviour and
is not deleterious.

The subsoil usually has a much lower loadbearing capacity than other
building materials. In order to prevent the acceptable stresses being
­exceeded, the loads to which the building subjects it must be distributed
over an adequately large foundation area. Loads spread over a wide area
in the subsoil, which means that the stress under the foundation dissi-
pates rapidly with increasing depth under the footing.

There are many types of soil, and they respond to loads in different Soil type
ways. The key factor affecting its properties is the grain size or the grain
size mixture. The way the soil responds to fluctuations in humidity is also
important. It is therefore essential to collect as much information as
possibl­e about soil material and soil humidity, and about groundwater

7:21 AM
77
 point footing strip footing slab footing

Fig. 87: Footing types

levels. A soil report is now a customary feature of smaller construction

projects as well.

Footing types Footings transfer loads into the subsoil. Soil stresses thus depend
on the area across which the loads are distributed, i.e. on the size of the
footing. A distinction is made between the following footing types:
—— Point footings are usually deployed to absorb the load from individ-
ual columns
—— Strip footings dissipate loads from walls, into the ground, for exam-
ple
—— Slab footings consist of a continuous concrete base that distrib-
utes the loads from the walls and columns standing on over the
whole area of the building. > Fig. 87

Footings can also be supplied to the building site in prefabricated

form, although this makes financial sense mainly for point footings. Fig-
ure 88 shows a bucket footing, into which the column fits as if into a
bucket. Once the prefabricated column has been adjusted precisely, the
joint between the footing and the column is filled with mortar, thus bond-
ing the two prefabricated elements securely.

Deep foundations If no loadbearing subsoil is to be found in the upper strata, it is pos-

sible to dissipate the loads by using deep foundations. To do this, holes
are drilled down to the firm stratum and then filled with concrete. These
bored piles then act as long columns on which the building stands in the
subsoil. The loads are then largely transferred via the tip of the bored pile,
but a firm hold in the subsoil can also be gained via the roughly concreted
surface shell of the pile. > Fig. 89

78 7:21 AM
 prefabricated column

adjusting the column

with wedges

the joint is filled

with mortar after
adjustment

prefabricated
bucket footing

Fig. 88: Bucket footing as a prefabricated element

ground
surface

soil
incapable of strip foundation
loadbearing or ice wall

firm soil frost-free

capable of depth
loadbearing

Fig. 89: Deep foundations using bore piles Fig. 90: Frost-free foundations

When the ground freezes, it expands because of the increased vol- Frost-free foundations
ume of the ice it contains. This produces perceptible uneven soil defor-
mations. It is therefore necessary to avoid frost action under the foot-
ings. The ground only freezes to a certain depth below the surface in
winter, so a continuous strip foundation is laid around the edge of the
building, extending into the frost-free area underground. The required
depth depends on the climate and can be anything from 80 cm to over a
metre. > Fig. 90

Cracks in the completed building usually indicate damaged founda- Damaged foundations
tions. Such damage is always the result of irregularities, possibly in the
building, possibly in the subsoil.

7:21 AM
79
 Changing soil properties inevitably cause problems, because gener-
ally speaking each soil type is subject to different degrees of sinking. In
terms of the building, problems can arise because parts of the building
that impose very different loads share foundations, or from having foun-
dations at different depths, because this will always cause different ten-
sions in the subsoil. It this is recognized at the planning stage, suitable
measures can be taken, either to dissipate the loads into the soil evenly,
or to avoid possible damage from different sinking rates, for example by
means of gaps in the structure.

80 7:21 AM
 In conclusion
Basics Loadbearing Systems is intended to provide an approach to
the complex field of loadbearing system theory. The knowledge collected
here should enable students to understand structural contexts, to con-
sider the demands made by support and loads when designing, and thus
to plan their designs realistically and holistically. Designing the loadbear-
ing structure ultimately helps to sharpen designers’ ideas of space and
can also further them by working creatively with the possibilities that
supporting structures offer. Thus, the quality of the loadbearing struc-
ture design is assessed first and foremost by whether it flows from the
design idea, or even helps to shape it. This happens primarily in the plan-
ning tasks whose function and structure make the loadbearing system
the determining element – such as when using large spans. Problems of
this kind can usually be solved only by addressing the design of the load-
bearing system in its full complexity.

Thus, the basics this volume conveys can be expanded upon as part
of the student’s own architectural development process by working cre-
atively or even playfully with loadbearing structures, and by interpreting
the laws of support structures in terms of individual requirements. To
sum up, there are three basic principles to be considered:

1. Loadbearing structural elements should run through all floors to

the foundations in a single line.
2. Spans should be kept as small as possible. Large spans demand
a great deal of expense and effort. They should be deployed only
when large spaces expressly demand them.
3. It is possible to handle large spans without difficulty if sufficient
height is allowed to construct them. Even if nothing is known at
this stage about the nature of the structure, sufficient statical
height should simply be allowed for such spans.

Anyone who would like to go beyond the material introduced in this

book and find out more about loadbearing structures will gain a better
understanding of how structural engineers work; it is then possible for
students to make their own calculations, thus increasing their ability to
determine dimensions more precisely, and enabling themselves to ­design
on the basis of the loadbearing structure itself.

7:21 AM
81
Appendix
Pre-dimensioning formulae
The formulae below can provide provisional results for dimensioning
structural components at the preliminary design stage. They do not fur-
nish conclusive proof of loadbearing capacity.

Floors and ceilings

Concrete floors or ceiling as flat units in multi-storey buildings:
—— Viable at spans up to 6.5 m
—— The formulae apply to simply supported beams
—— Thickness to provide necessary sound insulation at least 16 cm
—— As a point-supported flat ceiling or carried on walls

at a span less than 4.3 m

li (m)
h(m) ≈ + 0.03 m
35

at a span greater than 4.30 m and given limited deflection as a re-

sult of light dividing walls on the floor

li 2(m)
h(m) ≈ + 0.03 m
150

Timber beam floor or ceiling:

—— Distance between beams 70–90 cm
—— Width of beams ≈ 0,6 • d ≥ 10 cm
li
Height of beams h ≈
17
IPE girders:
—— Load around the strong axis
—— Where h = section height in cm, q = distributed load in KN/m,
l = span in m

3
h≈ 50 • q • l2 –2

82 7:22 AM
 HEB girders:
—— Load around the strong axis
—— Where h = section height in cm, q = distributed load in KN/m, l =
span in m

h≈ 3 17.5 • q • l2 –2

Wide-span roof-bearing structures

Laminated timber beams (parallel):
—— Span 10–35 m
—— Distance between trusses 5–7.5 m

Height h = l
17

Timber trussed beams with parallel chords:

—— Span 7.5–60 m
Distance between trusses 4–10 m

Overall height h > l to l

12 15

Steel solid web girder:

—— Span up to 20 m
—— IPE girder up to 600 mm high

Girder height h ≈ l ... l

30 20
Steel trussed girder:
—— Span up to 75 m

Girder height h ≈ l ... l

15 10

7:22 AM
83
LITERATURe
James Ambrose: Building Structures, 3rd edition, Wiley, Hoboken 2011
James Ambrose, Patrick Tripeny: Simplified Engineering for Architects
and Builders, 12th edition, Wiley, Hoboken 2016
Francis D.K. Ching: Building Construction illustrated, 5th edition, Wiley,
Hoboken 2014
Andrea Deplazes (ed.): Constructing Architecture, 3rd, expand. edition,
Birkhäuser, Basel 2013
Heino Engel: Structure Systems, 4th edition, Hatje Cantz, Ostfildern
2009
Thomas Herzog, Michael Volz, Julius Natterer, Wolfgang Winter, Roland
Schweizer: Timber Construction Manual, Birkhäuser, Basel 2003
Russell C. Hibbeler: Structural Analysis, 6th edition, Prentice Hall
­Publisher, Englewood Cliffs/NJ 2005
Friedbert Kind-Barkauskas, Bruno Kauhsen, Stefan Polonyi, Jörg Brandt:
Concrete Construction Manual, Birkhäuser, Basel 2002
Angus J. Macdonald: Structure and Architecture, 2nd edition,
­Architectural Press, Oxford 2001
Bjørn Normann Sandaker, The Structural Basis of Architecture, 2nd, rev.
edition, Routledge New York 1992
G.G. Schierle: Structures in Architecture, USC Custom Publishing,
Los Angele­s 2006
Helmut C. Schulitz, Werner Sobek, Karl-J. Habermann: Steel
­Construction Manual, Birkhäuser, Basel 2000

84 7:22 AM
picture credits
Figure page 8: Colonnade in front of the Old National Gallery, Berlin,
Friedrich August Stüler
Figure page 34: AEG Turbine Hall, Peter Behrens
Figure page 58: Berlin Central Station, von Gerkan, Marg und Partner
Figure 7, left, right, Figure 41, left, ­centre; Figure 55, left, right: Institut
für Tragwerksplanung, Professor Berthold Burkhardt, Technische
Universität Braunschweig

All other figures are supplied by the author.

7:23 AM
85
Series editor: Bert Bielefeld ­ aterial is concerned, ­specifically the rights of
m
Concept: Bert Bielefeld, Annette Gref translation, reprinting, re-use of illustrations,
Translation from German into English: ­recitation, broadcasting, reproduction on
Michael Robinson ­microfilms or in other ways, and storage in
English copy editing: Monica Buckland ­data­bases. For any kind of use, permission
Layout print edition: ­Andreas Hidber of the copyright owner must be obtained.
EPUB production: Kösel Media, Krugzell
This publication is also available as a softcover
Library of Congress Cataloging-in-Publication (ISBN 978-3-7643-8107-3) and in a German
data language edition (ISBN PDF 978-3-0356-1255-4;
A CIP catalog record for this book has been ISBN EPUB 978-3-0356-1173-1).
­applied for at the Library of Congress.
© 2007 Birkhäuser Verlag GmbH, Basel
Bibliographic information published by the P.O. Box 44, 4009 Basel, Switzerland
­German National Library Part of Walter de Gruyter GmbH, Berlin/­Boston
The German National Library lists this publica-
tion in the Deutsche Nationalbibliografie; ISBN 978-3-0356-1282-0 PDF
­detailed bibliographic data are available on the ISBN 978-3-0356-1207-3 EPUB
Internet at http://dnb.dnb.de.

This work is subject to copyright. All rights are

reserved, whether the whole or part of the www.birkhauser.com

7:23 AM