Professional Documents
Culture Documents
MCCM 69 3 s192-197 Krause
MCCM 69 3 s192-197 Krause
1984
www.ptcer.pl/mccm
Abstract
This article contributes to a meaningful interpretation of results gathered by in situ measurements of sonic velocity, electrical conductiv-
ity and the formation of calcium-aluminium-hydrates during setting and curing of LC refractory castables. All said monitoring techniques
are well known in the refractory community and are well documented in the literature. However, the time dependent changes of the said
properties are not well correlated to mineralogical and in consequence technological changes of the material during setting and curing.
The basic interest of refractory users of course is to define the time at which the installation or the pre-shape construction element can
be demoulded. This is in principle possible with the methods listed above. However, after water addition the time dependent changes
of sonic velocity, electrical conductivity and temperature are as diverse as there are possible combinations of cements, microfines and
surface-active additives. In further the ambient conditions, temperature and relative humidity have a strong influence on these proper-
ties and this does not only mean a simple time-shift. Up to now, the results are more confusing than helpful to determine the best time
for demoulding pre-shapes and refractory linings.
Recent research at Koblenz University of Applied Science contributes to a deeper understanding of the setting behaviour, because
besides the physical evolution of the said parameters the time dependent formation of hydrate phases is also investigated by means of
a gravimetric method. The paper will show further insights in the evolution of refractory castables during setting and curing.
After a two years’ research project performed with Forschungsgemeinschaft Feuerfest e.V. and Institute of Ceramics and Building
Materials, Refractory Materials Division, we can contribute to the understanding as follows: The strength evolution recorded by means
of ultrasonic velocity is mostly dependent on the dispersing agent that leads to typical time resolved strength patterns. However, gravi-
metrical analysis of the water content bond to the hydrate phases determined by the freeze-drying method clearly show that, depending
on the dispersing agent, the first increase of the sonic velocity does not correlate with the formation of hydrate phases. Additional mea-
surements of the electrical conductivity assist the assumption that the said increase does not stand for the formation of hydrate phases.
We suppose that, in the early stage of hardening, gelation or coagulation are the predominant reactions that occur. However, further
investigations are necessary to gain a deeper understanding for this pre-setting reaction in highly dispersed refractory castables.
Keywords: Refractory castables, Setting kinetics, Mixing energy, Cement testing, Hydration process
Artykuł wnosi wkład w pełną treści interpretację wyników zgromadzonych za pomocą pomiarów in situ prędkości fali dźwiękowej,
przewodności elektrycznej i powstawania hydratów glinokrzemianowych w czasie wiązania i dojrzewania niskocementowych betonów
ogniotrwałych. Wszystkie wymienione techniki monitorowania znane są w środowisku technologów zajmujących się materiałami ognio-
trwałymi i dobrze udokumentowane w literaturze. Jednakże, zależność zmian tych właściwości w czasie nie jest dobrze skorelowana
ze zmianami mineralogicznymi, a w konsekwencji technologicznymi materiału w czasie wiązania i dojrzewania.
Głównym przedmiotem zainteresowania użytkowników materiałów ogniotrwałych jest określenie czasu, po którym szalunek lub forma
odlewanego elementu mogą być zdemontowane. W zasadzie jest to możliwe za pomocą przedstawionych powyżej metod, jednak po
wprowadzeniu wody prędkość fali dźwiękowej, przewodnictwo elektryczne i temperatura są tak zróżnicowane jak możliwe kombinacje
cementów, mikroziarna i dodatków powierzchniowo czynnych. Ponad to, warunki otoczenia, temperatura i wilgotność mają silny wpływ
na te właściwości i nie oznacza to tylko prostego przesunięcia w czasie. Jak dotąd otrzymane wyniki są bardziej mylące niż pomocne
w określeniu najkorzystniejszego czasu demontażu form lub oszalowania wyłożeń. Obecne badania przeprowadzone na Uniwersytecie
Nauk Stosowanych w Koblencji pozwalają na głębsze zrozumienie zachowania w czasie wiązania, ponieważ obok fizycznej ewolucji
wspomnianych parametrów badano również metodą grawimetryczną tworzenie się hydratów w zależności od czasu. Artykuł pokaże
dalsze rozeznanie ewolucji betonów w czasie wiązania i dojrzewania. Po dwóch latach realizacji projektu wspólnie z Forschungsgeme-
inschaft Feuerfest e.V i Oddziałem Materiałów Ogniotrwałych Instytutu Ceramiki i Materiałów Budowlanych możemy wnieść następu-
jący wkład w zrozumienie problemu: Zmiany wytrzymałościowe rejestrowane metodą ultradźwiękową są zależne głównie od dodatku
dyspergującego, co prowadzi do typowych zalesności wytrzymałości od czasu. Jednakże analiza grawimetryczna ilości wody związa-
nej przez fazy hydratów, przeprowadzone metodą suszenia przez zamrażanie, jasno pokazują, że zależnie od zastosowanego dodatku
dyspergującego pierwszy wzrost prędkości ultradźwięków nie koreluje z tworzeniem się hydratów. Dodatkowe pomiary przewodnictwa
elektrycznego prowadzą do wniosku, że ten wzrost wytrzymałości nie jest związany z powstawaniem hydratów. Przypuszczamy, że
żelowanie lub koagulacja są dominującymi reakcjami, które występują we wczesnym stadium twardnienia betonu. Konieczne są jednak
dalsze badania w celu głębszego zrozumienia reakcji wczesnego wiązania silnie zdyspergowanych betonów ogniotrwałych.
Słowa kluczowe: materiały ogniotrwałe odlewane, kinetyka osadzania, energia mieszania, cementu, badanie cementu, proces hydratacji
sample weight after the freeze drying, and mh after heat- 4. Results
ing to 850 °C. wp and wh are determined independently and
should sum up to the total amount of water addition. 4.1. Setting and hardening kinetics of the
RTE-castables
m − md
w p = f (1) The hardening of RTE starts between 7.5 and 9 hours
mh and is mostly finished between 12.5 and 14 hours. In
this stage, the sonic velocity increases to approximately
5000 m/s. All three attempts, as depicted in Fig. 1, show
m − mh
w h = d (2) almost this end value. However, the starting point of the
mh hardening varies around 1.5 hours, as a result of slight
Fig. 1. Setting kinetics of the RTE castable (Table 1). Different colours indicate three single measurements of the same material. The
upmost diagram show the evolution of the sonic velocity and the electrical conductivity (dashed line). The diagram in the middle the
temperature progression and in the diagram below, the gravimetric measurement of the hydrate phase content is shown.
variations of the mixing process. This indicates that con- the time of the highest hydrate formation. In this period,
crete setting is sensitive to the mixing procedure. Especially the electrical conductivity decreases rapidly, due to a sig-
V1 shows an initial temperature that lies approximately nificant loss of pore water that is now bond to the hydrate
5 °C higher than for V2 and V3. This primarily indicates phases.
deviations in mixing energy input, because all mixes where For the determination of the hydrate phase content, the
stabilized to a temperature of 19.5 °C to 19.9 °C before the first sample was taken 5 minutes after the mixing process. At
mixing process was initiated. Secondly the higher initial this time, the admixtures already show a significant amount
temperature for V1 leads to an earlier temperature peak at of hydrate phases. Within the first 6 hours, the amount of
11.5 hours that indicates the maximum crystallization rate hydrate phases increases rapidly and remains almost at
of the hydrate phases. Different to that, V2 and V3 show a constant value between (1.8–2)% hydrate water for the
a temperature peak after 12.5 hours. This is remarkable next 6 hours. Between 11.5 and 12.5 hours, a significant
because all mixtures are stabilized to 20 °C in the climate increase of the hydrate phase content is visible that lasts for
cabinet approximately 5 hours after mixing. This indicates about one hour. This period is in good correlation with the
that the initial temperature increase caused by mixing sig- instant of time at which electrical conductivity, sonic veloc-
nificantly influences the setting kinetics. ity, and temperature indicate the maximum formation rate
The temperature peaks of all attempts correlate properly of hydrate phases. Thereafter the crystallization rate fades
with the steep increase of the sonic velocity. More precisely, out to an end value of approximately 2.9%.
the time of the temperature peaks superposes with the sec-
ond turning point of the steep slope of the ultrasonic curve 4.2. Setting and hardening kinetics of the
progressions. The first turning point was already indicated MS-castables
as the first setting that must be correlated with the interlac-
ing of the organic dispersing agent chains [4]. Fig. 2 envisages the curve progressions for the sonic ve-
The development of electrical conductivity typically be- locity, temperature, electrical conductivity, and the hydrate
haves vice-versa to the ultra-sonic signal over the moni- formation of the MS-castables. In this case, the castable
tored 48 hours. In all attempts, the electrical conductivity composition contains 3% microsilica and FS 20 replaces
increases systematically till the temperature peak indicates FS 60 as it was used in the plain RTE-composition.
Fig. 2. Setting kinetics of the MS castable (Table 1). Different colours indicate three single measurements of the same material. The
upmost diagram show the evolution of the sonic velocity and the electrical conductivity (dashed line). The diagram in the middle the
temperature progression and in the diagram below, the gravimetric measurement of the hydrate phase content is shown.
A steep increase of the sonic velocity is already visible ly document the first development of strength, hardening
at about one hour after mixing. Between 10 and 15 hours, should apply for the formation of hydrate phase, and curing
the sonic velocity proceeds with an almost constant velocity for the late hydrate phase evolution due to the formation of
for the next 20 hours. Thereafter, all admixtures show again Katoit, the stable cubic phase C3AH 6.
a significant increase from about 2900 m/s to 4000 m/s. Monitoring the electrical conductivity delivers results that
In contrast to the RTE castable, the electrical conductiv- can be best correlated with the formation of the hydrate
ity exhibits no significant increase at the initial time of the phase. A drop of the specific electrical conductivity always
setting process. In all cases this increase is only visible for correlates with a diminution of pore water due to the forma-
a short period of about one hour. Thereafter the electrical tion of hydrate phases. Therefore, this method does better
conductivity decreases mirroring the curve progression of apply for monitoring the hydrate phases formation in the
the sonic velocity. The temperature profile shows no sig- concrete than the sonic velocity.
nificant peak over the entire monitoring time of 48 hours. The curve pattern of the sonic velocity describes the
The hydrates formation, as it can be directly observed by physical state of the concrete which does not directly cor-
plotting the hydrate pore water against time (Fig. 2), starts relate with the formation of hydrate phases. However, a dis-
immediately after mixing with a shallow slope till 20 hours. tinct dispersing agent leads to a curve pattern that makes
Thereafter the hydrates formation continues with a slightly different dispersing agents distinguishable. Therefore, curve
increased formation rate. There is no indication that the hy- patterns of the sonic velocity allow determining the nature
drates formation is already finished after 48 hour, because of the dispersing agent.
there is still an increase visible following the same slope In conclusion, all discussed monitoring methods provide
of the curves that is already apparent after 20 hours. The a more detailed insight on the reaction kinetics in CAC
derived hydrate phase concentration as a function of time based refractory castables, whereby the monitoring tools
is mirrored by the evolution of the electrical conductivity. deliver complementary results that really show the physi-
Over all monitored parameter it should be obvious, that cal and chemical processes that occur after the addition
the first strength increase derived from the sonic velocity of water. However, all mentioned methods are simply de-
should not be correlated with a strength formation based scriptive methods that cannot shed light on the reactions
on the hydration progress. It should be correlated with the that occur in the castables. Further research is necessary
interlacing of the organic dispersing agent chains or a co- to gain a basic knowledge which processes lead to the
agulation process. early strength that is clearly entailed by the nature of the
dispersing agent.
5. Conclusions
Acknowledgement
The envisaged methods for the investigation of the set-
ting behaviour of hydraulic bond refractory castables show We would like to thank the German Federation of Indus-
in combination a detailed view on the reactions occurring trial Research Associations (AiF) for its financial support of
during setting, hardening and curing. the research project IGF-No. 119 EN. This project was car-
The setting, hardening, and curing process is very sensi- ried out under the auspices of AiF and financed within the
tive to the mixing conditions. Even under strictly controlled budget of the Federal Ministry of Economics and Technol-
mixing conditions as they were defined for the presented ogy (BMWi) through the programme to promote collective
formulations, minor time shifts of the setting and hardening industrial research (IGF).
reaction are unavoidable. Minor time shifts already influ-
ence the green strength if tested after 48 hours. References
The gravimetric determination of the hydraulic bond wa-
ter content over the setting and curing time by the freeze- [1] Götz-Neunhoeffer, F.: Modelle zur Kinetik der Hydrata-
tion von Calciumaluminatcement mit Calciumsulfat aus
drying method clearly show reliable results that enables to
kristallchemischer und mineralogischer Sicht, Erlanger
approximate the hydrate phase concentration at any time Forschungen, Reihe B, Naturwissenschaften und Medizin,
during setting and curing. For a precise determination, the Band 29, 2006.
prevailing phases should be known and can be additionally [2] Roesel, R. E., MacZura, G., Rothenbuhler, P. T.: Calcium
Aluminate Cements for High Strength Refractory Monoliths,
determined e.g. by XRD methods.
Interceram, 5, (1982), 519–523.
The freeze-drying method depicts that it is necessary [3] Kopanda, J. E., MacZura, G.: Production Processes, Prop-
to clearly distinguish between setting and hardening of erties and Applications for Calcium Aluminate Cements in
hydraulic bond refractory castables. Both castables show Alumina Chemicals, Science and Technology Handbook,
Westerville, USA, (1990),171–183.
that the first strength evolution does not correlate with the
[4] Krause, O., Cölle, D., Schaffhauser, H., Jahn, D.: Setting Ki-
formation of hydraulic phases. This behaviour is more ob- netics of High-dispersed Refractory Castables, Refractories
vious for the microsilica containing material and is typical Worldforum, 4, 4, (2012), 81–88.
for all castables in which FS 20 is used. It should be cor- [5] Proske, T.: Frischbetondruck bei Verwendung von selbst-
verdichtendem Beton, PhD thesis, Institut für Massivbau der
related with a setting period where predominantly coagula-
TU Darmstadt, Darmstad 2007.
tion takes place. [6] Boumiz, A., Vernet, C., Tenoudji, F .C.: Mechanical proper-
Therefore, a prescribed terminology will help to prevent ties of cement pastes and mortars at early ages, Adv. Ce-
misapprehension. We propose to use setting, hardening, ment Based Mat., 3, 3–4, (1996), 94–106.
[7] Parr, C., Wöhrmeyer, C.: The investigations into the interac-
and curing in a very strict way whereby setting should sole-
tions of deflocculated castables which control placing and
♦
Received 29 August 2017, accepted 6 September 2017.