See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/273129401

Heat evolution due to cement hydration in

foamed concrete

Article in Magazine of Concrete Research · December 2010

DOI: 10.1680/macr.2010.62.12.895

CITATIONS READS

15 259

4 authors, including:

Elsabé Kearsley Derek Mostert

University of Pretoria University of Pretoria
82 PUBLICATIONS 677 CITATIONS 15 PUBLICATIONS 53 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Grizelda PhD View project

Smart structures View project

All content following this page was uploaded by Elsabé Kearsley on 09 March 2015.

The user has requested enhancement of the downloaded file.

 Magazine of Concrete Research, 2010, 62, No. 12, December, 895–906
doi: 10.1680/macr.2010.62.12.895

Heat evolution due to cement hydration in

foamed concrete
A. S. Tarasov*, E. P. Kearsley*, A. S. Kolomatskiy† and H. F. Mostert*

University of Pretoria; Belgorod Shukhov State Technological University; Institute of Concrete Technology

Foamed concrete is nowadays considered an attractive building material as a result of flexible production
technology and excellent target properties, along with economic benefits. Heat generated in foamed concrete during
cement hydration is found to differ significantly from normal-weight concrete. Foamed concrete that is placed in a
regular-shaped well-insulated mould experiences significant core temperature rises and, although several docu-
ments discussing the thermal effect in foamed concrete have been published, there is still a lack of information
revealing the interrelations of foamed concrete parameters and their effect on temperature rise. Laboratory tests
were carried out with foamed concrete of variable densities and volumes under semi-adiabatic curing conditions.
The observed trends were used as the initial input for a mathematical model of temperature prediction in foamed
concrete. The heat evolution process under real production conditions can be controlled by optimising the mix
composition. If the thermal energy in foamed concrete is properly employed it can result in considerable improve-
ment of the final product characteristics while production costs decrease.

Notation S mass of all solid ingredients in the given foamed

concrete mix (kg/m3 )
A overall surface area of foamed concrete cube in s/c sand/cement ratio
a sealed box (m2 ) w/c water/cement ratio
a/c ash/cement ratio V volume of foamed concrete in a sealed box (m3 )
Cem cement content in foamed concrete (kg/m3 ) Vf volume of foam (l)
Cv volumetric heat capacity of foamed concrete (kJ/ ºi thermal conductivity of constituents (cement,
(m3 8C)) filler, water, foaming agent and the air) (W/
ci specific mass heat capacities of components (m K))
(cement, filler, water, foaming agent and the air) ºv thermal conductivity values for foamed concrete
(J/(kg 8C)) mixes (W/(m K))
D design dry density of hardened foamed concrete ri real density of a constituent in the mix (kg/m3 )
(kg/m3 ) rv casting density of foamed concrete (kg/m3 )
F sum of the weights of filler content in foamed
concrete (ash and/or silica sand etc.) (kg/m3 )
mi mass content of the ingredients in foamed Introduction
concrete as it is prescribed in mix compositions
(kg/m3 ) Foamed concrete is a versatile material which has
RDc relative density of cement become widely used in the construction industry,
RDa relative density of ash mostly for cast-in-situ applications in roof slopes, floor
RDs relative density of sand levelling and insulating layers of wall constructions, as
well as for any kind of void filling (mines, tunnels,
road basements, ground stabilisation, land reclamation
and others), owing to its free-flowing properties, excel-
* Department of Civil Engineering, University of Pretoria, Pretoria lent heat-insulating characteristics, low density and
0002, South Africa
† Belgorod Shukhov State Technological University, Russia
affordable strength values. In this case dry densities of
300–500 kg/m3 are normally applied as there are
(MACR 900180). Paper received 28 September 2009; last revised 28 no strict requirements to strength characteristics, and
January 2010; accepted 16 February 2010 thermal conductivity plays the more dominant role
895

www.concrete-research.com 1751-763X (Online) 0024-9831 (Print) # 2010 Thomas Telford Ltd

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
Tarasov et al.

(Aldridge, 2000; 2005; Basuirski and Wells, 2001; Cox (a) characteristics of the cement
and Van Dijk, 2003; Giannakou, 2002; Jones and (i) volume of cement per cubic metre of foamed
McCarthy, 2005b; Kearsley and Mostert, 2005b; concrete
Tisyachuk and Svinarev, 2005; Vasiljev and Lundyshev, (ii) mineralogical composition of cement clinker
2003; Wimpenny, 1996). and cement fineness
Foamed concrete has also found its application as an (b) water/cement ratio
alternative to autoclaved aerated concrete (ÀÀÑ) pro- (c) heat capacity of 1 cubic metre of foamed concrete
ducts for semi-structural purposes, especially where (i) foamed concrete density
small-capacity production facilities are more economic- (ii) type and amount of aggregate
ally efficient as foamed concrete plants require much (d ) initial temperature of foamed concrete mix
lower capital expenditures owing to the simplicity of (e) size of foamed concrete mix pouring
the technology in comparison with AAC factories ( f ) pore size, distribution and connectivity.
(Tarasov et al., 2005). In this case the most common
dry density range lies within 500–900 kg/m3 as both Environmental factors
compressive strength and insulating properties of the
(a) insulation
foamed concrete need to meet standard requirements.
(b) parameters of the outside environment (air tem-
The most progressive technologies for the manufacture
perature and humidity, air velocity)
of foamed concrete products include high-volume mass
(c) other external influences (pressure, heating, differ-
castings with subsequent cutting of the green concrete
ent types of rays etc.).
into products with given dimensions (Kolomatskiy,
2003). Both types of concrete (normal-weight and foamed
Hence, both cast-in-situ and product items entail concrete) comprise solid, liquid and gaseous phases,
high volumes of foamed concrete to be poured in one but the latter phase is dominant (up to 70% by volume)
section, which may lead to a significant rise in core in the cellular concrete. The gaseous phase (which in
temperature owing to the heat of cement hydration foamed concrete is normally presented by the air) acts
(Jones and McCarthy, 2006; Kolomatskiy, 2003; Tara- as a heat insulator, preventing heat dissipation. The
sov et al., 2005; Tarasov, 2007). Although several docu- evolved heat is transferred mostly through interporous
ments discussing the thermal effect in foamed concrete partitions by means of heat conductivity. It has been
have been published, there is still a lack of information shown (Anikanova, 2007) that in cellular concrete,
revealing the interrelations of foamed concrete para- especially low-density concretes, the heat is also trans-
meters and their effect on temperature rise (Jones and ferred through the heat boundary layer that is formed
McCarthy, 2006; Sakharov, 2002). between a pore and interporous partition, while convec-
The main goal of the present study is to investigate tion is negligible.
the influence of foamed concrete density, volume of The rate of hydration is faster for finer cement, high-
casting and fine aggregate type on temperature profiles er initial temperature of mix and higher water/cement
that can develop in foamed concrete due to cement ratios (Bentz, 1998; De Schutter and Taerwe, 1995).
hydration. Thermal gradients between core and surface These factors were not investigated in this project.
of foamed concrete specimen were also registered to Temperature development in cellular concrete contain-
estimate the risk of cracking. An attempt was made to ing Portland cement seems to be influenced mostly by
create a prediction model for temperature development density and pouring size, with the other factors being
in foamed concretes. comparable to normal-weight concrete. The influence
of cement content and fly ash as fine aggregate on
temperature profiles is, however, unclear.

Parameters influencing heat evolution in

foamed concrete Experimental set-up
Heat evolution in foamed concrete is affected by a The aim of the laboratory investigation was to deter-
greater number of parameters than normal-weight con- mine the effect of foamed concrete density and pour
cretes (Jones and McCarthy, 2006). Results have been size on the temperature profiles. Temperature develop-
published to indicate the significance of the parameters ment in foamed concrete subjected to semi-adiabatic
influencing the temperature development of foamed conditions in insulated boxes was measured.
concrete (Tarasov, 2007). The factors affecting heat The cement used in the investigation is Portland ce-
evolution of cement in foamed concrete can be divided ment (denoted PC) with a density of 3140 kg/m3 , which
into two main groups, namely foamed concrete internal can be classified as a CEM I 42.5 R according to the
parameters and outside environmental factors, as de- South African specification SANS 50197-1:2000
scribed below. (SABS, 2002). The fly ash, with a density of 2200
Internal foamed concrete parameters kg/m3 , is an unclassified ash where approximately 40%
896 Magazine of Concrete Research, 2010, 62, No. 12

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
 Heat evolution due to cement hydration in foamed concrete

of the particles have a particle size exceeding 45 ìm. where x is the cement content (kg/m3 ); w/c is the
The chemical composition of the cement and the ash water/cement ratio (indicating total water to cement
can be seen in Table 1. The foaming agent consists of mass in the foamed concrete mix); a/c is the ash/
hydrolysed proteins and is diluted with water to a 2% cement ratio; s/c is the sand/cement ratio; Vf is the
concentration and then aerated to form foam with a volume of foam (l); RDc is the relative density of ce-
density of 75 kg/m3 . ment; RDa is the relative density of ash; and RDs is the
A density grade (D) was used to calculate the relative density of sand.
foamed concrete mix composition. This density grade The calculation of foamed concrete mix composi-
is the design dry density of cellular concrete (Tarasov tions can be seen in Table 2 and Table 3. In Table 2 the
et al., 2005) stipulated in Russian standard GOST cement content was kept constant for different mix
25485-89 (Gosstandart, 1990). The dry densities of compositions in order to look into the influence of
foamed concrete were chosen in the range D400– foamed concrete density and pour size on temperature
D1200 with silica sand as fine aggregate and D400– development. In the mixtures presented in Table 3, the
D800 with fly ash as fine aggregate. These are the dry cement content varied for foamed concrete of different
densities, which are commonly used in the construction densities while the water/cement ratio was kept con-
industry. The effect of Portland cement content was stant and no fine aggregates were incorporated into the
established while varying the dry densities from D250 mixes.
to D650 without using any fine aggregate. It was Foamed concrete mixes were produced using pre-
assumed that the water bound in the chemical hydration formed foam and mortar slurry mixed in a rotary drum
reactions will result in the addition of 20% to the mass mixer with a horizontal main shaft. The foam was gen-
of the unhydrated cement, while assuming that fly ash erated by diluting a hydrolysed protein-based foaming
plays the sole role of a filler and no amounts of fly ash agent with water to a 2% concentration and then blend-
are hydrated (Kearsley, 1999; Kearsley and Mostert, ing in compressed air until the foam had a density of
2005a; Tarasov et al., 2005). Although the fly ash will 75 kg/m3 . As soon as the homogeneous mortar was
start taking part in the hydration process, the rate of prepared, the preformed foam was added and incorpo-
reaction is a function of temperature and to simplify rated. The plastic density of the foamed concrete mix
the modelling process this variable was not included in was measured with a tolerance of
50 kg/m3 .
the current study. The following equation was produced The foamed concrete was then poured into specially
prepared insulated cubic moulds of different sizes with
D ¼ 1:2Cem þ F (1) side lengths of 100, 400 and 800 mm. The ‘hot curing’
boxes were constructed with 100 mm thick polystyrene
where D is the design dry density of hardened foamed sheets covered by 15 mm plywood boards reinforced
concrete (kg/m3 ), Cem is the cement content (kg/m3 ) with wooden beams. Once the box was filled with
and F is the sum of the weights of filler content in foamed concrete, it was levelled by trowel and covered
foamed concrete (ash and/or silica sand, etc.) (kg/m3 ). with a polystyrene sheet and plywood on top. K-type
Besides the dry density formula (see Equation 1), the thermocouples were placed into the centre and surface
sum of the volume of all the constituent materials of each foamed concrete mass using copper tubes filled
should be 1 cubic metre (or 1000 litres) (Kearsley, with synthetic oil. The layout and illustration of the
1999; Kearsley and Mostert, 2005a; Kearsley and experimental set-up is shown in Figure 1.
Wainwright, 2001) as shown in the formula below The temperature development caused by heat of ce-
ment hydration in foamed concrete was recorded using a
x xð a=cÞ xð s=cÞ Graphtec GL 400/350 Midi Logger (Graphtec Corpora-
1000 ¼ þ xð w=cÞ þ þ þ Vf (2)
RDc RDa RDs tion, Japan). The data were captured at 10 min intervals.
Afterwards all received data were transferred to compu-
Table 1. Chemical composition of Portland cement and fly
ters and processed to obtain final temperature profiles.
ash

Main oxides Portland cement:* % Fly ash: %

Experimental results
CaO 61.7 5.0
SiO2 21.2 54.8 Temperature profiles
Al2 O3 4.6 31.7
The effect of foamed concrete density and pour size
Fe2 O3 1.8 3.8
Na2 O 0.1 0.8 was examined with the mixes as shown in Table 2 (keep-
K2 O 0.7 0.8 ing the cement content constant). Dry densities of foamed
MgO 4.3 1.1 concrete with fly ash varied from D400 to D800, and with
SO3 2.0 0.3 sand varied from D400 to D1200. The temperature pro-
Loss on ignition 0.8 0.8
files obtained are given in Figure 2 and Figure 3, for fly
* Phase composition: C3 S ¼ 45.8%, C2 S ¼ 26.2%, C3 A ¼ 9.2%, ash and silica sand as fine aggregates, respectively.
C4 AF ¼ 5.5%. Temperatures inside the 800 mm and 400 mm cubes
Magazine of Concrete Research, 2010, 62, No. 12 897

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
Tarasov et al.

Table 2. Foamed concrete mix compositions with fly ash/silica sand* (constant PC content)

Name of the component Specific heat Thermal conductivity Unit Content in 1 m3 of foamed concrete for density grades:
characteristic capacity, at 08Ñ: coefficient, at 208Ñ:
J/(kg 8Ñ){ (W/(m K))§ D400 D600 D800 D1200

300 300 300 

Cement 750 1.55 kg
300  300 300

40 240 440 
Fine aggregate 750 1.73 kg
40  440 840

111:3 157:8 201:8 

Water 4180 0.604 litre
111:8  179:8 247:8

773 639 503 

Foamy 4180 0.604 litre
777  549 340

0:95 0:78 0:62 

Air entrained 1005 0.023 kg
0:93  0:66 0:41

509 746 980 

Calculated casting density kg/m3
510  961 1413

476 750 978 

Tested average casting density kg/m3
497  926 1423

* Upper values in the table correspond to mixes with fly ash, and lower values with silica sand.
† Heat conductivity and heat capacity values valid for foamed concentrate are assumed to be the same as water parameters.
{
Bentz, 1998; De Schutter and Taerwe, 1995; Springenschmid, 1998.
§
Bentz, 1998, 2007; De Schutter and Taerwe, 1995; Kim et al. 2003; Zaporozhets et al., 1966.

Table 3. Foamed concrete mix compositions without aggregates, w/c ratio 0.35

Name of component/characteristic Unit Consumption of materials per 1 m3 of foamed concrete and values for target
density grade

D250 D350 D650

Cement kg 220 300 540

Water litre 73.5 105 189
Foam litre 853 800 639
Calculated casting density kg/m3 347 466 777
Tested average casting density kg/m3 325 432 732

reached their maximums within 12–18 h from casting. for D400 foamed concrete (the lowest density in these
Peak temperature values in the 800 mm cubes are 1.75– sets of experiments), both for fly ash and silica sand
2.25 times higher than in 100 mm cubes for fly ash, and fine aggregates. This trend becomes more evident as
2.05–2.15 times for those with silica sand, depending pouring size is increased.
on the foamed concrete density. Foamed concrete in As soon as peak temperatures are reached the tem-
400 mm cubes evolves much more heat in comparison perature starts going down as the amount of the subse-
with 100 mm cubes, and the margins of the registered quently evolved heat is no longer sufficient to maintain
temperatures were 1.7–1.95 and 1.55–1.85 times higher temperature growth. The heat continues to be consumed
for fly ash and silica sand, respectively. It can be seen by constituent products and dissipated into the environ-
from these graphs that cube size drastically changes the ment through the box insulations. The slopes of the
heat evolution curves of the foamed concrete. temperature curves beyond the turning point depend on
The other observed peculiarity of the foamed con- the heat-accumulating ability (thermal inertia) of the
crete heat evolution is related to the material’s density. foamed concrete, which is mostly determined by the
The lighter the material, the better it conserves the heat coefficient of thermal diffusivity.
inside the mass, preventing its dissipation from the The curves in Figure 4 illustrate temperature profiles
cube centre. The highest temperature values, as well as for foamed concrete where the cement content is varied
the highest rate of temperature increase, were recorded as shown in Table 3. This foamed concrete contained
898 Magazine of Concrete Research, 2010, 62, No. 12

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
 Heat evolution due to cement hydration in foamed concrete

°C

hour
Graphtec GL 400/350 Midi Logger
PC

K-type thermocouple
in oil bath

Plywood board
scaffolding

100 mm
100 mm expanded cube mould
polystyrene
800 mm
cube mould
Foamed concrete 400 mm
placed into cube cube mould
mould

Figure 1. Layout and illustration of sealed-box foamed concrete curing test

100 Foamed concrete 100 Foamed concrete

(fly ash fine agregate) (silica sand fine agregate)
cast in cubic insulated moulds cast in cubic insulated moulds
90 90
D800, 800 mm
Foamed concrete core temperature: °C

Foamed concrete core temperature: °C D1200, 800 mm

D800, 400 mm D1200, 400 mm
80 D800, 100 mm 80 D1200, 100 mm
D600, 800 mm D800, 800 mm
70 D600, 400 mm
70 D800, 400 mm
D600, 100 mm
D800, 100 mm
D400, 800 mm
60 D400, 800 mm
D400, 400 mm 60 D400, 400 mm
D400, 100 mm
D400, 100 mm
50
50

40
40

30
30

20
1 10 100 20
Time after casting: h (logarithmic scale) 1 10 100
Time after casting: h (logarithmic scale)
Figure 2. Influence of pour size and density on temperature
profiles of foamed concrete with fly ash Figure 3. Influence of pour size and density on temperature
profiles of foamed concrete with silica sand
no fine aggregate and the dry densities varied from
D250 to D650. The cube sizes were the same as in the
previous sets of experiments. Size effect in these mix- ture range 20–808C (Malinina, 1977; Shakhova, 2007).
tures had a similar influence on the temperature gradi- These differences caused excessive internal stresses in
ent as foamed concrete with a constant cement content. the foamed concrete matrix, resulting in its collapse.
The density effect was, however, overwhelmed by ce- The other probable reason could be protein coagulation
ment content, which was the dominant influencing fac- at high temperatures that induced irrevocable destruc-
tor in these systems. tion of the porous structure. Foamed concrete mixtures
Fluctuations of temperature curves after more than with a synthetic type of foaming agent were not investi-
2 days following casting, especially for 100 mm cubes, gated in the present paper. Nevertheless, this phenom-
resulted from daily room temperature variations, indi- enon has to be taken into consideration when high
cating insufficient insulation of the moulds. temperatures occur in foamed concrete.
It should also be noted that some temperature curves
in the 800 mm cubes were visibly dropping quickly as Core–surface temperature differentials in foamed
they reached high temperatures. This happened because concrete masses
of different volumetric thermal expansion values of The temperature differentials between the core and
constituents of the foamed concrete, mostly due to air surface of sealed cured foamed concrete with fly ash as
bubbles having significantly (over 100 times) higher fine aggregate were registered in the 800 mm and
coefficient of volumetric thermal expansion values than 400 mm cube moulds. The resulting curves are pre-
cement stone or (over 10 times) water at the tempera- sented in Figure 5.
Magazine of Concrete Research, 2010, 62, No. 12 899

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
Tarasov et al.
100 Insulated cubic mould
Foamed concrete
(without aggregates) 20 400 ⫻ 400 ⫻ 400 mm
cast in cubic insulated
90
Foamed concrete core temperature: °C moulds

80 D650, 800 mm 16
Density grade:

Temperature gradient: °C
D650, 400 mm
D800
70 D650, 100 mm
D350, 800 mm D600
12
D350, 400 mm D400
60 D350, 100 mm
D250, 800 mm
D250, 400 mm 8
50
D250, 100 mm

40 4

30
0
0 12 24 36 48 60 72
20
1 10 100 Time after casting: h
Time after casting: h (logarithmic scale)

Figure 4. Temperature profiles of foamed concrete curing in Insulated cubic mould

sealed boxes with different cement contents 20 800 ⫻ 800 ⫻ 800 mm

Density grade:
The initial temperature differential of 2–38 disap- 16
D800
pears within the first 3 h after casting in low-density Temperature gradient: °C D600
foamed concretes. After 3 to 4 h the difference starts D400
increasing and reaches a maximum after 30 to 40 h in 12
the 800 mm cubes. After 42 h the difference reaches a
maximum of 21.38 for D800 foamed concrete. After-
wards the difference decreases, but after 72 h the tem- 8
perature differences in D800 and D600 foamed
concretes are still high (178 and 98 respectively). Mean-
while D400 foamed concrete has a low difference of 4
temperatures (2–48) for the whole registered period.
In general it can be concluded that the heavier the
0
foamed concrete, and the bigger the size of a mould, the 0 12 24 36 48 60 72
higher the thermal difference. Taking into the account Time after casting: h
the risk of thermal cracks occurring, low-density foamed
concretes are preferable for use in large pours. Figure 5. Influence of density and mould size on core–
surface temperature differentials in foamed concrete (fly ash
fine aggregate)
Discussion of results
tude and time of the peak occurrence. Lighter foamed
Rate of temperature development in foamed concrete concretes always show higher rates of temperature rise
The temperature profiles given in Figure 2 and Fig- due to less heat capacity in comparison with heavier
ure 3 have been interpreted to obtain a rate of tempera- foamed concretes. Considering similar densities in
ture development in the centres of insulated foamed 800 mm cubes, foamed concrete with fly ash shows a
concrete 800 mm and 400 mm cubes (see Figure 6). higher rate of temperature rise than that with silica
The peaks in the curves indicate the maximum rate sand. The D800 foamed concrete with silica sand in
of temperature rise and time of its occurrence in the 400 mm cubes was the only specimen with an unex-
centres of foamed concrete cubes. Regardless of cube pected rate of heat development.
size (either 800 mm or 400 mm) all the peaks lie in The maximum temperature is reached when the rate
the time interval between 5.5 and 8 h after casting; of heat development becomes negative, as can be seen
but for larger (800 mm) cubes the peaks are shifted in Figure 6(b). The heavier the foamed concrete, the
earlier by 0.5–1.5 h. Heat evolution starts accelerating more time it required to get to the maximum tempera-
earlier in larger foamed concrete cubes because ture regardless of the type of aggregate. It is also noted
foamed concrete cores are more insulated by the that it takes more time for higher volumes of material
material itself. (800 mm cubes) to heat up to maximum than for smal-
Definite trends can be observed for both the magni- ler volumes (400 mm cubes). However, foamed con-
900 Magazine of Concrete Research, 2010, 62, No. 12

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
 Heat evolution due to cement hydration in foamed concrete
10
25 3
800 mm cube centre 800 mm cube centre
6 6
Rate of temperature development: °C/h

Rate of temperature development: °C/h

from 0 to 8 h from 8 to 36 h
20 2 3
1 D800, fly ash (fa)
2 D600, fly ash (fa)
3 D400, fly ash (fa) 1 5 2
15 4 D1200, silica sand (ss)
5 D800, silica sand (ss)
5 5 1
6 D400, silica sand (ss)
10
4
4
5 0
D400 fa
D400 ss
0 D600 fa
0 2 4 6 8 D800 fa
Time: h D800 ss
D1200 ss
⫺5 ⫺5
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
Time: h

6 10
25
400 mm cube centre 5 400 mm cube centre
3
Rate of temperature development: °C/h

Rate of temperature development: °C/h

from 0 to 8 h 2 5 from 8 to 36 h
20 1 D800, fly ash (fa)
2 D600, fly ash (fa) 3
3 D400, fly ash (fa) 5
15
4 D1200, silica sand (ss)
5 D800, silica sand (ss) 6
2
10 6 D400, silica sand (ss)
1

4
5 0
D400 ss
4 D400 fa
1 D800 ss
0
0 2 4 6 8 D600 fa
D800 fa
Time: h D1200 ss
⫺5 ⫺5
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
Time: h
(a) (b)

Figure 6. Rate of temperature development in foamed concrete cores of 800 mm and 400 mm sealed cubes

crete with fly ash reaches the peak temperatures earlier retards the dormant and acceleration periods. It also
than that with silica sand. accelerates hydration after the acceleration period.
The test results of the current investigation (see
Effect of aggregate type and cement content on Figure 7) indicate that fly ash contributes towards the
temperature profiles temperature effect in foamed concrete. Comparing
The experimental programme of the present study identical densities (D400, D800) and pouring volumes
allowed investigation of the influence of aggregate type (100, 400, 800), foamed concrete with fly ash showed a
on heat evolution in foamed concrete. The literature higher rate (see Figure 6) and maximum values of tem-
proves (Atis, 2002; Jones and McCarthy, 2005a, 2006; perature development than foamed concrete with silica
Kearsley and Wainwright, 2001) that the replacement sand (see Figure 7), regardless of water consumption.
of a part of cement with fly ash is beneficial, such as in It is interesting to note that for the same cube size
the reduction of core temperatures and crack develop- the difference between fly ash and silica sand foamed
ment, long-term strength improvement and so on. Fly concrete is higher for D800 than for D400 because of
ash can also be used as fine aggregate in foamed con- larger amounts of fly ash introduced into the mix to
crete but this should logically increase the total gener- obtain higher density. This confirms that fly ash con-
ated heat. However, Jones and McCarthy (2006) report tributes to heat evolution and that chemical reactions
that replacement of sand with fly ash as fine aggregate are being activated at elevated temperatures. The dotted
led to a reduction of temperature, although heat capa- trend lines in Figure 7 illustrate the prognosed tempera-
cities of the aggregates are similar (from 700 to 800 J/ ture development in extra lightweight foamed concretes
kg 8Ñ), while it is significantly less than the heat capa- at cube sizes over 400 mm. These types of foamed
city of water (4180 J/kg 8Ñ). The authors explained that concrete could not practically reach their expected
the higher water content in the mixes with fly ash maximums due to extreme (close-to-boiling) tempera-
means a greater amount of energy is required to pro- tures leading to structural collapse and large heat
duce a temperature increment. Langan et al. (2002) losses.
reported that fly ash increases initial hydration but Cement content has an impact on temperature
Magazine of Concrete Research, 2010, 62, No. 12 901

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
Tarasov et al.
100

2·3°C
Insulated cube moulds, ment added to the initial cement content of 220 kg/m3
300 kg/m3 PC content
produces a temperature increase of 7.48C.
90
Predicted
trend lines Effect of foaming agent on the kinetics of cement heat
Maximum temperature: °C

80

13·5°C
evolution
Foam is a two-phase system where air bubbles are
70
surrounded by liquid films. The liquid phase consists of
foam concentrate dissolved in water. Depending on the
60 foamed concrete technology applied, the foam concen-
trate may be incorporated either by mixing all the com-
D400 silica sand
50 ponents in a blending mixer or by adding foam into
D800 silica sand
cement slurry. In any case, foam substance is present in
D1200 silica sand
40 D400 fly ash
the water–cement hardening system, and as soon as the
D600 fly ash foam is no longer required in its functional role of porous
D800 fly ash structure former, it starts to retard cement hydration.
30
0 200 400 600 800 1000 Preliminary calorimetric studies have been conducted
Cube size: mm to evaluate the rate of heat evolution of cement hydration
Figure 7. Effect of aggregate type on maximum temperature in the presence of foaming agent during a 72 h period of
rise in foamed concrete interaction. The limited study was conducted on a small
scale and further investigations will be required.
A sample of 2 g of cement was taken and mixed with
development in foamed concrete under semi-adiabatic
2 g of water or water containing foaming agent. The
conditions, as shown in Figure 8. There is a near-linear
percentages of foaming agent were 0.04, 0.2, 1.0% of
relationship between cement content and maximum
cement mass. The results of the tests are shown in
temperature recorded in foamed concrete of 100 mm
Figure 9.
and 400 mm cube size, and only one point for the high-
According to the results, the foaming agent addition
est cement content and biggest cube size (800 mm)
affects both the magnitude and the rate of heat evolu-
falls out of the relationship, owing to the inability of
tion. Within the first hour of cement and water inter-
the foamed concrete structure to bear such internal
action high heat evolution rates occur which are
temperature- and moisture-derived stresses, resulting in
explained mostly by a wetting effect. Then, after a
cellular skeleton destruction and heat energy release
dormant period, the heat evolution rates increase again,
from the foamed concrete core. The trend line shows
reaching another peak value within 7 to 13 h. The
that a temperature over 1008C should have been
higher percentage of foaming agent has a greater im-
reached hypothetically if total energy potential is used
pact on heat evolution rates. For the sample with 1% of
without any side effects. For the specific laboratory
foaming agent content (of the cement mass), the maxi-
conditions it was established that every 100 kg of ce-
mum rate reaches 2.8 W/mg, that is 1.75 times less
120 than that of the control samples without foam addition.
Foamed concrete However, the integral amounts of generated heat are
structure collapsed
virtually the same for all the samples after a 72 h
100 period of interaction and this constitutes about 435 W/
mg on average.
Maximum temperature effect: °C

⬇ 13·4°C

The foaming agent additive both reduces the rate of

80
heat evolution and delays the time taken to reach the
⬇ 37·6°C

maximum temperature. Most typical foamed concrete

60 mix designs have foaming agent percentages between
0.2 and 0.7% of cement mass. Further research, such as
testing plastic strength development of foamed concrete
40 with different foaming agent content, should be con-
ducted.
100 mm sealed cube
20 400 mm sealed cube
Effect of density and cube size on temperature
800 mm sealed cube
development
0 The temperature profiles as shown in Figures 2–4
0 200 400 600
Cement content: kg/m3
have been studied to establish the influence of the
foamed concrete density on maximum temperatures
Figure 8. Effect of cement content on maximum temperature that developed in the foamed concrete cores under
rise in foamed concrete semi-adiabatic hardening, in comparison with potential
902 Magazine of Concrete Research, 2010, 62, No. 12

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
 Heat evolution due to cement hydration in foamed concrete
6
1
70 2 2

5
60
3 Foaming agent dosage
Heat evolution rate: W/mg

of cement weight:
50 3 4
1 Pure
4 2 0·04%
40
3 0·20% 3
1
4 1%
30
2
20

1
10
4

0 0
0 0·2 6 12 18 24 30 36 42 48 54 60 66 72
Time: h

Figure 9. Kinetics of cement heat evolution with the presence of foaming agent addition

heat evolution due to cement hydration in totally iso- where º i is the thermal conductivity of constituents
lated conditions. (cement, filler, water, foaming agent and the air) as
Heat capacities of each foamed concrete mixture specified in Table 2 (W/(mK)); m i is the mass content
with silica sand as an aggregate were calculated. The of the relevant ingredient in the foamed concrete mix
total heat capacity of one cubic metre of foamed con- taken from Table 2 and Table 3 (kg/m3 ); and r i is the
crete is the sum of the heat capacities of all constitu- real density of the corresponding constituent in the mix
ents. The volumetric heat capacity of the foamed (kg/m3 ).
concrete mix (Cv ) can be calculated using Equation 3 Calculations of heat capacity and thermal conductiv-
X ity values of foamed concrete were done for the initial
Cv ¼ ci mi (3) stage of hardening and results of the calculations are
given in Table 4. The values will change as water binds
where c i is the specific mass heat capacities of compo- into hydrates having lower heat capacity values. The
nents (cement, filler, water, foaming agent and the air values are also prone to change as temperature in-
taken from Table 2) (J/(kg 8C)); and m i is the mass creases with time owing to cement heat evolution
content of the ingredients in foamed concrete as it is (Bentz, 1998). These changes are complicated and war-
prescribed in mix compositions in Table 2 and Table 3 rant further investigations.
(kg/m3 ). In fully isolated conditions the heat resulting from
Thermal conductivity values for foamed concrete cement hydration will be used only to heat the constitu-
mixes (ºv ) have been calculated using Equation 4, ents. Based on the heat capacity equation the resultant
assuming that the material structure is uniform and the temperature increment can be calculated showing the
total value is determined from a volumetric contribu- maximum possible temperature to be developed in
tion of each component in the given mix, considering foamed concrete. The total heat generated in a period of
thermal conductivity of cement and water as fresh ce- 3 days is determined as the product of cement content
ment paste that depends on initial water volume frac- and integral value of specific heat evolution. The results
tion, as reported by Bentz (2007). of these calculations are presented in Figure 10.
It can be seen that the effect of foamed concrete den-
X ºi mi
ºv ¼ (4) sity and pour size on maximum temperature in foamed
ri concrete is more evident at lower foamed concrete den-

Table 4. Calculated heat capacity and thermal conductivity of foamed concrete mixes,
300 kg/m3 PC content, silica sand fine aggregate

Characteristics Unit Values for dry density grades

350* 400 800 1200

Volumetric heat capacity, Cv kJ/(8Ñ m3 ) 915.70 966.90 1479.40 1997.80

Thermal conductivity, ºv W/(m K) 0.19 0.28 0.53 0.84

* Values valid for foamed concrete mix without aggregate.

Magazine of Concrete Research, 2010, 62, No. 12 903

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
Tarasov et al.

sity and larger volume of material (cement content being temperature rise, the heat losses played a dominant role
constant). However, the theoretical thermal potential is and resulted in reduced reliability of the recorded data.
not reached by any of the mixes in the investigation. The Semi-adiabatic conditions resulted in extreme tem-
difference between calculated and measured values indi- perature development. It is believed that this can be a
cates that truly adiabatic conditions did not exist. The function of numerous parameters, the same as those for
thickness of the polystyrene layer will have to be in- heat evolution described in the earlier section on ‘Para-
creased significantly to ensure fully adiabatic conditions. meters influencing heat evolution in foamed concrete’.
For the given volume of foamed concrete pouring, Taking into account all obvious factors that should have
the highest values of internal temperature would, there- an impact on the maximum foamed concrete tempera-
fore, be reached in foamed concrete of the lowest den- ture in the current experimental programme, the follow-
sity that is practically achievable keeping cement ing general function was derived
content unchanged while meeting other product re-

Tmax ¼ f Cem, ðCem=S Þ, V , rv , Cv , ºv , A (5)
quirements (e.g. design compressive strength, non-col-
lapsible structure, foamed concrete mix workability, where Tmax is the maximum registered temperature in a
etc.). Reducing cement content for the purpose of sealed box core of foamed concrete (8C); Cem is the
further decrease of foamed concrete density would in- cement content in foamed concrete (kg/m3 ); S is the
evitably result in decreased maximum temperature, as mass of all solid ingredients in the given foamed con-
the generated heat is then mostly consumed by heat crete mix (kg/m3 ); V is the volume of foamed concrete
capacity or lost to the environment. The resulting heat in a sealed box (m3 ); rv is the casting density of
is no longer enough to produce the same temperature foamed concrete (kg/m3 ); Cv is the volumetric heat
effect. capacity of foamed concrete (kJ/(m3 8C)); ºv is the
thermal conductivity coefficient of foamed concrete
(W/(m 8C)); and A is the overall surface area of foamed
Prediction of temperature rise in foamed concrete cube in a sealed box (m2 ).
Heat evolution is directly related to cement content
concrete
in foamed concrete. Cement-to-solid ratio in the func-
Foamed concrete, as any other building material, is tion considers the presence of inert fillers which are
produced under environmental conditions. Heat evolu- accumulators of evolved heat. The more filler there is
tion is a spontaneous process that should be adjusted in the solid phase, the lower is the maximum tempera-
either to isothermal or adiabatic thermal conditions, not ture. Volume of foamed concrete mass represents the
only to enable mathematical estimation and prediction influence of cube size on heat evolution. Foamed con-
of temperature development but also to minimise the crete density, heat conductivity and heat capacity are
influence of heat losses (Springenschmid, 1998). How- combined in a term of thermal diffusivity that describes
ever, even in laboratory conditions it was not possible the velocity of temperature change in the mass and it
to exclude heat losses. Owing to unexpectedly high depends on the physical characteristics of the initial

170 Calculated values

Calculated trend effect of
cement content on peak 300 kg/m3 800 mm3
temperatures PC constant,
150 400 mm3
silica sand
100 mm3
800 mm3
130
PC varies, 400 mm3
Temperature: °C

no aggregate 100 mm3

110 3
540 kg/m PC Calculated values
content
Calculated maximum
90 temperature under ideal
adiabatic conditions
220 kg/m3 PC
content
70

50

30
0 200 400 600 800 1000 1200
Dry density of foamed concrete: kg/m3

Figure 10. Influence of foamed concrete density on the maximum recorded temperature rise in comparison with calculated
thermal maximums for foamed concrete under adiabatic conditions
904 Magazine of Concrete Research, 2010, 62, No. 12

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
 Heat evolution due to cement hydration in foamed concrete

ingredients in the mix. Overall surface area is an indir- adjusted model needs to be developed to predict core
ect parameter of heat losses of foamed concrete masses, temperatures in foamed concrete.
with higher surface area resulting in higher heat losses.
Multi-variable regression analysis was conducted to
establish the effect of each of the identified parameters
Conclusions
on the temperatures recorded. The best correlation was
found for the following function The experiments have shown that foamed concrete
 0:17 with dry densities ranging from D250 to D1200 experi-
: a Cem2 V 2 ence high temperature development under semi-
Tmax ¼ 120 29 (6)
S A2 adiabatic conditions due to heat of cement hydration.
The rate of temperature rise as well as maximum tem-
where a is the thermal diffusivity coefficient of foamed perature reached are affected by numerous factors, the
concrete (m/s2 ) using Equation 7 most important of these being cement content, plastic
ºv density and pour size of foamed concrete.
a¼ (7) The investigations have shown that the highest tem-
Cv rv
perature peaks are recorded for foamed concrete with
Actual measured temperatures are compared to tem- the lowest possible density and large-sized moulds. The
peratures predicted using Equation 6, and this compari- higher heat-insulating capacity of low-density foamed
son is shown in Figure 11. The computed temperatures concretes helps to conserve the heat inside, initiating
compare well with measured temperatures. The R2 sta- further acceleration of cement hydration. The inner
tistic indicates that the model as fitted explains 89.3% layers of large masses are at almost adiabatic condi-
of the variability in temperatures. More efficient heat tions; however, the heat still dissipates outwards. The
insulation of moulds would probably give a still higher pour size plays one of the dominant roles in tempera-
correlation. Although this function can only be used to ture development and the size effect is more noticeable
predict temperatures under particular laboratory condi- at lower foamed concrete densities.
tions, it does reflect the general trend indicating that the There is an inverse relationship between the peak
most dominant factors influencing temperature develop- temperature under semi-adiabatic conditions and
ment are cement content and volume of pouring. foamed concrete density with constant cement content.
The prediction models proposed by many researchers The lowest thermal differences between core and sur-
(Alshamsi, 1994; Babitsky, 2005; Ballim, 2004; Bentz, face of cubic moulds were recorded in the lowest den-
1998; De Schutter and Taerwe, 1995; Kearsley and sity concrete and medium-sized moulds. In this case the
Mostert, 2005a; Ross, 1949; Springenschmid, 1998; temperature difference lies in the range 4–68, which
Zaporozhets et al., 1966) were developed specifically should not result in problems with cracking. Heat-insu-
for cement pastes or normal-weight concretes and lating foamed concretes (D400 and lower) are more
neither thermal properties at lower densities of cellular efficient for applications in large pours.
concretes nor volume of casting were considered. An Temperature rise in foamed concrete is directly re-
lated to cement content. Mix designs should therefore
include as little cement as possible. This can be
100 achieved by partial replacement with extenders, such as
fly ash.
Existing hydration models should be carefully imple-
80 R 2 ⫽ 0·894 mented to foamed concrete in order to take into ac-
count the main influencing aspects which are assumed
Measured temperature: °C

to be negligible in normal concretes (e.g. thermal

60 characteristics of the constituent materials and density).
Negative consequences of heat of hydration in
foamed concrete should be taken into account when it is
40 used in large volumes in order to prevent or to diminish
mix settling, long-term strength losses, shrinkage, crack
development and other problems. As test results showed,
20 foamed concrete subjected to self-heating may experi-
ence porous structure collapse if temperatures reach 908.
However, the natural heat resulting from the exothermal
0 reaction of cement hydration in foamed concrete should
0 20 40 60 80 100
Predicted temperature: °C
also be considered as a favourable factor speeding up
curing. Further investigation is necessary to obtain pre-
Figure 11. Measured temperature plotted against predicted diction models for temperature development in foamed
temperature in the foamed concrete concrete of different shapes.
Magazine of Concrete Research, 2010, 62, No. 12 905

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
 Tarasov et al.

on the Properties of Foamed Concrete. PhD thesis, Leeds Univer-

Acknowledgements sity, UK.
Kearsley EP and Mostert HF (2005a) Designing mix compositions of
The authors would like to express appreciation for
foamed concrete with high fly ash contents. Proceedings of an
the research grants made by Duraset branch of the International Conference on Uses of Foamed Concrete Global Con-
Aveng Group and the THRIP-program of the South struction: Ultimate Concrete Opportunities, Dundee, Scotland.
African National Research Foundation to carry out the Thomas Telford, pp. 29–36.
work reported in this paper. Kearsley EP and Mostert HF (2005b) Opportunities for expanding the
use of foamed concrete in the construction industry, Proceedings of
an International Conference on Uses of Foamed Concrete Global
References Construction: Ultimate Concrete Opportunities, Dundee, Scotland.
Thomas Telford, pp. 143–154.
Aldridge D (2000) Foamed concrete for highway bridge works. Semi- Kearsley EP and Wainwright PJ (2001) The effect of high fly ash
nar Notes on Foamed Concrete: Properties, Applications and Po- content on the compressive strength of foamed concrete. Cement
tential. University of Dundee, Scotland, pp. 33–41. and Concrete Research 31(1): 105–112.
Aldridge D (2005) Introduction into foamed concrete: what, why, Kim K-H, Jeon S-E, Kim J-K and Yang S (2003) An experimental
how? Proceedings of an International Conference on Global Con- study on thermal conductivity of concrete. Cement and Concrete
struction: Ultimate Concrete Opportunities, Dundee, Scotland. Composites 33(3): 363–371.
Thomas Telford, pp. 1–14. Kolomatskiy AS (2003) Cement hardening processes in foam con-
Alshamsi AM (1994) Temperature rise inside pastes during hydration crete. Proceedings of the International Conference Foamconcrete-
in hot climates. Cement and Concrete Research 24(2): 353–360. 2003, Belgorod, 108–116.
Anikanova T (2007) Heat-Insulating Foamed Concretes with Acceler- Langan BW, Weng K and Ward MA (2002) Effect of silica fume and
ated Setting. Technical Science Dissertation, Belgorod Shukhov fly ash on heat of hydration of Portland cement. Cement and Con-
State Technological University. crete Research 32(7): 1045–1057.
Atis CD (2002) Heat evolution in high volume fly ash concrete. Malinina LA (1977) Thermal Treatment of Heavy-weight Concrete.
Cement and Concrete Research 32(5): 751–756. Stroyizdat, Moscow.
Babitsky VV (2005) The prognoses for the concrete hardening Ross AD (1949) The prediction of temperatures in mass concrete by
kinetics with the thermos processing of structures. Magazine of numerical computation. Magazine of Concrete Research 1(1):
Construction Materials, Equipment and Technologies of XXIst Cen- 9–21.
tury 4(75): 66–67. SABS (South African Bureau of Standards) (2002) SANS 50197-
Ballim Y (2004) A numerical model and associated calorimeter for 1:2000 Cement, Part 1: Composition, specifications and conformity
predicting temperature profiles in mass concrete. Cement and Con- criteria for common cements. SABS Standard Division, South
crete Composites 26(6): 695–703. Africa.
Basuirski J and Wells D (2001) The use of foamed concrete in Sakharov GP (2002) Non-autoclaved porous concrete and a problem
construction and civil engineering. Conspectus: 64–73. of energy saving. Formula of Construction Journal 1(12): 20–21.
Bentz DP (1998) Prediction of adiabatic temperature rise in conven- Shakhova LD (2007) Production Enhancement of Non-Autoclaved
tional and high-performance concretes using a 3-D microstructural Foamed Concrete with the Given Properties. PhD thesis, summary,
model. Cement and Concrete Research 28(2): 285–297. Belgorod Shukhov State Technological University, Belgorod.
Bentz DP (2007) Transient plane source measurements of the thermal Springenschmid R (1998) Prevention of Thermal Cracking in Con-
properties of hydrating cement pastes. Materials and Structures crete at Early Ages. RILEM R Technical Committee, New York,
40(10): 1073–1080. USA, Report 119-TCE.
Cox L and Van Dijk S (2003) Foam concrete for roof slopes and floor Tarasov AS (2007) The Increase in Foamed Concrete Effectiveness
leveling. Concrete 37(2): 37–39. Owing to Internal Energy Potential. Technical Science thesis, Bel-
De Schutter G and Taerwe L (1995) General hydration model for gorod Shukhov State Technological University.
Portland cement and blast furnace slag cement. Cement and Con- Tarasov AS, Lesovik VS and Kolomatskiy AS (2005) Industrial man-
crete Research 25(3): 593–604. ufacture of foamed concrete products. Proceedings of the Interna-
Giannakou A (2002) Potential of foamed concrete to enhance the tional Conference Poreconcrete-2005, Belgorod, pp. 128–143.
thermal performance of low-rise dwellings. Proceedings of the Tisyachuk VV and Svinarev AV (2005) Production and application of
International Conference Concrete 2002: Innovations and Develop- monolithic foamed concrete in construction. Proceedings of the
ments in Concrete Materials and Construction, University of Dun- International Scientific and Practical Conference Poreconcrete-
dee, Scotland. Thomas Telford, pp. 533–543. 2005, Belgorod, pp. 97–102.
Gosstandart (1990) GOST 25485-89 Cellular concretes: Specifica- Vasiljev VD and Lundyshev IA (2003) Experience of monolithic
tions. Gosstandart, Russia. foamed concrete use in construction. Proceedings of the Interna-
Jones MR and McCarthy A (2005a) Utilising unprocessed low-lime tional Conference Foamconcrete-2003, Belgorod, pp. 105–107.
coal fly ash in foamed concrete. Fuel 84(11): 1398–1409. Wimpenny DE (1996) Some aspects of the design and production of
Jones MR and McCarthy A (2005b) Behaviour and assessment of foamed concrete. Appropriate concrete technology. Proceedings of
foamed concrete for fill and highway applications. Proceedings of the International Conference on Concrete in the Service of Man-
an International Conference on Uses of Foamed Concrete Global kind (Dhir RK and McCarthy MJ (eds)), London, pp. 245–254.
Construction: Ultimate Concrete Opportunities, Dundee, Scotland. Zaporozhets ID, Okorokov SD and Pariyskiy AA (1966) Heat Evolu-
Thomas Telford, pp. 61–88. tion of Concrete. Stroyizdat, Leningrad.
Jones MR and McCarthy A (2006) Heat of hydration in foamed
concrete: effect of mix constituents and plastic density. Cement and
Concrete Research 36(6): 1032–1041. Discussion contributions on this paper should reach the editor by
Kearsley EP (1999) The Effect of High Volumes of Ungraded Fly Ash 1 June 2011

906 Magazine of Concrete Research, 2010, 62, No. 12

Delivered by ICEVirtualLibrary.com to:

IP: 137.215.9.20
On: Wed, 08 Dec 2010 14:21:21
View publication stats