International Journal of Concrete Structures and Materials

Vol.11, No.1, pp.17–28, March 2017

DOI 10.1007/s40069-016-0175-2
ISSN 1976-0485 / eISSN 2234-1315

Effect of Elevated Temperature on Mechanical Properties

of Limestone, Quartzite and Granite Concrete
Muhammad Tufail1), Khan Shahzada1), Bora Gencturk2),* , and Jianqiang Wei2)

(Received April 3, 2016, Accepted October 11, 2016, Published online December 27, 2016)

Abstract: Although concrete is a noncombustible material, high temperatures such as those experienced during a fire have a
negative effect on the mechanical properties. This paper studies the effect of elevated temperatures on the mechanical properties of
limestone, quartzite and granite concrete. Samples from three different concrete mixes with limestone, quartzite and granite coarse
aggregates were prepared. The test samples were subjected to temperatures ranging from 25 to 650 C for a duration of 2 h.
Mechanical properties of concrete including the compressive and tensile strength, modulus of elasticity, and ultimate strain in
compression were obtained. Effects of temperature on resistance to degradation, thermal expansion and phase compositions of the
aggregates were investigated. The results indicated that the mechanical properties of concrete are largely affected from elevated
temperatures and the type of coarse aggregate used. The compressive and split tensile strength, and modulus of elasticity decreased
with increasing temperature, while the ultimate strain in compression increased. Concrete made of granite coarse aggregate showed
higher mechanical properties at all temperatures, followed by quartzite and limestone concretes. In addition to decomposition of
cement paste, the imparity in thermal expansion behavior between cement paste and aggregates, and degradation and phase
decomposition (and/or transition) of aggregates under high temperature were considered as main factors impacting the mechanical
properties of concrete. The novelty of this research stems from the fact that three different aggregate types are comparatively
evaluated, mechanisms are systemically analyzed, and empirical relationships are established to predict the residual compressive
and tensile strength, elastic modulus, and ultimate compressive strain for concretes subjected to high temperatures.
Keywords: concrete, fire resistance, limestone, quartzite, granite, mechanical properties.

1. Introduction (phase expansion, condensation, evaporation and vapor

diffusion), chemical (thermo-chemical damage and dehy-
Concrete is the most widely used construction material in dration) and mechanical (thermo-mechanical damage,
the world because of its durability, strength, and resistance spalling and cracking) phenomena take place in concrete
to fire. Structural concrete must satisfy the fire safety that results in deterioration of its properties (Heikal 2000).
requirements identified in building codes (EN 1992; ACI As temperature is increased, the water on the surface of
2014). Although concrete, in general, has a high resistance concrete and the capillary water is lost, and this process is
to fire, its mechanical properties such as strength and elastic accelerated by the reduced cohesive forces between water
modulus decrease when subjected to high temperatures. molecules due to water expansion (Hager 2013). At a
The failure at elevated temperatures is mainly due to for- temperature of 105 C, the free water starts evaporating
mation of cracks parallel to the heat surface, changes in the rapidly. In the temperature range from 80 to 150 C,
chemistry, and an increase in pore pressure as a results of dehydration of ettringite takes place followed by the
evaporation of water (Gluekler 1979; Ali et al. 2004; Peng decomposition of gypsum between 150 and 170 C (Hager
and Huang 2008). At high temperatures, various physical 2013). When the temperature reaches to 300 C, the
chemically bound water starts to evaporate, which in turn
decreases the compressive strength of concrete. As tem-
1)
Department of Civil Engineering, University of perature is further increased, decomposition of portlandite
Engineering and Technology Peshawar, Peshawar, occurs between 400 and 540 C. When temperature
Pakistan. increases beyond 400 C, the concrete strength decreases
more rapidly due to the degradation of calcium–silica–hy-
2)
Sonny Astani Department of Civil and Environmental drate (C–S–H). Second phase of the C–S–H decomposes in
Engineering, University of Southern California, Los the temperature range from 600 to 800 C forming b-C2S
Angeles, CA, USA. (Hager 2013). At a temperature of 900 C, the C–S–H
*Corresponding Author; E-mail: [email protected] breaks down completely. Therefore, the critical temperature
for concrete ranges from approximately 400 to 900 C. In
Copyright  The Author(s) 2016. This article is published
this range concrete loses most of its strength.
with open access at Springerlink.com

17
 Aggregates normally occupy 50–75 % of the volume in temperature resistance and low density are desired for con-
concrete, therefore, the behavior at elevated temperatures is crete. There have also been studies on the effect of tem-
strongly influenced by the type and properties of the perature on high strength and/or fiber-reinforced concrete
aggregate used. Naturally found aggregates are usually with or without supplementary cementitious materials (Poon
stable up to 300 to 350 C. With regards to the high tem- et al. 2001; Cheng et al. 2004; Savva et al. 2005; Husem
perature properties of aggregates, thermal expansion and 2006; Behnood and Ziari 2008; Behnood and Ghandehari
conductivity, and chemical and thermal stability are known 2009; Ghandehari et al. 2009). The literature review here is
to play an important role (Bentz 2007; Bentz et al. 2011). limited to those studies that investigated the effect of
Abrams (1971) reported that carbonate and lightweight aggregate type on high temperature resistance of regular
concretes exhibit similar strength characteristics at high Portland cement concrete without supplementary cementi-
temperatures and retain approximately 75 % of their original tious materials for their relevance to the presented research.
strengths at 650 C, while siliceous concrete yield a lower The objective of this research is to systematically inves-
strength, particularly above 430 C. It was also found that tigate the temperature dependency of mechanical properties
the original strength of the concrete has little influence on the of concrete containing different types of aggregates and to
percentage of residual strength at high temperatures. explore the corresponding mechanisms by considering phase
According to Sakr and El-Hakim (2005), the mechanical transformation, decomposition, thermal expansion and
properties of concrete are inversely proportional to temper- deterioration of aggregates. The focus is placed on lime-
ature, and among the heavy weight concretes for nuclear stone, quartzite and granite coarse aggregates. It is expected
reactors containing gravel, barite and ilmenite aggregate, the that the results from this research will contribute to the
ilmenite concrete, shows the highest density, modulus of existing knowledge by expanding the experimental database
elasticity, strength, and hence, it is the most resistant to high and providing empirical formulas to predict the residual
temperatures. Li and Liu (2016) investigated the tensile mechanical properties of concretes containing different types
properties of hybrid fiber-reinforced concrete exposed to of coarse aggregates subjected to high temperatures.
elevated temperatures from 20 to 900 C, and found that the
basic reason for the degradation of mechanical properties of
fiber-reinforced concrete is the deterioration of its
microstructure by forming a number of pores in matrix and 2. Experimental Program
micro-cracks along the bonding interface between steel
fibers and matrix. Arioz (2007) showed that the relative 2.1 Materials
strength of concrete decreases as the exposure temperature The physical properties of the coarse aggregates, i.e.,
increases from 200 to 1200 C, and the reduction is very limestone, quartzite and granite, were determined according
sharp beyond 800 C. A more pronounced temperature to pertinent standards [i.e., ASTM C-29 (2009), ASTM C-33
dependence is observed in the case of river gravel aggregate (2013), ASTM C-127 (2015) and ASTM C-566 (2013)] as
in comparison to crushed limestone. Koksal et al. (2012) shown in Table 1. Figure 1 shows pictures of coarse
studied the physical and mechanical properties of four dif- aggregates used in this study. As seen in Fig. 1, all the three
ferent concrete mixtures with varying amounts of expanded aggregates were angular. Limestone had the lowest angu-
vermiculite (3–6 % by volume) subjected to temperatures larity, but quartzite exhibited smoother texture than the other
ranging from 300 to 1100 C for 6 h. It was found that two aggregates. Three different concretes were prepared,
expanded vermiculate improves the high temperature sta- each one having one of these coarse aggregates. The fine
bility of concrete significantly, by increasing the critical aggregate was kept the same for all three concretes. The
temperature for loss of strength to 900 C. Therefore, it was physical properties of the fine aggregate were also deter-
concluded that the expanded vermiculate could be a good mined [according to ASTM C-70 (2013) and ASTM C-128
alternative to other lightweight aggregates when high (2015)], and provided in Table 1. The particle size

Table 1 Physical properties of coarse and fine aggregates.

Properties Coarse aggregate Fine aggregate
Limestone Quartzite Granite
Maximum size of aggregate 19 19 19 n.a.
(mm)
Moisture content (%) 1 0.8 0.5 0.56
Water absorption (%) 1.95 1.87 1.78 1.79
Specific gravity 2.42 2.64 2.68 2.65
3
Unit weight (kg/m ) 1540.5 1655.7 1649.4 n.a.
Fineness modulus n.a. n.a. n.a. 2.65
n.a. not applicable.

18 | International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017)
 Fig. 1 Coarse aggregates.

(a) (b)
100 100

80 80
Percent Passing

60 Percent Passing 60

40 40

20 Limestone 20
Quartzite
Granite
0 0
40 30 20 10 9 8 7 6 5 4 5 4 3 2 1 0.5 0.1
Sieve Size (mm) Sieve Size (mm)

Fig. 2 Particle size distribution of a coarse and b fine aggregates.

distribution of the coarse and fine aggregates was determined temperature and across all the temperature-concrete combi-
[according to ASTM C-136 (2014)] as shown in Fig. 2a and nations were within 5 %. Compressive and split tensile
b, respectively. The chemical composition of the coarse strength tests were performed on the samples after being
aggregate was determined by X-ray fluorescence test, and subjected to six different temperatures as described below.
the results are shown in Table 2. Samples from each mixture design were also tested as
control (i.e., no temperature exposure). ASTM C-39 (2014)
2.2 Sample Preparation and ASTM C-496 (2011) were followed in compression and
The control mix (limestone concrete) was designed to tensile tests that were performed in a universal testing
yield a concrete strength of 21 MPa at 28 days. A propor- machine with a loading rate of 1 mm/s. Data loggers were
tioning of 1:1.97:2.97 of cement, fine, and coarse aggregate, attached to the universal testing machine to record the dis-
respectively, with a water-to-cement (w/c) ratio of 0.55 was placements and loading.
determined. The aggregates were mixed in saturated surface The slump values were obtained for all three mixtures
dry condition. To ensure that there are only two factors that according to the ASTM C-143 (2015). The slump values for
affect concrete mechanical properties, namely aggregate type the control mix, quartzite and granite concretes were 86, 88
and temperature, the same proportions as limestone concrete and 85 mm, respectively. The slight differences were
were used for the granite and quartzite concretes. The potentially due to the different shapes and textures of the
amounts of materials used in the control mix to obtain 1 m3 aggregates. As was shown in Fig. 1, the limestone had a
of concrete are shown in Table 3 (which were identical in more rounded shape than quartzite and granite. Although
proportion for the other concretes). quartzite was angular, it exhibited the smoothest surface
Forty-two concrete cylinder specimens, 150 mm in texture among all three aggregates.
diameter and 300 mm in height, were cast from each mix (2 A compressive strength of 20.48 MPa was obtained [ac-
types of tests 9 7 temperatures 9 3 repetitions per test- cording to ASTM C-39 (2014)] for the control (limestone)
temperature combination). The coefficient of variation of all concrete that was not subjected to high temperatures. This
the tests conducted on the same type of concrete at the same was very close to the target compressive strength of 21 MPa.

International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017) | 19
 Table 2 Chemical composition of coarse aggregates.
Element Limestone Quartzite Granite
SiO2 2.69 91.14 63.05
Fe2O3 0.57 2.23 5.85
Al2O3 0.16 3.38 11.26
CaO 53.36 Traces 13.19
MgO 0.21 Traces 4.17
Na2O 0.32 2.71 0.69
K2O 0.41 0.3 1.08
Loss on ignition 42.2 0.1 0.6

Table 3 Amounts of materials for 1 m3 of concrete.

Material Quantity (kg/m3)
Cement 354
Fine aggregate 699
Coarse aggregate 1050
Water 195

Fig. 3 Temperature exposure of concrete cylinders: a rotary kiln, b specimen being subjected to high temperature.

Quartzite and granite concretes yielded compressive reached at the same heating rate. The temperature was con-
strengths of 28.92 and 35.42 MPa, which were 41 and 72 % stantly monitored using an infrared thermometer to ensure that
higher, respectively, than the control mix. it remains constant for the period of exposure, i.e., 2 h. Fire was
put off after two hours and the cylinders were left in the kiln for
2.3 Heating Procedure cooling. The cylinders were removed from the kiln after
After casting, concrete cylinders were cured at room tem- cooling to room temperature. All mechanical tests were per-
perature in a water tank. After curing for 28 days, the cylinders formed afterwards under room temperature (25 ± 1 C).
were burnt in rotary kiln, as shown in Fig. 3, at 95 ± 2 C
(*200 F), 205 ± 2 C (*400 F), 315 ± 2 C (*600 F),
425 ± 2 C (*800 F), 535 ± 3 C (*1000 F) and 2.4 Resistance to Degradation and X-ray
650 ± 3 C (*1200 F) for a duration of 2 h. Six cylinders Diffraction Tests
(pertaining to one mixture and one test temperature) were In addition to the mechanical properties of concrete, the
placed in the kiln at a given time. Fire was burnt at the fire exit effect of elevated temperatures on strength of the three
point of the kiln and the desired inside temperatures were coarse aggregates was investigated. These three aggregates

20 | International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017)
were burnt in the same conditions as concrete (at the six high Limestone and granite yielded highest and lowest loss by
temperatures for 2 h). The resistance to degradation of abrasion, respectively, under elevated temperatures. The
original and heat treated aggregates was determined using a main difference occurs when temperature exceeded 300 C:
Los Angeles abrasion machine according to ASTM C131/ the loss by abrasion of limestone increased by 44.2 %,
C131M (2014). however a slight increase (17.8 %) was observed in granite.
Phase compositions of these three aggregates were As reported by Zhang et al. (2015), 300 C is the damage
investigated on the fine powders using X-ray diffraction threshold temperature for limestone, and in the range from
method. The powder samples of original aggregates, and 300 to 600 C, significant damage occurs due to formation
those of the heat treated aggregates at 315 and 650 C, were of internal defects, such as micropores, cracks and trans-
collected after abrasion. X-ray diffraction analysis was per- granular fractures. This is in agreement with the observations
formed on an X-ray diffractometer with a h–2h configuration of Chen et al. (2009) that microcracks appear when lime-
using CuKa source (k = 1.54 Å). The samples were scan- stone is heated to 300 C and mineral particles crack and
ned in step mode with a step size of 0.05 (2h) at -40 kV clear fractures appear when heated to 500 C. Accordingly,
and 30 mA. based on the results here and data in literature, it may be
anticipated that limestone concrete will have a lower
strength than quartzite and granite concretes after heat
3. Results and Discussion treatment. It will be shown later in this paper that the
strengths of concretes subjected to elevated temperatures
3.1 Effect of High Temperature on Resistance parallel these observations.
to Degradation of Aggregates
The resistance to degradation of the three aggregates under 3.2 Temperature Dependence of Mechanical
elevated temperatures is shown in Fig. 4. Prior to heat Properties of Concrete
treatment, granite and limestone exhibited the lowest and Subjecting the samples to high temperatures decreased the
highest losses by abrasion and impact, indicating that these concrete compressive strength at 650 C to 6.77 MPa, 10.00
two aggregates have the highest and lowest strength, and 19.08 MPa, respectively, for limestone, quartzite and
respectively. This might be attributed to the difference in granite concrete. The compressive strength values at differ-
mineral compositions and aggregate shapes. Limestone and ent temperatures are shown in Fig. 5a. The percent loss
quartzite mainly consisted of calcite and quartz, respectively. values with respect to the original strength of concretes is
The primary constituents of granite were quartz and feld- shown in Fig. 5b. In absolute terms, granite concrete had
spars. With a hardness of 7.0, quartz is harder and stronger higher compressive strength in room temperature and
than calcite under room temperature. However, as shown in retained higher strength at increasing temperatures compared
Fig. 4, quartzite aggregate, which contained the highest to quartzite and limestone concrete. As shown in Fig. 5b,
quartz fraction with an evident angular shape, yielded a granite concrete exhibited a considerably less relative (to
strength value in between. This is believed to be the main original) strength reduction rate than quartzite and limestone
reason for the difference in compressive strength of con- concrete, which showed similar percent loss values with
cretes. Under elevated temperature, cracks gradually gener- respect to their original strength as temperature increased.
ated and the cracking density, intracrystalline crack length, This might be partially attributed to the low temperature
and crack width of aggregates increased. As a result, the dependence of the strength of granite as presented in Fig. 4.
elastic modulus and strength were reduced considerably. As seen in Fig. 5a, a first order polynomial fitted well to the
compressive strength-temperature curve in linear scale. It
was obtained that the compressive strength of limestone,
60 quartzite and granite aggregate decrease by approximately
Limestone 2.16, 2.94 and 2.65 MPa, respectively, for every 100 C
Quartzite
50 Granite
increase in temperature, which were found to be comparable.
Loss by Abrasion and Impact (%)

At 25 C limestone concrete provided a tensile strength of

2.47 MPa while the same for quartzite and granite concretes
40 was 3.14 and 3.99 MPa (27 and 62 % higher), respectively.
At 650 C the tensile strength of limestone, quartzite and
30
granite concretes was 0.60, 0.95 and 1.77 MPa, respectively.
The tensile strength at different temperatures is shown in
Fig. 6a while the percent loss values are shown in Fig. 6b.
20 The line fits indicate that the tensile strength of limestone,
quartzite and granite concrete decrease by 0.29, 0.34 and
0.36 MPa, respectively, for every 100 C increase in tem-
10
0 100 200 300 400 500 600 700 perature. These results indicate that limestone concrete per-
Temperature ( C)
forms slightly better in terms of retaining its absolute
Fig. 4 Resistance to degradation of aggregates subjected to strength with temperature compared to quartzite and granite
high temperature. concrete which showed very similar results. However, from

International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017) | 21
 (a) (b)
40 70
Limestone

Reduction in Compressive Strength (%)

y = 36.17 − 0.0265 x Limestone
35 R 2 = 0.998 Quartzite 60 Quartzite
Compressive Strength (MPa)

Granite Granite
30
50
25
40
20
y = 29.85 − 0.0294 x 30
15 R 2 = 0.994
y = 21.32 − 0.0216 x 20
10
R 2 = 0.995
5 10

0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
Temperature (° C) Temperature (° C)

Fig. 5 a Compressive strength of concretes at different temperatures, b change in compressive strength with temperature.

(a) 4 (b) 80
y = 4.05 − 0.0036 x Limestone Limestone
3.5 R 2 = 0.997 Quartzite
Reduction in Tensile Strength (%)
70 Quartzite
Granite Granite
3
Tensile Strength (MPa)

60
y = 3.22 − 0.0034 x
2.5 R 2 = 0.999
50

2 40

1.5 y = 2.52 − 0.0029 x 30

R 2 = 0.996
1 20

0.5 10

0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
Temperature (° C) Temperature (° C)

Fig. 6 a Tensile strength of concretes at different temperatures, b change in tensile strength with temperature.

Fig. 6b, a similar trend as compressive strength was had the highest ultimate compressive strain passed 200 C
observed in the rate of relative strength loss under increasing followed by limestone and granite concrete. For tempera-
temperature. This again indicated that, given its higher tures lower than 200 C, limestone concrete showed the
original strength and low temperature sensitivity of granite,
40
the granite concrete yields a lower degradation rate of rela- Ultimate Strain
tive strength than limestone and quartzite concrete as the 35
temperature increases.
Compressive Stress (MPa)

30
The stress–strain responses were recorded for all the
25
samples during compressive testing. The results are shown
25 C
in Figs. 7, 8 and 9 for limestone, quartzite and granite 20
95 C
concrete, respectively. The ultimate compressive strains 205 C
15
(taken as the value corresponding to the peak stress) were 315 C
425 C
observed to be different for different concretes and depen- 10
535 C
dent on the temperature. For the same concrete, increasing 5 650 C

the temperature resulted in an increase in the ultimate strain

0
values (hence the ductility of the material). The ultimate 0 1 2 3 4 5
Compressive Strain -3
strain values for all of the mixes at different temperatures are x 10
shown in Fig. 10a while the percent increase values are
Fig. 7 Stress-strain curves for limestone concrete at different
shown in Fig. 10b. As seen in Fig. 10a, quartzite concrete temperatures.

22 | International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017)
 40 change in its ultimate compressive strain with increasing
Ultimate Strain
35 temperature while quartzite and granite concretes showed a
similar behavior.
Compressive Stress (MPa)

30 25 C The compressive chord modulus of elasticity of concrete

95 C
25 205 C was calculated according to ASTM C-469 (2014). At 25 C,
315 C the limestone concrete provided an elastic modulus of 23.0
20
425 C GPa while the same for quartzite and granite concretes was
15 535 C 31.2 and 38.2 GPa (36 and 66 % higher), respectively. The
10 650 C elastic moduli at different temperatures are shown in
Fig. 11a while the percent change is provided in Fig. 11b.
5
The granite concrete had the highest elastic modulus at all
0
0 1 2 3 4 5
temperatures followed by quartzite and limestone concrete.
Compressive Strain
x 10
-3 A linear fit to the data indicated that the modulus of elasticity
of limestone, quartzite and granite concrete decreased 2.41,
Fig. 8 Stress-strain curves for quartzite concrete at different 3.26, and 2.86 GPa, respectively, for every 100 C increase
temperatures. in temperature. The information in Fig. 11b indicates that,
the same observations as for compressive and tensile
40 strength could be made, that is the rate of decrease of the
Ultimate Strain
25 C relative modulus of granite concrete is lower than that of
35
95 C
limestone and quartzite concrete.
Compressive Stress (MPa)

30 205 C
315 C
25 425 C 3.3 Thermal Expansion
535 C
20 Given that concrete is a heterogeneous composite mixture,
650 C
the deterioration of concrete under elevated temperature
15
consists of complex interactions between physical and
10 mechanical processes, such as dehydration and decomposi-
5
tion of cement paste, degradation of aggregates, and damage
to the matrix due to the differences in the thermal expansion
0 behavior of different components. In this study, the dehy-
0 1 2 3 4 5
Compressive Strain
x 10
-3 dration of cement paste can be assumed to be the same for all
concretes, as a result, the differences in their temperature
Fig. 9 Stress-strain curves for granite concrete at different dependencies were due to the degradation of aggregates and
temperatures.
cracking caused by thermal stresses generated by the
highest ultimate compressive strain followed by quartzite imparity in thermal expansion behavior of cement paste and
and granite concrete. Again, a linear relationship was found aggregates.
to represent well the change of ultimate compressive strain Thermal expansion represents the volume change of a
with temperature. Limestone concrete showed the least material due to temperature change and it is important

(a) (b)
-3
x 10
4
Increase in Ultimate Compressive Strain (%)

120
y = 0.0018 + ( 3.4065 ×10−6 ) x
Limestone
3.5 R 2 = 0.993
100 Quartzite
Ultimate Compressive Strain

3 Granite

80
2.5
y = 0.0022 + (1.6823 ×10−6 ) x
2 60
R 2 = 0.984
1.5 y = 0.0015 + ( 2.5471× 10−6 ) x
40
R 2 = 0.993
1
Limestone
Quartzite 20
0.5
Granite
0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
Temperature (° C) Temperature (° C)

Fig. 10 a Ultimate compressive strain of concretes at different temperatures, b change in ultimate compressive strain with
temperature.

International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017) | 23
 (a) (b)
40 80
y = 39.00 − 0.0286 x Limestone

Reduction in Modulus of Elasticity (%)

2 Limestone
35 R = 0.998 Quartzite 70 Quartzite
Granite
Modulus of Elasticity (GPa)

Granite
30 60
y = 32.15 − 0.0326 x
25 R 2 = 0.998 50

20 40

15 30

10 20
y = 23.37 − 0.0241x
5 10
R 2 = 0.996
0 0
0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700
Temperature (° C) Temperature (° C)

Fig. 11 a Modulus of elasticity of concretes at different temperatures, b change in modulus of elasticity with temperature.

because it can result in thermal stresses and structural change of calcite, limestone and limestone concrete have
movement that can lead to concrete cracking and spalling potentially shown the lowest thermal expansion among the
(Naus 2010). Thermal expansion and contraction behavior of three aggregates and concretes, respectively. However, it
concrete varies primarily with aggregate type, cementitious should be noted that, different from the pure volume
material content, w/c ratio, temperature range, concrete age, expansion of quartzite and granite, cracking of mineral
and ambient relative humidity, and of these factors, aggre- particles and the formation of fractures in limestone are the
gate type was found to have the greatest influence on the most important factors for heat expansion of limestone
expansion and contraction of concrete (Yao and Zheng (Chen et al. 2009). Moreover, the anisotropic expansion
2007). Moisture content, w/c ratio, and type of cement were behavior can be observed in these aggregates under elevated
found to only affect the thermal expansion at relatively low temperature (Chen et al. 2009; Plevova et al. 2016). Due to
temperatures, i.e., T \ 200 C (Naus 2010). For a constant the thermal shrinkage of cement paste and the considerable
amount of water, the coefficient of thermal expansion of thermal expansion of aggregates, a net expansion of concrete
concrete reduces with decreasing cement content (Yao and is exhibited and inner thermal stresses are generated. Once
Zheng 2007). In addition, in a temperature range from 100 to these stresses become greater than the Portland cement
1000 C, the coefficient of thermal expansion decreases with tensile capacity or aggregate strength, micro-cracking
increasing porosity of aggregates, therefore, a lower expan- occurs. Under high temperature, the strength of cement
sion is observed in concretes with lightweight (porous) pastes is reduced due to dehydration and decomposition.
aggregate in comparison to those that contain normal weight This in turn reduces the stress threshold for crack generation
aggregates (Uygunoglu and Topçu 2012). It should be noted and propagation. These cracks generate new surfaces that are
that, for temperatures below 150 C, as the temperature is directly exposed to high temperatures, hence the temperature
increased, a volume expansion in cement paste takes place. gradient may significantly change and the cracking deterio-
When the temperature exceeds 150 C, Portland cement ration could accelerate (Pancar and Akpinar 2016).
paste starts contracting with increasing temperature (Cruz
and Gillen 1980; Naus 2010), which is attributed to the 3.4 Phase Transformation
dehydration of cement hydration products, such as calcium In addition to dehydration of cement paste, thermal
hydroxide and C–S–H gels. expansion and cracking, crystal transformation and mineral
Among the three aggregates considered here, quartzite decomposition of aggregates are considered as important
exhibits the highest linear thermal expansion coefficient, reasons for deterioration of concrete under elevated tem-
therefore, the quartzite concrete has potentially yielded a perature. As shown in Fig. 12, the limestone aggregate used
higher volume expansion than limestone and granite con- in this study composed primarily of calcium carbonate
cretes. This might be attributed to the high silica (quartz) (CaCO3) in the form of the mineral calcite and a small
fraction in quartzite. Naus (2010) presented that the largest amount of quartz. As temperature increased, the intensity of
expansion always occurs for aggregates having the highest quartz increased first and then decreased. This is attributed to
percentage of silica by weight. In the presence of anorthite, the partial decomposition of calcite and phase transition of
which has an even lower thermal expansion than calcite, quartz. At atmospheric pressure, decomposition of calcite
granite shows a lower coefficient of thermal expansion than does not start until temperature exceeds 850 C. In fact, the
quartzite. Due to the low temperature sensitivity of volume conversion starts at a temperature of about 600 C, and

24 | International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017)
 (a) Cc 25 °C (b) ZSM (c) 25 °C
25 °C
315 °C Q Q/Ab 315 °C
Q 315 °C
650 °C 650 °C An/Ab 650 °C
ITQ ZSM Ab An
Q Ag
Z
CHClNO Z Q
CcQ Cc Ag

10 20 30 40 50 60 70 80 10 20 30 40 50 60 10 20 30 40 50 60 70
2 theta (θ) 2 theta (θ) 2 theta (θ)

Fig. 12 XRD pattern of mineral components in a limestone, b quartzite and c granite aggregates (Notes: Cc calcite, Q quartz,
Z Zeolite, ITQ calcined ITQ-4 (Si24O48), ZSM ZSM-23 (Si24O48), Ab albite, An anorthite, Ag augite).

finishes at 850 C with a loss of 44 wt % corresponding to indicates that amorphous phases were formed in quartzite
the stoichiometric CO2 amount in CaCO3 (Rodriguez- aggregate under high temperature. As a result, the degree of
Navarro et al. 2009). It should be noted that the decompo- crystallinity of quartzite aggregate decreased by about 78 %.
sition of limestone occurs in a similar temperature range as The main constituents of granite aggregates in this study,
thermal degradation of C–S–H. This partial thermal as indicated in Fig. 12 were quartz and two plagioclase
decomposition of calcite, combined with the inner defects feldspars: albite (NaAlSi3O8) and anorthite (CaAl2Si2O8).
(e.g., cracks and fractures) formed under elevated tempera- Augite and zeolite were also detected in XRD pattern as
tures, resulted in significant changes in mechanical proper- minor phases. With increasing temperature, a same phase
ties of the aggregates (as shown in Fig. 4). Therefore, transition of quartz as that in quartzite aggregate was
although limestone has the lowest thermal expansion, the observed. Given the short heat treatment time, no significant
limestone concrete yielded the lowest strength at all test disordering or phase transformation was detected in the
temperatures and the highest relative rate of strength feldspars within the test temperature range of this study. This
decrease. was found to be in good agreement with the previous
As shown in Fig. 12, the quartzite aggregates mainly observations indicating that low and high albites are
consisted of quartz. Peaks of zeolite and ZSM-23 (Si24O48) stable below 650 and 725 C, respectively, under low
were also detected. In the range of 25 to 315 C, the effect of pressure (disordering rates of low albite on dry heating at
temperature on the phase composition of quartzite was mild: atmospheric pressure is of the order of 150–200 days at
only the change of degree for quartz peaks and the disap- 950 C) (Brown 1989). The anorthite phase is also
pearance of ZSM-23 were detected. The degree of crys- stable under low pressure, and even at high temperature, the
tallinity was increased by 2.4 %. When temperature was breakdown reaction of feldspar anorthite is difficult due to
raised to 650 C, the quartz phase transformed substantially. the slow reaction rate at temperatures below 1200 C
The intensity of zeolite phase decreased and two new peaks (Goldsmith 1980). Given the low fraction of quartz phase,
for calcined ITQ-4 and C22H20Cl2N4O8 were observed. granite was more stable than quartzite under elevated tem-
Under normal pressure, trigonal quartz (also named a-quartz peratures and this agrees well with the temperature depen-
or low quartz) transforms into hexagonal b-quartz (high dence of resistance to abrasion of aggregates as shown in
quartz) at 573 C, and upon further heating the SiO2 trans- Fig. 4.
forms into hexagonal b-tridymite at 870 C and later to According to the test results and analysis above, it can be
cubic b-cristobalite at 1470 C (Wenk and Bulakh 2004; concluded that deterioration of these concretes were caused
Akhavan 2005). In the temperature range of this study, only by different mechanism. Among the three aggregates, the
the first phase transition occurred. The density of a-quartz most serious degradation was observed in limestone, mainly
and hexagonal b-quartz is 2.65 and 2.53 g/cm3, respectively attributed to the formation of inner fractures and cracks of
(Akhavan 2005). As a result, a slight discontinuous thermal mineral particles, as well as the partial decomposition of
expansion caused by phase change was observed. As this calcite under higher temperature. As a result, the highest
process is reversible only when the temperature changes relative strength reduction was yielded by the limestone
very slowly, it is believed that there was residual hexagonal concrete at all test temperatures, although this aggregate had
b-quartz after cooling. In addition, it should be noted that a the lowest thermal expansion among the three groups. In
wide dispersion peak was found between 6.5 and 13 (2h) contrary to limestone, the significant thermal expansion
for the quartzite sample after 650 C (Fig. 12). This behavior of quartz phase and its phase transition when

International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017) | 25
temperature exceeded 573 C were identified to be the main under increasing temperatures. However, the lowest rate
reasons for the reduction of strength in quartzite concrete. of increase of ultimate compressive strain was yielded by
Given its low quartz fraction (indicating low thermal the limestone concrete.
expansion and low impact from phase transformation) and • Degradation of coarse aggregates was considered as one
the high stability of the two plagioclase feldspars (albite and of the primary reasons for the decreased mechanical
anorthite), granite concrete showed the lowest temperature properties of concretes under elevated temperature. The
dependence of mechanical properties among the three resistances to abrasion and impact of all these three
groups. aggregates decreased with increasing temperature. Lime-
stone and granite yielded the highest and the lowest loss
by abrasion, respectively. The main difference occurred
4. Conclusions when temperature exceeded 300 C, which is the
damage threshold temperature of limestone to form
In this paper, the effect of different coarse aggregates; inner defects.
namely, limestone, quartzite and granite, on the mechanical • Due to high quartz fraction, quartzite exhibits the
properties of concretes subjected to elevated temperatures highest linear thermal expansion coefficient among the
from 95 to 650 C were studied. The following conclusions three aggregates. When the temperature exceeded
were drawn from the obtained results: 150 C, Portland cement paste contracted with increas-
ing temperature. The thermal shrinkage of the cement
• The type of aggregate used in this study had no
paste and expansion of aggregates have potentially
considerable effect on the workability of the concrete
resulted in thermal stresses in the cement pastes
mix as very similar slump values were achieved for
causing micro-cracking. Due to dehydration and
limestone, quartzite and granite. Since the aggregates
decomposition of C–S–H phases, which substantially
were in surface dry saturated condition, the slight
decrease the tensile strength of cement pastes, the
differences in the slump values may be attributed to
damage was significantly accelerated under high
the shape and texture of the aggregates.
temperature.
• At 28 days and no elevated temperature exposure,
• Although no significant phase transformation occurred in
granite and quartzite concrete, respectively, yielded 41
the limestone, a partial thermal decomposition was
and 72 % higher compressive strength, 27 and 62 %
detected under high temperature. Different from quartzite
higher tensile strength, and 36 and 66 % higher elastic
and granite, the thermal expansion of limestone was
modulus in comparison to limestone concrete. This
mainly caused by cracking of mineral particles and the
indicates that properties of aggregate substantially
formation of fractures. Therefore, although it showed the
impact the mechanical properties of concrete.
lowest temperature sensitivity in volume change, lime-
• For every concrete mix, the compressive and tensile
stone aggregate and limestone concrete yielded the
strength, and elastic modulus were observed to be
highest relative strength and modulus reduction among
inversely proportional to temperature; while the ultimate
the three concretes.
strain in compression was directly proportional.
• According to the XRD patterns and the phase diagram,
Although ultimate strength of these concretes were
a phase transition of quartz occurred in quartzite and
reduced by different amounts, their deformability in
granite aggregates when temperature exceeded 573 C.
compression was improved after heat treatment.
Given the low fraction of quartz phase and the
• Granite concrete had the highest compressive and tensile
presences of stable albite and anorthite phases in
strength and elastic modulus at all test temperatures in
granite, less phase conversion and thermal expansion
comparison to quartzite and limestone concretes. The
occurred in granite in comparison to quartzite. There-
compressive and tensile strength, and elastic modulus,
fore, granite exhibited better resistance to degradation
respectively, of granite concrete was 182, 195 and 182 %
than limestone and quartzite at all test temperatures.
higher than that of limestone concrete, and 191, 186 and
This is considered as the main reason for the highest
97 % higher than that of quartzite concrete at 650 C.
resistance of mechanical properties of granite concrete
• A linear relationship represented well the temperature
among the three groups when subjected to elevated
dependency of mechanical properties. The compressive
temperatures.
and tensile strength, and the elastic modulus of lime-
stone, quartzite and granite concrete, respectively,
decreased by 2.16, 2.94 and 2.65 MPa (compression),
0.29, 0.34 and 0.36 MPa (tension), and 2.41, 3.26 and Open Access
2.86 GPa (modulus) for every 100 C increase in
temperature. This article is distributed under the terms of the Creative
• Compared to limestone and quartzite concrete, a lower Commons Attribution 4.0 International License
relative (to original value) rate of decrease of compres- (http://creativecommons.org/licenses/by/4.0/), which per-
sive strength, tensile strength and elastic modulus under mits unrestricted use, distribution, and reproduction in any
elevated temperatures were observed for granite concrete medium, provided you give appropriate credit to the original

26 | International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017)
 ASTM. (2015). Standard test method for density, relative den-
author(s) and the source, provide a link to the Creative
sity (specific gravity), and absorption of fine aggregate.
Commons license, and indicate if changes were made.
West Conshohocken, PA: American Society of Testing
Materials (ASTM). ASTM C128-15.
ASTM. (2015). Standard test method for slump of hydraulic-ce-
References
ment concrete. West Conshohocken, PA: American Society
for Testing of Materials (ASTM). ASTM C143/C143M-15.
Abrams, M. S. (1971). Compressive strength of concrete at
Behnood, A., & Ghandehari, M. (2009). Comparison of com-
temperatures to 1600F. ACI Special Publication, 25, 33–58.
pressive and splitting tensile strength of high-strength
ACI. (2014). Building code requirements for structural concrete
concrete with and without polypropylene fibers heated to
(ACI 318-14) and commentary (ACI 318-14R). American
high temperatures. Fire Safety Journal, 44(8), 1015–1022.
Concrete Institute (ACI): Farmington Hills, MI.
Behnood, A., & Ziari, H. (2008). Effects of silica fume addition
Akhavan, A. C. (2005). The quartz page. Retrieved 07/24,
and water to cement ratio on the properties of high-strength
2016. http://www.quartzpage.de/gen_mod.html.
concrete after exposure to high temperatures. Cement &
Ali, F., Nadjai, A., Silcock, G., & Abu-Tair, A. (2004). Out-
Concrete Composites, 30(2), 106–112.
comes of a major research on fire resistance of concrete
Bentz, D. P. (2007). Transient plane source measurements of the
columns. Fire Safety Journal, 39(6), 433–445.
thermal properties of hydrating cement pastes. Materials
Arioz, O. (2007). Effects of elevated temperatures on properties
and Structures, 40(10), 1073–1080.
of concrete. Fire Safety Journal, 42(8), 516–522.
Bentz, D. P., Peltz, M. A., Duran-Herrera, A., Valdez, P., &
ASTM. (2009). Standard test method for bulk density (‘‘Unit
Juarez, C. A. (2011). Thermal properties of high-volume fly
Weight’’) and voids in aggregate. West Conshohocken, PA:
ash mortars and concretes. Journal of Building Physics,
American Society for Testing of Materials (ASTM). C29/
34(3), 263–275.
C29M-09.
Brown, W. L. (1989). Alkali feldspars: Ordering rates, phase
ASTM. (2011). Standard test method for splitting tensile
transformations and behaviour diagrams for igneous rocks.
strength of cylindrical concrete specimens. West Con-
Mineralogical Magazine, 53(369), 25–42.
shohocken, PA: American Society of Testing Materials
Chen, L.-J., He, J., Chao, J.-Q., & Qin, B.-D. (2009). Swelling and
(ASTM). ASTM C496-11.
breaking characteristics of limestone under high temperatures.
ASTM. (2013). Standard specification for concrete aggregates.
Mining Science and Technology (China), 19(4), 503–507.
West Conshohocken, PA: American Society for Testing of
Cheng, F., Kodur, V., & Wang, T. (2004). Stress–strain curves
Materials (ASTM). C33/C33M-13.
for high strength concrete at elevated temperatures. Journal
ASTM. (2013). Standard test method for surface moisture in
of Materials in Civil Engineering, 16(1), 84–90.
fine aggregate. West Conshohocken, PA: American Society
Cruz, C. R., & Gillen, M. (1980). Thermal expansion of Port-
of Testing Materials (ASTM). ASTM C70-13.
land cement paste, mortar and concrete at high tempera-
ASTM. (2013). Standard test method for total evaporable
tures. Fire and Materials, 4(2), 66–70.
moisture content of aggregate by dryin. West Con-
EN. (1992). Design of concrete structures. Part 1–2: General
shohocken, PA: American Society of Testing Materials
rules—structural fire design. Brussels, Belgium: European
(ASTM). ASTM C566-13.
Standards (EN).
ASTM. (2014). Standard test method for compressive strength
Ghandehari, M., Behnood, A., & Khanzadi, M. (2009). Resid-
of cylindrical concrete specimens. West Conshohocken,
ual mechanical properties of high-strength concretes after
PA: American Society of Testing Materials (ASTM).
exposure to elevated temperatures. Journal of Materials in
ASTM C39/C39M-14.
Civil Engineering, 22(1), 59–64.
ASTM. (2014). Standard test method for resistance to degra-
Gluekler, E. L. (1979). Local thermal and structural behavior of
dation of small-size coarse aggregate by abrasion and
concrete at elevated temperatures. 5th international con-
impact in the Los Angeles machine. West Conshohocken,
ference on structural mechanics in reactor technology
PA: American Society for Testing and Materials (ASTM).
Berlin: International Association.
C131/C131M-14.
Goldsmith, J. R. (1980). The melting and breakdown reactions
ASTM. (2014). Standard test method for sieve analysis of fine
of anorthite at high pressures and temperatures. American
and coarse aggregates. West Conshohocken, PA: American
Mineralogist, 65, 272–284.
Society for Testing of Materials (ASTM). ASTM C136/
Hager, I. (2013). Behaviour of cement concrete at high tem-
C136M-14.
perature. Bulletin of the Polish Academy of Sciences:
ASTM. (2014). Standard test method for static modulus of
Technical Sciences, 61, 145.
elasticity and Poisson’s ratio of concrete in compression.
Heikal, M. (2000). Effect of temperature on the physico-me-
West Conshohocken, PA: American Society of Testing
chanical and mineralogical properties of homra pozzolanic
Materials (ASTM). ASTM C469/C469M-14.
cement pastes. Cement and Concrete Research, 30(11),
ASTM. (2015). Standard test method for density, relative den-
1835–1839.
sity (specific gravity), and absorption of coarse aggregate.
Husem, M. (2006). The effects of high temperature on com-
West Conshohocken, PA: American Society of Testing
pressive and flexural strengths of ordinary and high-per-
Materials (ASTM). ASTM C127-15.
formance concrete. Fire Safety Journal, 41(2), 155–163.

International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017) | 27
Koksal, F., Gencel, O., Brostow, W., & Lobland, H. E. H. normal- and high-strength pozzolanic concretes at elevated
(2012). Effect of high temperature on mechanical and temperatures. Cement and Concrete Research, 31(9),
physical properties of lightweight cement based refractory 1291–1300.
including expanded vermiculite. Materials Research Inno- Rodriguez-Navarro, C., Ruiz-Agudo, E., Luque, A., Rodriguez-
vations, 16(1), 7–13. Navarro, A. B., & Ortega-Huertas, M. (2009). Thermal
Li, H., & Liu, G. (2016). Tensile properties of hybrid fiber- decomposition of calcite: Mechanisms of formation and
reinforced reactive powder concrete after exposure to ele- textural evolution of CaO nanocrystals. American Miner-
vated temperatures. International Journal of Concrete alogist, 94, 578–593.
Structures and Materials, 10(1), 29–37. Sakr, K., & El-Hakim, E. (2005). Effect of high temperature or
Naus, D. (2010). In: D. J. Naus (Ed.), A compilation of elevated fire on heavy weight concrete properties. Cement and
temperature concrete material property data and informa- Concrete Research, 35(3), 590–596.
tion for use in assessments of nuclear power plant rein- Savva, A., Manita, P., & Sideris, K. K. (2005). Influence of ele-
forced concrete structures. U.S. Nuclear Regulatory vated temperatures on the mechanical properties of blended
Commission, Office of Nuclear Regulatory Research. cement concretes prepared with limestone and siliceous
Report no: NUREG/CR-7031, Washington, DC. aggregates. Cement & Concrete Composites, 27(2), 239–248.
Pancar, E. B., & Akpınar, M. V. (2016). Temperature reduction Uygunoğlu, T., & Topçu İ. B. (2012). Effect of aggregate type
of concrete pavement using glass bead materials. Interna- on linear thermal expansion of self-consolidating concrete
tional Journal of Concrete Structures and Materials, 10(1), at elevated temperatures. Science and Engineering of
39–46. Composite Materials, 19(3), 259.
Peng, G.-F., & Huang, Z.-S. (2008). Change in microstructure Wenk, H. R., & Bulakh, A. (2004). Minerals: Their constitution
of hardened cement paste subjected to elevated tempera- and origin. Cambridge: Cambridge University Press.
tures. Construction and Building Materials, 22(4), Yao, W., & Zheng, X. (2007). Effect of mix proportion on
593–599. coefficient of thermal expansion of concrete. Tongji Daxue
Plevova, E., Vaculikova, L., Kozusnikova, A., Ritz, M., & Xuebao/Journal of Tongji University, 35(1), 77–81 ? 87.
Simha Martynkova, G. (2016). Thermal expansion beha- Zhang, W., Qian, H., Sun, Q., & Chen, Y. (2015). Experimental
viour of granites. Journal of Thermal Analysis and study of the effect of high temperature on primary wave
Calorimetry, 123(2), 1555–1561. velocity and microstructure of limestone. Environmental
Poon, C.-S., Azhar, S., Anson, M., & Wong, Y.-L. (2001). Earth Sciences, 74(7), 5739–5748.
Comparison of the strength and durability performance of

28 | International Journal of Concrete Structures and Materials (Vol.11, No.1, March 2017)