Construction and Building Materials 168 (2018) 431–441

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Effects of re-curing on microstructure of concrete after high temperature

exposure
Abdullah Huzeyfe Akca a,b,⇑, Nilüfer Özyurt a
a _
Boğaziçi University, Department of Civil Engineering, 34342 Bebek/Istanbul, Turkey
b _
Yıldız Technical University, Department of Civil Engineering, 34220 Esenler/Istanbul, Turkey

h i g h l i g h t s

High temperature exposure and re-curing cause morphological changes in concrete.

Hydrates disappeared after heating, resulting in a decrease in mechanical properties.

Rehydration of CaO caused further deterioration of concrete after air re-curing.

Regeneration of hydrates resulted in mechanical recovery in water re-cured concrete.

a r t i c l e i n f o a b s t r a c t

Article history: Concrete contains many phases such as ettringite, portlandite, C-S-H etc. which undergo phase changes
Received 10 November 2017 by forming some active products such as lime and calcium silicates during heating. These new active
Received in revised form 9 February 2018 phases give reactions with water and carbon dioxide in air and affect the residual properties of concrete
Accepted 19 February 2018
after subsequent days of cooling. In order to understand better these phase changes and their effects on
Available online 27 February 2018
residual properties, 3 different types of concrete were produced depending on mineral admixture type.
Concrete specimens were heated to 1000 °C and one face heating procedure was applied in the furnace.
Keywords:
Then concrete specimens were subjected to air and water re-curing processes. Microstructural investiga-
High temperature
Fiber reinforced concrete
tions XRD, TGA and SEM/EDX were conducted on the samples obtained from concrete specimens.
Recovery Ó 2018 Elsevier Ltd. All rights reserved.
Deterioration
Microstructure

1. Introduction chemical activity continues at the post-heating stage, causing

Due to excessive temperature increases in concrete hydration
products which determine the strength and durability perfor-
mance of concrete start to alter by losing their water content due
to dehydration and also by releasing carbon dioxide due to decar-
bonation [1–3]. Firstly, ettringite, C-S-H and other hydrates which
have more than one water molecule in their compositions deterio-
rate due to dehydration [4,5]. Then, dehydration of Ca(OH)2 takes
place around 450 °C and decarbonation of carbonates begins after
600 °C [4,5]. After 800 °C most of the constituents in concrete
become decomposed and weight loss of concrete slows down [4].
These new compounds are similar to compounds of unhydrated
cement and they include calcium silicates, calcium aluminates
and calcium oxides etc. which are chemically active [5]. Therefore,
⇑ Corresponding author.
E-mail addresses: abdullah.akca@boun.edu.tr (A.H. Akca), nilufer.ozyurt@boun.
edu.tr (N. Özyurt).
changes in physical and mechanical properties of concrete
[3,6,7]. Water vapor and carbon dioxide in air can react with some
oxides and calcium silicate phases after cooling and resulting prod-
ucts determine the final state of concrete [3,7,8]. For example, dis-
integration and further deterioration of concrete can be observed
due to CaO rehydration. Because of the fact that, Ca(OH)2 is pro-
duced at the end of this rehydration and its volume is 44% higher
than CaO [7–9]. Due to expansive behavior of this rehydration, size
and numbers of existing cracks can increase and mechanical per-
formance of concrete may decrease after cooling [9,10]. In litera-
ture it was reported that using mineral admixtures in concrete
could eliminate or reduce the negative effect of CaO rehydration
since slag consumed Ca(OH)2 after rehydration [8].
On the other hand, when concrete come in contact with water
after cooling, healing and recovery of concrete can be possible in
the case of regeneration of C-S-H and carbonates phases in con-
crete [3,9,11]. However, water re-curing is a debated application
since some of researches inferred that water re-curing after cooling

https://doi.org/10.1016/j.conbuildmat.2018.02.122
0950-0618/Ó 2018 Elsevier Ltd. All rights reserved.
432 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441

could be beneficial for the recovery of concrete and many other specimens were demolded and placed in curing water for 27 days.
researchers found water re-curing application detrimental for con- Then they were conditioned in laboratory environment for addi-
crete [4,5,11–13]. Unfortunately, there are very limited numbers of tional 2 months before tests [14,16,17]. Three cubic specimens
study about water re-curing after heat exposure. Moreover, some were used to determine initial properties of concrete before heat-
of the curing techniques proposed in these researches are not ing. Remaining 12 cubic specimens were heated to 1000 °C. Three
easily applicable on full scale structural members. of them were used to monitor temperatures in concrete cubes dur-
Consequently, a comprehensive research project was designed in ing heating, 3 of them were tested to determine properties after
order to understand disintegration and recovery processes at micro cooling and remaining 6 cubic specimens were subjected to air
level and the effects of microstructural changes on residual mechan- and water re-curing. Therefore, in tables and graphs 0, X, Z and
ical properties of concrete. Hence, concrete groups were subjected W represent tests before heating, after cooling, after air re-curing
to air and water re-curing applications after heating to 1000 °C. In and after water re-curing, respectively.
this project, one face heating conditions were applied and also only
the heated faces of concrete specimens were subjected to water dur- 2.2. Heating procedure
ing water re-curing since these methods can be considered more
realistic and applicable. Ground granulated blast furnace slag An electrical furnace which has a 1250 °C maximum operation
(GGBFS) and pulverized fly ash (PFA) as mineral admixtures were temperature was used in this study. Two stages of heating were
added into concrete mixes in order to evaluate their possible effects determined and during the first stage specimens were heated to
on deterioration and recovery of concrete at the post-heating stage. 1000 °C and during the second stage specimens were kept at this
PP fiber reinforced and air entrained concrete groups were produced temperature until the end of the total heating time (200 min).
in order to reduce the spalling risk of concrete by forming a more Fig. 1 represents the heating procedure and also shows thermocou-
permeable network in concrete during heating [10]. ple positions in a cubic concrete specimen. Electrical furnace was
operated at full power and temperature inside the furnace reached
2. Experimental study 1000 °C in approximately 120 min. Since furnace heating capacity
was limited, heating rate was 20 °C/min at the beginning but the
2.1. Materials and specimens rate decreased to 5 °C/min at the end of the first stage of heating.
After the second stage completed, hot concrete specimens were
CEM I type Portland cement (PC), ground granulated blast fur- not taken out until the furnace cooled down to 100 °C. This heating
nace slag (GGBFS) and F type pulverized fly ash (PFA) were used rates applied in this study can be considered as too high for mate-
in concrete groups as cementitious materials. PC represents Port- rial characterization due to the fact that higher heating rates cause
land cement concrete group, SC represents concrete incorporating thermal stresses which may result in further damage in concrete
GGBFS and FC represents concrete incorporating PFA. Cement during heating [18]. However, the aim of this study was to mimic
replacement ratios of GGBFS and PFA were 40% and 30%, respec- real fire conditions as close as can be done with the available facil-
tively since these ratios were found to be optimum in previous ities. Four concrete cubes were placed together in the electrical fur-
researches [14,15]. Total amount of cementitious materials in 1 nace and in order to simulate one face heating conditions they
m3 of concrete was 450 kg in every concrete group and all concrete were insulated with aerated concrete blocks as shown in Fig. 1.
groups had a water to binder ratio of 0.45. Water contents were For temperature monitoring K-Type thermocouples were placed
determined according to EN 206-1 and TS 13515 since cement inside a concrete specimen during heating.
equivalence factors for GGBFS and PFA are given to be 0.8 and
0.4, respectively. Table 1 shows mix proportions of concrete 2.3. Re-curing processes
groups. As chemical admixtures, oil alcohol and ammonium salt
based air entraining admixture (AEA) and modified polycarboxy- After heating and cooling processes 3 specimens were subjected
late based superplasticizer were used in concrete mixes. AEA con- to air re-curing and 3 specimens were subjected to water re-curing
tent was the same and 0.3 kg (0.7% of total weight of binder) in all for 28 days. Only the heated faces of water re-cured specimens
concrete groups. Super plasticizer amount changed in order to were subjected to tap water. Air re-cured specimens were kept in
obtain slump levels in S4 limits given in EN 206-1. PP fibers were laboratory environment which has a relative humidity of 65 ±
used as low melting point fibers to reinforce concrete groups and 10% and temperature of 20 ± 2 °C during re-curing period.
PP fibers were used 0.2% of volume of concrete. River sand and
siliceous gravel were used in all concrete groups as aggregates. 2.4. Macroscopic observation
15 cubic specimens with dimensions of 15  15  15 cm were
produced for each concrete group. One day after production, Visual changes on heated or fire exposed concrete give informa-
tion about maximum temperature experienced and amount of
Table 1 deterioration of concrete [1,10,19]. Heated surfaces of each con-
Mix proportions. crete group were monitored after re-curing processes.
Materials PC (kg/m3) SC (kg/m3) FC (kg/m3)
2.5. Microscopic observations
Portland Cement 450 270 315
GGBFS – 180 –
PFA – – 135 Microstructural investigations were conducted on samples
Water 203 186 167 which were obtained from a specimen before heating and from
Aggregate No1 484 493 496 specimens after cooling, after air re-curing and after water re-
Aggregate No2 484 493 496
Crushed Sand 564 574 577
curing, in order to determine phase changes such as CaO/Ca(OH)2
River Sand 225 229 230 and CaO/CaCO3 conversions. These samples were taken from the
Super Plasticizer 6.0 7.8 9.6 center of heated surface of concrete specimens and from the cen-
Air Entraining Admixture 0.3 0.3 0.3 troid of concrete specimens by using a percussion drill in order
PP Fiber 1.8 1.8 1.8
not to use cooling water during cutting process (the maximum
Density 2285 2280 2296
length on 1 dimension was 1 cm).
 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441
Fig. 1. (a) Heating curve, (b) Heat insulation of concrete specimens in electrical furnace, (c) positions of the K-type thermocouples used to monitor temperature change.
SEM analyses were conducted on these samples as they were by
using an environmental scanning electron microscope (Philips
XL30 ESEM-FEG/EDAX) to understand the morphological changes
on the heated surfaces and at the centroid of specimens. Secondary
electrons (SE) were measured on platinum coated samples to
obtain pure images and applied accelerating voltage (Acc V) and
working distance (WD) were in the range of 12–20 kV and 9.4–
13.3 mm, respectively. Since water re-cured samples were satu-
rated before SEM, they were dried at 40 °C in an oven for two hours
not to have a vacuum problem. On the other hand, samples were
pounded to obtain powder samples for X-ray diffraction (XRD)
and thermogravimetric analyses (TGA).
2.6. Residual compressive strength and specimen stiffness
measurements
Compression tests were carried out on cubic concrete specimens
before heating, after cooling and after re-curing periods in order to
evaluate changes in mechanical properties of concrete. During tests,
a compression press which has 3000 kN of maximum loading capac-
ity was used and loading rate was set as 0.6 MPa/s. Since heated face
of a specimen became its weakest part after heating, compression
load was not applied on this surface. During loading the plane of
heated face of a specimen was parallel to loading direction as can
be seen in Fig. 1(c). Also, a compressometer frame which includes
two LVDTs was used on cubic concrete specimens to be able to mea-
sure length changes and to determine specimen stiffness. The com-
pressometer frame was removed after 40% of ultimate strength of
concrete specimens was reached and tests continued until failure
occurred to determine compressive strength values.
3. Results
3.1. Temperature monitoring
Monitoring internal temperature of concrete is important since
measured temperatures give information about the morphological
Table 2
Maximum temperatures at predefined depths from the surface, (after 200 min of heating and 150 min of cooling).
Thermocouple No TC#1 TC#2
Depth (mm) 0 19
PC (°C) 1000 770
SC (°C) 1000 801
FC (°C) 1000 791
433
changes which can be beneficial during SEM investigations. There-
fore, temperatures of heated surface and inner parts of concrete
cubes were monitored during heating and temperature monitoring
continued further during cooling of the furnace (150 min more). In
cooling period, although heating of furnace was stopped tempera-
tures in concrete continued to increase. Maximum monitored tem-
peratures in concrete groups were given in Table 2. When the
whole heating and cooling period is considered it can be seen that
core level of concrete groups experienced at least 590 °C of tem-
perature which may result in severe strength losses [10,14,20].
Moreover, temperatures measured at deeper parts of PFA concrete
were higher than other concrete types.
3.2. Heated surfaces after air and water re-curing
Heated surfaces of concrete specimens were monitored at the
end of re-curing periods as can be seen in Fig. 2. Water re-cured
specimens kept their stabilities and no crack growth was observed
in these specimens. On the other hand, severe cracks occurred on
the heated surfaces of concrete specimens after air re-curing per-
iod due to morphological degradation. These morphological
changes will be evaluated in detail in upcoming sections.
3.3. Results of microstructural analyses
Changes in microstructure of concrete due to thermal applica-
tions determine the behavior of macroscopic material under speci-
fic conditions [21,22]. Bonds between molecules can be broken
with a temperature increase and breaking of most of the bonds
occurs after 750 °C [3,9]. Therefore, residual mechanical properties
of concrete drastically decrease after 750 °C [10,14,20]. In order to
understand these morphological changes SEM/EDX, XRD and TGA
were carried out on samples obtained from concrete specimens.
3.3.1. SEM/EDX investigations
3.3.1.1. Changes in concrete morphology at the surface level after
heating. During SEM investigations, C-S-H gels were traced via EDX
TC#3 TC#4 TC#5 TC#6
38 57 75 113
659 614 590 564
647 619 600 570
675 648 619 607
434 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441

Fig. 2. Heated surfaces of concrete specimens after re-curing processes.

analyses since the molar ratio of Ca/Si gives information about the
existence and types of C-S-H [21,23]. In literature it is reported that
if C-S-H gel has a loose fibril-like morphology Ca/Si ratios around
0.6 can be measured. On the other hand, if C-S-H loses its exten-
sions by forming dense granular-like morphology, measured Ca/
Si values reported to be around 2.0 [23]. Evaluated SEM pictures
of samples obtained from heated surfaces of Portland cement con-
crete, concrete incorporating GGBFS and PFA before heating and
after cooling were given in Figs. 3–5, respectively. (It should be
noticed that in some SEM images magnifications were seen as
20000X due to an update of SEM software. Actually all the pictures
were captured at the same magnification level of 50000X since the
scale bar length are the same and set as 1 mm.)
When the specimens produced for this study were examined,
initially, C-S-H gels were found to have sharp extensions in PC con-
crete and Ca/Si ratios were around 1.3. On the other hand, in min-
eral added concrete groups, C-S-H gels were like dense granular
particles and the molar ratios of Ca/Si of concrete incorporating
GGBFS and concrete incorporating PFA were around 1.7 and 1.6,
respectively. After heating, C-S-H gels shrank and densified and
represented a granular form as can be seen in Figs. 3–5b [21]. Lin
and colleagues report that C-S-H turns to C2S (belite) and C3S
(alite) by losing its water molecules at 600–700 °C [3]. Also Ca/Si
ratio has a tendency to increase after heating as Taylor stated in
his book [24]. In this study, Ca/Si ratios increased and measured
around 1.9, 1.9 and 1.7 in PC concrete, concrete with GGBFS and
concrete with PFA, respectively.
3.3.1.2. Changes in concrete morphology at the surface level after re-
curing. SEM investigations were conducted on samples obtained
from heated surfaces of air re-cured and water re-cured concretes
as given in Figs. 6–8. As can be seen in SEM pictures of air re-cured
samples, after air re-curing period, forms of dehydrated C-S-H par-
ticles remained almost same for all concrete groups. On the other
hand, prominent changes were observed in microstructures of
water re-cured concrete. Mainly, granular formations seen after
heating disappeared and new formations which have sharp exten-
sions were observed in all concrete samples. These extensions
probably occurred due to rehydration and carbonation and may
play a role in recovery of concrete after water re-curing. Because
of the fact that, during SEM observations cracks filled with new

Fig. 3. SEM pictures obtained from PC (a) before heating (b) after heating and cooling.
 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441 435

Fig. 4. SEM pictures obtained from SC (a) before heating (b) after heating and cooling.

Fig. 5. SEM pictures obtained from FC (a) before heating (b) after heating and cooling.

Fig. 6. SEM pictures obtained from PC after (a) air re-curing (b) water re-curing.

products, especially ettringite and calcium carbonate, were moni-
tored as can be seen in Fig. 9.
3.3.1.3. Changes in concrete morphology at the center after heating
and re-curing. SEM pictures obtained from the center of concretes
after cooling can be seen in Fig. 10. When monitored temperatures
at the center (75 mm depth) of PC, SC and FC as given in Table 2
were examined it can be understood that temperature at the center
of concrete were around 600 °C during heating. Therefore, C-S-H
gels lost some amount of its water content and shrank similar to
surface level investigations. Measured molar ratios of Ca/Si values
were around 1.5, 1.7 and 1.7 for PC, SC and FC, respectively. The
increases in molar ratios of Ca/Si values measured at the center
part were relatively lower than increases in Ca/Si values measured
436 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441

Fig. 7. SEM pictures obtained from SC after (a) air re-curing (b) water re-curing.

Fig. 8. SEM pictures obtained from FC after (a) air re-curing (b) water re-curing.

Fig. 9. Cracks filled with new forms after water re-curing.

at the surface level. This also indicated that center part of the spec-
imens experienced lower temperatures and they were less deteri-
orated than surface level after heating.
After air re-curing period microstructure of concrete did not
change prominently as can be seen in Fig. 11. During SEM investi-
gations almost no Ca(OH)2 was observed on the samples obtained
from the center of the air re-cured concrete specimens. On the
other hand, new formations were observed on water re-cured sam-
ples. Newly formed C-S-H and Ca(OH)2 crystals can be seen in
Fig. 12. This means that, although only the heated faces were sub-
jected to water re-curing, water molecules were able to diffuse in
concrete via capillary action. At the surface level Ca(OH)2 crystals
 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441 437

Fig. 10. Microstructure of the center of concretes after cooling (a) PC, (b) SC, (c) FC.

Fig. 11. Microstructure of the center of air re-cured concretes (a) PC, (b) SC, (c) FC.
438 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441

Fig. 12. Microstructure of the center of water re-cured concretes (a) PC, (b) SC, (c) FC.

were not common in water re-cured concrete since most of them
dissolved in curing water.
3.3.2. XRD analyses
XRD results give information about the presence of phases and
compounds in sample. XRD results of samples obtained from PC
group can be seen in Fig. 13. According to results, C-S-H and Ca
(OH)2 (CH) disappeared after heating and calcium silicate (CS) for-
mations were observed on the heated surface. After air re-curing
period, similar to TGA results CaCO3 (calcite) and Ca(OH)2 (CH) for-
mations were observed on the surface. On the other hand, CaCO3
(calcite) and C-S-H peaks can be seen on XRD diagram of sample
obtained from water re-cured specimen. Also the intensity of cal-
cium silicate peak decreased after water re-curing since a part of
them formed C-S-H after rehydration. Ca(OH)2 (CH) probably dis-
solved in curing water and its negative effect on heated surface
due to expansion after rehydration was eliminated.
3.3.3. Thermogravimetric analyses
Concrete pieces were taken from the heated surfaces of speci-
mens and they were pounded in a steel bowl in order to obtain
powder samples. Thermogravimetric analyses were conducted on
15 mg of powder samples. Heating rate was 20 °C/min and samples
were heated to 800 °C in Nitrogen atmosphere with a purge rate of
60 mL/min. After thermogravimetric measurements differential
thermal analysis (DTA) diagrams were drawn as can be seen in

Fig. 13. XRD results of samples obtained from PC group.

 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441 439

Fig. 14. DTA diagrams of concrete groups.

Fig. 14. Peaks in DTA diagrams give information about the phase
changes and transformations. DTA peaks of a concrete sample
around 450 °C represent Ca(OH)2 dehydration and around 700 °C
represent CaCO3 decarbonation [25–27].
After cooling DTA diagrams showed that there were no Ca(OH)2
and CaCO3 in samples obtained from heated surfaces because of
the fact that most of the crystals and gels deteriorated during heat-
ing. Since maximum temperature at the center of concrete speci-
mens were around 600 °C, at this part of concrete CaCO3 did not
totally decarbonated (DTA diagrams of after cooling samples
obtained from center region of specimens). Most of Ca(OH)2 crys-
tals dehydrated at the center part of concrete specimens. After
air re-curing period both Ca(OH)2 and CaCO3 peaks reappeared.
The presence of Ca(OH)2 on the surface of the heated specimen is
hypothesized to be the reason of newly formed cracks due to its
expansion history during air re-curing.
On the other hand, there were no Ca(OH)2 peaks on DTA dia-
grams of water re-cured concrete samples obtained from surface
region of specimens. Since CaOH2 is a soluble crystal in water [28],
newly formed CaOH2 crystals can be dissolved in curing water.
Absence of CaOH2 on the heated surfaces of concrete can result in
a slowdown of new crack formation. After water re-curing period
specimens kept their stabilities and the numbers and sizes of exist-
ing cracks did not change. There were CaOH2 peaks on DTA diagrams
of samples obtained from the center region of water re-cured con-
crete specimens as was observed during SEM investigations.
Ca(OH)2 and CaCO3 contents can be calculated by using mass
loss data obtained from TGAs. Measured mass loss around Ca
(OH)2 peak represents evaporated water and was shown as DmCH
in Eq. (1) and measured mass loss around CaCO3 peak represents
released carbon dioxide was shown as DmCC in Eq. (2). Based on
this and by using Eqs. (3) and (4) the ratios between measured
mass loses around Ca(OH)2 and CaCO3 peaks and actual amounts
of Ca(OH)2 and CaCO3 can be calculated.
Table 3 shows calculated Ca(OH)2 and CaCO3 contents of con-
crete samples. According to the results, there were initially Ca
(OH)2 and CaCO3 in all non-heated samples. Moreover, amount of
Ca(OH)2 measured in mineral added concrete samples is approxi-
mately 50% less than the measured amount of Ca(OH)2 in Portland
cement concrete. After cooling, negligable amounts of Ca(OH)2 and
CaCO3 were found in all types of concrete samples at the surface
level. After air re-curing period, 2.5% of Ca(OH)2 was found in sam-
ple obtained from Portland cement concrete surface and 2% of Ca
(OH)2 were found in samples obtained from mineral added con-
crete surfaces. CaCO3 ratio reached 7.5% at the surface level of Port-
land cement concrete while less amount of carbonation was
observed in mineral added concrete groups. After water re-
curing, negligible amounts of CaOH2 were calculated in all kind
of concrete samples. When the center parts of concrete were eval-
uated, it can be seen that CaOH2 increased to 6.1% in Portland
cement concrete while they were 3.3% and 1.4% in concrete incor-
porating GGBFS and PFA, respectively. Less amount of CaOH2 in

Table 3
Calculated Ca(OH)2 and CaCO3 percentages in concrete samples.

Before heating Sample from After cooling Air re-cured Water re-cured
(0) (X) (Z) (W)
Ca(OH)2 CaCO3 Ca(OH)2 CaCO3 Ca(OH)2 CaCO3 Ca(OH)2 CaCO3
(%) (%) (%) (%) (%) (%) (%) (%)
PC 8.3 5.8 Surface 0.5 0.3 2.5 7.5 0.1 7.0
Center 0.4 2.7 0.2 6.4 6.1 8.0
SC 4.3 6.4 Surface 0.1 1.6 2.0 2.7 0.4 3.1
Center 1.2 6.2 0.9 5.7 3.3 7.6
FC 4.5 7.1 Surface 0.3 0.8 1.9 1.1 0.1 2.9
Center 0.2 6.0 0.6 6.7 1.4 6.2
440 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441
Table 4
Average residual compressive strength and specimen stiffness percentages.
Group Residual compressive strength (%)
X Z
PC 50.5 41.3
SC 56.4 42.8
FC 56.0 42.8
mineral added concrete groups can be the result of pozzolonic
activity.
a ð1Þ
a ð2Þ
mCH ¼ DmCH  74=18 ð3Þ
mCC ¼ DmCC  100=44 ð4Þ
3.4. Results of compressive strength and specimen stiffness
measurements
Since only one face of the concrete specimens was directly sub-
jected to high temperatures a heterogeneous degradation from the
heated surface to opposite one was obtained. Therefore, the reduc-
tion in compressive strength and specimen stiffness measured in
the study is the result of this degradation and is purely qualitative
in nature. Consequently, it should be remembered that reduction
percentages measured in this study are only applicable to make a
comparison when similar heating conditions are applied on speci-
mens which have similar dimensions.
Residual compressive strength and specimen stiffness of con-
crete groups were given in Table 4. As a summary, results of the
mechanical part of the study showed that compressive strength
and specimen stiffness of concrete decreased 46% and 79% after
heating, respectively. Deterioration in mechanical properties con-
tinued in air re-cured specimens and compressive strength and
specimen stiffness of air re-cured concrete decreased 58% and
89% after heating, respectively. On the other hand, regaining in
strength and recovery in specimen stiffness of concrete were
observed in water re-cured specimens. According to compression
test results, reduction in compressive strength and specimen stiff-
ness of water re-cured concrete were 31% and 42%, respectively.
4. Conclusions
In this study, disintegration and recovery of various kinds of
concrete under air and water re-curing regimes were investigated.
SEM/EDX, XRD and TGAs were conducted on concrete samples in
order to understand the morphological changes which may give
rise to either disintegration or recovery of concrete. The conclu-
sions drawn were given below.
Mechanical properties of concrete decreased drastically after
heating because of the fact that initial morphology of concrete
prominently changed after heating. XRD and TGA results showed
that most of the crystals and gels disappeared after heating and
according to SEM investigations overall microstructure evolved
to a granular form. Molar ratio of Ca/Si of C-S-H changed and den-
sification of C-S-H was observed after heating. Increased free space
in microstructure and cracks occurred during heating resulted in
Residual specimen stiffness (%)
W X Z W
67.0 19.4 11.6 59.3
75.4 25.0 10.5 57.4
65.7 18.8 9.8 57.6
higher reduction in specimen stiffness than reduction in compres-
sive strength.
While overall morphology of concrete remained same at the
micro level, sizes and numbers of cracks increased at macro level
during air re-curing period. According to TGA results on samples
obtained from the heated surfaces of concrete, peaks representing
Ca(OH)2 and Ca(CO)3 were observed in air re-cured samples. There-
fore, the reason of disintegration of air re-cured concrete was
hypothesized to be CaO expansion. Since there was no apparent
change in microstructure, the new cracks occurred during this per-
iod can be considered as the reason of further decrease in residual
mechanical properties after air re-curing period.
The main effect of water re-curing was to prevent heated sur-
face of concrete from cracking during curing period by eliminating
the negative effects of CaO expansion because of the fact that Ca
(OH)2 could dissolve in curing water. As it was expected Ca(OH)2
was not seen in water re-cured samples obtained from heated sur-
faces and also no crack growth was observed on these surfaces. On
the other hand, morphology of concrete changed significantly after
water re-curing, especially rehydration of C-S-H and CaCO3
improved microstructure of concrete resulting in increase in resid-
ual mechanical properties. However, recovery in specimen stiffness
was higher than compressive strength owing to the cracks filled
with new rehydration and carbonation products and decreased
free space in microstructure of concrete.
According to microstructural investigations conducted on sam-
ples obtained from center of concrete specimens, Ca(OH)2 were
observed both in air re-cured and water re-cured concrete. How-
ever, pre-mentioned negative effect of CaO expansion was found
to be limited at this zone (cracks observed on surface didn’t extend
to center). This was probably because; comparatively lower tem-
peratures experienced in this zone (based on thermocouple read-
ings) leaving center part relatively stronger when compared to
the surface.
Acknowledgements
The authors gratefully acknowledge the financial support of Bog
aziçi University Research Fund [Project Code 14A04D2]. The sup-
˘
port of AKÇANSA Cement and BASF-YKS Construction Chemicals
is also acknowledged. The authors also would like to thank Ümit
Melep, Bilge Uluocak and Melike Babucci for their support during
experimental measurements. The first author is grateful for the
financial support given by The Scientific and Technical Research
_
Council of Turkey (TÜBITAK).
References
[1] J.P. Ingham, Application of petrographic examination techniques to the
assessment of fire-damaged concrete and masonry structures, Mater.
Charact. 60 (2009) 700–709.
[2] G. Peng, Z. Huang, Change in microstructure of hardened cement paste
subjected to elevated temperatures, Constr. Build. Mater. 22 (2008) 593–599.
[3] W.M. Lin, T.D. Lin, L.J. Powers-Couche, Microstructures of fire-damaged
concrete, ACI Mater. J. 93 (1996) 199–205.
[4] D.N. Crook, M.J. Murray, Regain of strength and firing of concrete, Mag. Concr.
Res. 22 (6) (1970) 149–154.
 A.H. Akca, N. Özyurt / Construction and Building Materials 168 (2018) 431–441
[5] M. Henry, K. Hashimoto, I.S. Darma, T. Sugiyama, Cracking and chemical
composition of cement paste subjected to heating and water re-curing, J. Adv.
Concr. Technol. 14 (2016) 134–143.
[6] H. Seleem, A.M. Rashad, T. Elsokary, Effect of elevated temperature on physico-
mechanical properties of blended cement concrete, Constr. Build. Mater. 25
(2011) 1009–1017.
[7] A. Mendes, J.G. Sanjayan, F. Collins, Long-term progressive deterioration
following fire exposure of OPC versus slag blended cement pastes, Mater.
Struct. 42 (2009) 95–101.
[8] A. Mendes, J.G. Sanjayan, F. Collins, Effects of slag and cooling method on the
progressive deterioration of concrete after exposure to elevated temperatures
as in a fire event, Mater. Struct. 44 (2011) 709–718.
[9] C. Alonso, L. Fernandez, Dehydration and rehydration processes of cement
paste exposed to high temperature environments, J. Mater. Sci. 39 (2004)
3015–3024.
[10] A.H. Akca, N. Özyurt Zihnioğlu, High performance concrete under elevated
temperatures, Constr. Build. Mater. 44 (2013) 317–328.
[11] C.H. Poon, S. Azhar, M. Anson, Y. Wong, Strength and durability recovery of
fire-damaged concrete after post-fire-curing, Cem. Concr. Res. 31 (9) (2001)
1307–1318.
[12] T. Horiguchi, S.L. Suhaendi, Recovery behavior of hybrid fiber reinforced
high strength concrete after fire exposure, J. Struct. Fire Eng. 1 (2010) 219–
229.
[13] R. Sarshar, G.A. Khoury, Material and environmental factors influencing the
compressive strength of unsealed cement Paste and Concrete at High
Temperatures, Mag. Concr. Res. 45 (1993) 51–61.
[14] C.S. Poon, S. Azhar, M. Anson, Y.L. Wong, Comparison of the strength and
durability performance of normal- and high-strength pozzolanic concretes at
elevated temperatures, Cem. Concr. Res. 31 (2001) 1291–1300.
[15] H.S. Lee, X.Y. Wang, L.N. Zhang, K.T. Koh, Analysis of the optimum usage of slag
for the compressive strength of concrete, Materials 8 (2015) 1213–1229.
441
[16] Y.N. Chan, G.F. Peng, M. Anson, Residual strength and pore structure of high-
strength concrete and normal strength concrete after exposure to high
temperatures, Cem. Concr. Compos. 21 (1999) 23–27.
[17] Y.F. Chang, Y.H. Chen, M.S. Sheu, G.C. Yao, Residual stress-strain relationship
for concrete after exposure to high temperatures, Cem. Concr. Res. 36 (2006)
1999–2005.
[18] T.C. Rilem, 129-MHT committee, test methods for mechanical properties of
concrete at high temperatures – Part 3: compressive strength, Mater. Struct.
28 (1995) 410–414.
[19] N. Yüzer, F. Aköz, L. Dokuzer Öztürk, Compressive strength–color change
relation in mortars at high temperature, Cem. Concr. Res. 34 (2004) 1803–1807.
[20] G.A. Khoury, Effect of fire on concrete and concrete structures, Struct. Eng.
Mater. 22 (2000) 429–447.
[21] S. Djaknoun, E. Quedraogo, A.A. Benyahia, Characterization of the behavior of
high performance mortar subjected to high temperatures, Constr. Build. Mater.
28 (2012) 176–186.
[22] A.H. Akca, N. Özyurt Zihnioğlu, Mechanical behavior and recovery of FRC after
high temperature exposure, Key Eng. Mater. 711 (2016) 457–464.
[23] D. Hou, Z. Li, H. Ma, Morphology of calcium silicate hydrate (C-S-H) gel: a
molecular dynamic study, Adv. Cem. Res. 27 (2015) 135–146.
[24] H.F.W. Taylor, Cement Chemistry, Academic Press Limited, London, 1992.
[25] E. Menendez, L. Vega, Analysis of the behaviour of the structural concrete after
the fire at the Windsor Building in Madrid, Fire Mater. 34 (2010) 95–107.
[26] F. Branda, G. Luciani, A. Costantini, C. Piccioli, Interpretation of the
thermogravimetric curves of ancient pozzolanic concretes, Archaeometry 43
(2001) 447–453.
[27] G. Villain, M. Thiery, G. Platret, Measurement methods of carbonation profiles
in concrete: thermogravimetry, chemical analysis and gammadensimetry,
Cem. Concr. Res. 37 (2007) 1182–1192.
[28] E.J. Reardon, R. Fagan, The calcite/portlandite phase boundary: enhanced
calcite solubility at high pH, Appl. Geochem. 15 (2000) 327–335.