CHARACTERIZING TEMPERATURE INDUCED STRENGTH DEGRADATION AND
EXPLOSIVE SPALLING IN ULTRA HIGH PERFORMANCE CONCRETE
By
Mahmood A. Sarwar

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Civil Engineering―Master of Science
2017

ABSTRACT
CHARACTERIZING TEMPERATURE INDUCED STRENGTH DEGRADATION AND
EXPLOSIVE SPALLING IN ULTRA HIGH PERFORMANCE CONCRETE
By
Mahmood A. Sarwar
Significant research and development in concrete technology have led to the evolution of
ultra-high-performance concrete (UHPC). UHPC possesses excellent durability and ductility
characteristics, and exemplary tensile and compressive strength properties. Therefore, owing to its
superlative properties UHPC is being progressively used in infrastructure and must satisfy fire
resistance requirements. Yet, evaluating fire resistance requires knowledge of elevated
temperature mechanical properties, and unfortunately such properties are lacking. Furthermore,
exploratory investigations reveal UHPC’s composition and construct precipitate the temperature
engendered dilapidation of its strength, and it is highly susceptible to fire induced spalling.
To fulfill this integral gap, a comprehensive experimental study was undertaken to develop
temperature induced degradation in compressive strength, along with the fire induced spalling
behavior of UHPC. Variables were introduced to the testing program through varying the content
and type of fibers, and varying the application of heating rate. Furthermore, UHPC’s results were
compared to four conventional concretes tested, and recommendations as prescribed in ASCE
(1992) and Eurocode 2 (2004). Results from the experimental testing reveal the compressive
strength and stress-strain behavior of UHPC deviates from conventional concretes, and there is a
vast disconnect amongst the experimental results and recommendations prescribed in codes.
Moreover, successful testing of UHPC at and above 300°C, was only possible through drastically
reducing the heating rate to less than 0.75°C/min. Otherwise, employing higher heating rates
induces explosive spalling at temperatures as low as 190°C.

Copyright by
MAHMOOD A. SARWAR
2017

To my Beevee, our three babies, Baby Girl, Iman and Afreen, and to all the children who have
suffered through war and tragedy.

iv

ACKNOWLEDGEMENTS

Sincere thanks go in particular to my supervisor Prof. Venkatesh Kodur for his valuable
guidance and constructive suggestions he provided in undertaking this research and in preparation
of this thesis. Also, special thanks to the distinguished faculty members who served on my
committee: Professors: Parviz Soroushian and Weiyi Lu. Thank you to all my committee members
for their support, patience, encouragement, useful suggestions and guidance.
Most importantly, I thank God who brought my wife and I together. Her continuous love,
care and patience is what keeps me on my feet. Also, I would like to thank the whole of our research
group, especially Mohannad Naser, Ankit Agrawal, Xu Dong, Saleh Alogla, Puneet Kumar, Roya
Solhmirzaei and Esam Aziz.
Thanks to the faculty and staff in the CEE department, Veterans Certification Team and
Student Veteran Resource Center for their support and help, especially the manager of the Civil
Infrastructure Laboratory Mr. Siavosh “Sia” Ravanbakhsh for his mentorship and everlasting
patience. Without his guidance, I would not have been able to complete my experimental testing.
Additionally, I would like to extend my thanks to Mrs. Laura Taylor, Mrs. Margaret Conner and
Mrs. Laura Post for all their help.
Lastly, thank you to all my family for their love and continuing prayers.

v

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................................... ix
LIST OF FIGURES ..................................................................................................................... x
CHAPTER ONE ......................................................................................................................... 1
1. INTRODUCTION ................................................................................................................... 1
1.1 General............................................................................................................................... 1
1.2 Fire resistance properties of concrete ................................................................................ 4
1.3 Mechanical properties of concrete at elevated temperature .............................................. 5
1.4 Objectives of this research ................................................................................................. 6
CHAPTER TWO......................................................................................................................... 9
2. STATE OF THE ART REVIEW ............................................................................................ 9
2.1 General............................................................................................................................... 9
2.2 Thermal properties of concrete ........................................................................................ 10
2.2.1 General .................................................................................................................. 10
2.2.2 Test methods for high temperature thermal properties .......................................... 10
2.2.3 Previous studies on thermal properties .................................................................. 12
2.2.4 Effect of temperature on thermal properties of conventional concretes................ 12
2.2.5 Effect of temperature on thermal properties of UHPC .......................................... 13
2.3 Mechanical properties of concrete ................................................................................... 13
2.3.1 General .................................................................................................................. 13
2.3.2 Testing methods for high temperature mechanical properties .............................. 15
2.3.3 Previous studies on mechanical properties ............................................................ 16
2.3.4 Compressive strength ............................................................................................ 17
2.3.4.1 Effect of temperature on compressive strength of conventional concretes . 17
2.3.4.2 Effect of temperature on compressive strength of UHPC ........................... 18
2.3.5 Tensile strength ..................................................................................................... 19
2.3.5.1 Effect of temperature on tensile strength of conventional concretes .......... 19
2.3.5.2 Effect of temperature on tensile strength of UHPC .................................... 20
2.3.6 Stress-strain relationship and elastic modulus ....................................................... 21
2.3.6.1 Effect of temperature on stress-strain relationship and elastic modulus of
conventional concretes ............................................................................................ 21
2.3.6.2 Effect of temperature on stress-strain relationship and elastic modulus of
UHPC ...................................................................................................................... 22
2.4 Fire induced spalling ....................................................................................................... 23
2.4.1 General .................................................................................................................. 23
2.4.2 Causes of spalling .................................................................................................. 23
2.4.2.1 Rate of heat and thermal stresses induced by thermal gradient .................. 24
2.4.2.2 Pore pressure build-up ................................................................................. 24
2.4.3 Previous studies on fire induced spalling .............................................................. 25
2.4.4 Methods for minimizing explosive spalling .......................................................... 26
vi

2.5 Research gaps .................................................................................................................. 28
CHAPTER THREE ................................................................................................................... 54
3. EXPERIMENTAL STUDIES ON COMPRESSIVE STRENGTH AND EXPLOSIVE
SPALLING OF UHPC AT ELEVATED TEMPERATURES ................................................. 54
3.1 General............................................................................................................................. 54
3.2 Design of mechanical property experiments ................................................................... 54
3.3 Mechanical properties tests ............................................................................................. 55
3.3.1 Batch mix proportions ........................................................................................... 55
3.3.2 Test specimens....................................................................................................... 56
3.3.3 Test apparatus ........................................................................................................ 58
3.3.4 Test procedure ....................................................................................................... 58
3.3.4.1 General procedure ....................................................................................... 58
3.3.4.2 Compressive strength, stress-strain relationship, and elastic modulus ....... 60
3.4 Results and discussion ..................................................................................................... 61
3.4.1 Sectional temperatures........................................................................................... 61
3.4.2 Compressive strength, elastic modulus, and stress-strain relationship.................. 63
3.4.2.1 NSC ............................................................................................................. 64
3.4.2.2 HSC ............................................................................................................. 65
3.4.2.3 UHPC .......................................................................................................... 67
3.4.3 Fire-induced spalling behavior .............................................................................. 70
3.4.3.1 Spalling behavior in experiments ................................................................ 70
3.4.3.2 Temperature gradient and rate of heat analysis ........................................... 74
3.5 Summary........................................................................................................................ 110
CHAPTER FOUR ................................................................................................................... 111
4. COMPARISON OF PROPERTIES OF UHPC WITH CONVENTIONAL CONCRETE
AND BUILDING CODES ...................................................................................................... 111
4.1 General........................................................................................................................... 111
4.2 Comparison of room temperature properties ................................................................. 111
4.3 Comparison of elevated temperature properties ............................................................ 112
4.3.1 Comparison of UHPC with conventional concretes tested ................................. 112
4.3.2 Comparison of UHPC with building codes ......................................................... 114
4.4 Comparison of spalling.................................................................................................. 114
4.4.1 Comparison of UHPC with conventional concretes tested ................................. 114
4.4.2 Comparison of UHPC with building codes ......................................................... 114
4.5 Summary........................................................................................................................ 123
CHAPTER FIVE ..................................................................................................................... 124
5. CONCLUSIONS AND RECOMMENDATIONS.............................................................. 124
5.1 General........................................................................................................................... 124
5.2 Key findings .................................................................................................................. 124
5.3 Recommendations for future research ........................................................................... 125
5.4 Research impact ............................................................................................................. 126

vii

APPENDIX ............................................................................................................................. 127
REFERENCES ........................................................................................................................ 131

viii

LIST OF TABLES

Table 2.1: Temperature rise and effect on concrete, adapted from Klingsch (2014) ................... 29
Table 2.2: Survey of UHPC/RPC and UHPFRC studies conducted at ambient conditions ......... 30
Table 2.3: Survey of UHPC/RPC and UHPFRC studies conducted at elevated temperatures..... 33
Table 3.1: Testing regime conducted as part of this study ........................................................... 78
Table 3.2: Mix proportions for UHPC design mixes tested as part of study ................................ 78
Table 3.3: Mix proportions for conventional concretes tested as part of study ............................ 79
Table 3.4: Density and 28-day compressive strengths of concrete cylinders tested as part of study
....................................................................................................................................................... 79
Table 3.5: Details of compressive strength tests of UHPC A & B cylinders ............................... 80
Table 3.6: Details of compressive strength tests of UHPC D cylinders ....................................... 81
Table 3.7: Details of compressive strength tests of UHPC F cylinders ........................................ 82
Table 3.8: Details of compressive strength tests of NSC-SF and NSC cylinders......................... 83
Table 3.9: Details of compressive strength tests of HSC-PP and HSC cylinders......................... 84
Table 3.10: Details pertaining to UHPC tests conducted at, and above 300°C ............................ 85
Table 3.11: Key points from thermal gradient with respect to time graphs for UHPC tests
conducted at, and above 300°C ..................................................................................................... 86
Table 3.12: Color legend for thermal gradient with respect to time graphs ................................. 87
Table A1: Additional notes pertaining to UHPC A & B specimen tests .................................... 128
Table A2: Additional notes pertaining to UHPC D specimen tests ............................................ 129
Table A3: Additional notes pertaining to UHPC F specimen tests ............................................ 130

ix

LIST OF FIGURES

Figure 1.1: Photos depicting explosive spalling in a NSC and HSC column under elevated
temperatures (Khaliq, 2012) ........................................................................................................... 8
Figure 2.1: Relative compressive strength versus temperature as per literature from the past 3
years .............................................................................................................................................. 51
Figure 2.2 Relative tensile strength versus temperature as per literature ..................................... 51
Figure 2.3 Relative elastic modulus versus temperature as per literature from the past 2 years .. 53
Figure 3.1: Prepared concrete specimens...................................................................................... 88
Figure 3.2: Table saw utilized to cut and prepare the concrete cylinders ..................................... 89
Figure 3.3: Electrical furnace utilized in heating concrete cylinders............................................ 90
Figure 3.4: Forney load controlled strength testing machine to conduct compressive strength
testing as part of this study............................................................................................................ 90
Figure 3.5: Compressometer with (2) LVDTs and (2) metal rods utilized to develop the stressstrain relationship as part of this study ......................................................................................... 91
Figure 3.6: Desired heating rates for high temperature compression tests ................................... 91
Figure 3.7: Temperature data acquisition device utilized as part of this study ............................. 93
Figure 3.8: Temperature versus time development for concrete cylinders tested at 600°C ......... 94
Figure 3.9: Compressive strength with respect to temperature of various concrete cylinders tested
as part of research ......................................................................................................................... 95
Figure 3.10: Elastic modulus with respect to temperature of various concrete cylinders tested as
part of research.............................................................................................................................. 96
Figure 3.11: NSC stress-strain relationships with respect to temperature .................................... 97
Figure 3.12: Failure modes of NSC at respective temperatures ................................................... 97
Figure 3.13: NSC-SF stress-strain relationships with respect to temperature .............................. 98
Figure 3.14: Failure modes of NSC–SF at respective temperatures ............................................. 98

x

Figure 3.15: HSC stress-strain relationships with respect to temperature .................................... 99
Figure 3.16: Failure modes of HSC at respective temperatures ................................................... 99
Figure 3.17: HSC-PP stress-strain relationships with respect to temperature ............................ 100
Figure 3.18: Failure modes of HSC – PP at respective temperatures ......................................... 101
Figure 3.19: UHPC A & B stress-strain relationships with respect to temperature ................... 102
Figure 3.20: Failure modes of UHPC A & B at respective temperatures ................................... 102
Figure 3.21: UHPC D stress-strain relationships with respect to temperature ........................... 103
Figure 3.22: Failure modes of UHPC D at respective temperatures........................................... 103
Figure 3.23: UHPC F stress-strain relationships with respect to temperature ............................ 104
Figure 3.24: Failure modes of UHPC F at respective temperatures ........................................... 104
Figure 3.25: Damage to furnace due to explosive spalling......................................................... 105
Figure 3.26: UHPC specimens following experiencing explosive spalling in electrical furnace106
Figure 3.27: Specimen UHPC B5 displaying water escaping concrete following 300°C test ... 107
Figure 3.28: Relation between rate of heating with respect to either, (1) the spalling temperature,
or (2) the maximum center temperature attained ........................................................................ 108
Figure 3.29: Thermal gradient with respect to time for UHPC A & B cylinders ....................... 108
Figure 3.30: Thermal gradient with respect to time for UHPC D cylinders ............................... 109
Figure 3.31: Thermal gradient with respect to time for UHPC F cylinders ............................... 109
Figure 4.1: Relative compressive strength with respect to temperature of various concrete
cylinders tested as part of research, and Eurocode 2 (2004) and ASCE (1992) ......................... 115
Figure 4.2: Relative elastic modulus with respect to temperature of various concrete cylinders
tested as part of research, and Eurocode 2 (2004) and ASCE (1992)......................................... 116
Figure 4.3: Relative stress-strain relationships for concrete cylinders tested ............................. 117

xi

CHAPTER ONE
1. INTRODUCTION
1.1 General
The use and implementation of concrete for construction can be traced as far back as over
3500 years (Hertz, 2003). Despite the extraordinary advancements in technology since then, the
compulsion for builders to utilize concrete continues to endure the test of time. On both a domestic
and global scale, the yearly manufacturing of concrete surpasses any other material, with 8 billion
cubic meters globally, and 180 million cubic meters within the United States (Ulm, 2008;
Kosmatka and Wilson, 2011). Whether dictated by consumer demand, or simply by availability of
resources, the relentless global appetite for concrete continues to further research and development
in its strength, durability, and sustainability properties.
In the early twentieth-century Abrams (1919) determined that water content was the critical
constituent that determines the resulting compressive strength of concrete. Approximately a
decade later, researchers discovered concrete’s strength could be increased during placement
through the application of pressure (Richard and Cheyrezy, 1995). Furthermore, it wasn’t until the
1970’s that concrete’s density was increased by the addition of silica fume, and/or water reducing
admixtures, allowing the compressive strength to drastically increase in the range between 50-120
megapascal (MPa) (Hertz, 2003; Buchanan and Abu, 2017). Such a concrete is referred to as high
strength concrete (HSC), and concrete with compressive strength less than 50 MPa is commonly
denoted as normal strength concrete (NSC). Finally, in the mid-1990s researchers developed a
concrete with a denser microstructure, named ultra-high-performance concrete (UHPC), otherwise
known as reactive powder concrete (RPC). UHPC provides compressive strength beyond 150
MPa, tensile strength (post-cracking) higher than 5 MPa, elastic modulus in the range of 40-70
1

gigapascal (GPa), protection from detrimental compounds (chloride or alkali silica reaction),
freeze-thaw resistance, blast resistance and increased ductility with the inclusion of steel fibers
(Kosmatka and Wilson, 2011; Graybeal, 2015; Gu et al., 2015; Li et al., 2015). Such supreme
mechanical properties enable UHPC to be a versatile and durable material. UHPC has a surfeit
amount of applications including construction of infrastructure, strengthening and rehabilitation
of structures, and within military, petroleum, marine, and nuclear facilities (Kosmatka and Wilson,
2011; Zheng et al., 2014a; Gu et al., 2015). In North America UHPC is predominantly utilized for
the construction of bridges to exploit its excellent compressive and tensile strength properties
(Graybeal, 2014). As of 2014, 55 bridges and overpasses were constructed employing UHPC
(Russell and Graybeal, 2013). The Hosokawa River Tunnel in Japan is the first in its class to be
retrofitted utilizing UHPC (Gu et al., 2015).
However, UHPC’s implementation is limited to certain areas due to autogenous shrinkage
during the early stages of curing, high cost and its poor fire resistance characteristics. Extreme
cases of autogenous shrinkage occur in very thin applications of UHPC, but can be avoided with
certain admixtures. The initial high cost is due to the volume and type of ingredients used (cement,
silica fume, etc.), and availability of such materials (Soliman and Nehdi, 2011; Canbaz, 2014;
YazÄącÄą et al., 2010; So et al., 2015). Nevertheless, implementing a stronger material as UHPC
allows the structural engineer to choose a reduced cross section (Russell and Graybeal, 2013).
Furthermore, the cost may be driven down by implementing industrial by-products within the
UHPC design mix. Contents of silica fume and cement may be reduced by substituting with ground
granulated blast furnace slag (GGBFS) or fly ash, while maintaining UHPC’s high mechanical
properties (Gesoglu et al., 2015; Xiao et al., 2014; YazÄącÄą et al., 2009; Alkaysi, 2016; So et al.,
2015). Whereas UHPC’s reduced cross section subsides the cost, with increasing temperatures it

2

exacerbates the strength degradation and chance of explosive spalling (Hertz, 2003; Kodur, 2014).
The spalling (Figure 1.1) process occurs when concrete is hastily heated, and the vaporized water
begins to increase and build up. If there is no clear avenue to release the built-up pore pressure
(due to low permeability), it will eventually surpass the concrete’s tensile strength and experience
explosive spalling (Khaliq, 2012; Buchanan and Abu, 2017). Depending on the type and size of
fire, a structure exposed to such an extreme environment signifies one of the most dire
circumstances (Kodur and Phan 2007; Zheng et al., 2012; Kodur, 2014). Given the predominant
use of UHPC within North America is for transportation structures, detailed data pertaining to such
structures affected by fire exposure is now presented. Transportation structures must endure higher
temperatures since the source of fire is petroleum, and fire temperatures can reach 815°Celsius
(°C) within the first three minutes (min), which is 66% greater than a standard temperature versus
time fire curve such as ASTM E119 (Kodur and Phan, 2007).
During the last 22 years, seven fires within tunnel structures have occurred. The Channel
Tunnel, connecting Britain and France was the first to suffer such a disaster while employing HSC
(Hertz, 2003; Ingason, 2004; Aydin and Baradan, 2012; Bei and Zhixiang, 2016). Due to HSC’s
low permeability the Channel Tunnel, along with the Great Belt Tunnel in Denmark, suffered
extreme damage in the form of spalling (Aydin and Baradan, 2012; Buchanan and Abu, 2017).
The Channel and Great Belt Tunnels’ fire reached a maximum temperature of 1,100 and 800°C,
respectively (Khoury, 2000). On the other hand, bridges experience greater damage due to fires.
From 1989 to 2000, 16 U.S. bridges collapsed due to fire, and from 2002 to 2015, the number
jumped to 27 (Naser and Kodur, 2015; Aziz, 2015). Furthermore, records from 2014 reveal, of the
200,000 fire occurrences via motor vehicles in the United States, nearly 90% occurred while
traveling on the highway resulting in 345 deaths and $1.5 billion worth of damage (NFDC, 2017).

3

Considering the increased desire for engineers to select UHPC as a building material, the fire
behavior of UHPC must be investigated, and as per local structural provisions these structures
must adhere to the fire resistance requirements. However, no requirements regarding fire resistance
currently exist for bridges further increasing the consternation, and proving to be quite contrary to
fire safety’s main objectives (Naser and Kodur, 2015; Buchanan, and Abu, 2017).
1.2 Fire resistance properties of concrete
The fire resistance of a structural member is its ability to withstand elevated temperatures
without compromising its structural integrity, stability, and the transmission of heat (Kodur, 2014).
Different criteria may be utilized, but fire resistance is typically measured in time. Moreover,
depending on the type of structure, it must maintain the previously mentioned conditions for a
given specified duration as prescribed in local codes. Thus, for infrastructure applications, fire
resistance is imperative. In addition, concrete as a material possesses excellent resistance
properties because of its non-combustible nature, coupled with its low thermal conductivity, and
high specific heat (Behloul et al., 2002; Kosmatka and Wilson, 2012; Kodur, 2014; Buchanan and
Abu, 2017). The combination of such qualities enable concrete to experience a gradual progression
of heat across the section and minimize the spread of fire throughout the structure (Kosmatka and
Wilson, 2011). However, with the inclusion of silica fume and reduction of water content to
achieve denser concrete, the fire resistance is inversely proportional to its strength, and explosive
spalling further thwarts its resisting ability. As opposed to NSC and HSC, UHPC’s rudiments are
attributed to its granular elements allowing reduction of the cross section, precipitating the
diminution of its fire resistance. Therefore, prior to implementing a novel material such as UHPC,
understanding the potential of its material behavior with increasing temperature is preeminent.
Exploiting the ambient condition mechanical properties of concrete allow the development of

4

varying structural arrangements. However, considering the inherent hazard a fire disaster poses
and to ensure the structure meets its fire resistance requirements, the mechanical properties of
concrete at elevated temperatures must be understood. The following section will discuss the
constituents of concrete influencing the principle mechanical properties, and the risk of
temperature induced strength dilapidation.
1.3 Mechanical properties of concrete at elevated temperature
Mechanical properties of materials experience a degradation with increasing temperatures.
Nevertheless, the composition and construct of concrete can serve as a catalyst to the rate of
reduction. The cement paste undergoes several decompositions and dehydration of chemically
bound water, resulting in strength loss at each respective temperature. Moreover, the amount of
cement can further exaggerate the decrease in strength (Klingsch, 2014). The type and size of
aggregate can influence the fire resistance as well. With the former, limestone’s specific heat is
much greater than quartz, therefore providing a greater resistance to fire exposure (Kodur and
Phan, 2007). Additionally, quartz’s dramatic expansion in volume at temperatures above 500°C,
results in further loss of strength (Schrefler et al., 2002; Jang et al., 2016). The latter, dictates the
permeability of the concrete, thereby influencing the strength. Utilization of silica fume or the use
of finer aggregates jointly with silica fume, will deliver a progressively increasing dilapidation of
the concrete’s mechanical properties and allow the concrete to be more vulnerable to explosive
spalling (Kodur, 2014). Including polymer fibers to denser concretes may alleviate the explosive
spalling but will decrease its strength.
As compared to conventional concretes, UHPC is designed on the premise to increase the
fineness of its elements to capitalize on its packing density (Kosmatka and Wilson, 2011; Ju et al.,
2013; Zdeb, 2013; Richard and Cheyrezy, 1995; Canbaz, 2014). The optimization of the packing
5

density of UHPC is achieved through minimizing the porosity of the UHPC, and exploiting any
available space within the concrete’s matrix (Magureanu et al., 2012; Zdeb, 2013). The dense
microstructure is achieved through a combination of, (1) low water-to-cement (w/c) ratio of 0.20
or lower, (2) addition of superplasticizers, silica fume, and steel fibers, (3) employing quartz sand
vice coarse aggregate (grain size is reduced by a factor of 50), and/or (4) through additional
treatments, such as thermal, and/or applied pressure (Hager et al., 2013; Zdeb, 2013; Wang et al.,
2015; Wang and Gao, 2016). Thereby, the mechanical properties will suffer a rapid degradation,
as opposed to the conventional concretes. Additionally, though UHPC has tremendous
compressive and tensile strength, the dense matrix causes it to be tremendously vulnerable to
explosive spalling and the data associated to UHPC at high temperature conditions is limited.
UHPC’s strongest attributes conversely serve as its paramount imperfection at elevated
temperatures.
1.4 Objectives of this research
The foremost aims of this research are to illustrate the variation of temperature-dependent
compressive strength of different UHPC types as compared to other conventional types of
concretes, and to highlight the details pertaining to explosive spalling within UHPC through an
inclusive experimental study. To successfully complete these goals, the research objectives listed
below were accomplished:
▪

Conduct a detailed state of the art literature review on the thermal and mechanical
responses of concrete at elevated temperatures. Also, to enumerate the pertinent research
gaps, specifically with respect to UHPC properties at elevated temperatures. The thorough
literature review will incorporate experimental and analytical studies.

6

▪

Evaluate compressive strengths of three design mixes of UHPC at various temperatures
(20, 100, 200, 300, 400, 600 and 700°C), immediately following exposure to elevated
temperatures. Also, undertake compressive strength tests with differing design mixes of
NSC and HSC for comparison.

▪

Trace the explosive spalling behavior of the UHPC specimens at the elevated temperatures
stated previously.

▪

Compare the temperature dependent compressive strength and explosive spalling behavior
of UHPC with other concrete types, ASCE (1992) and Eurocode 2 (2004).

7

Figure 1.1: Photos depicting explosive spalling in a NSC and HSC column under elevated
temperatures (Khaliq, 2012)

8

CHAPTER TWO
2. STATE OF THE ART REVIEW
2.1 General
UHPC’s paragon properties allow the employment of such a material in diverse functions.
However, with increasing temperatures, UHPC will result in a disparate degradation and
associated properties given its granular microstructure; Thus posing a hindrance in reaching the
required fire resistance. Therefore, for evaluating fire resistance of structural members, properties
at elevated temperatures are required. Moreover, understanding the temperature induced
physiochemical changes provides further insight regarding concrete’s deterioration.
Klingsch (2014) presents a consolidated table that is consistent with other research
conducted, providing the rising temperature with the effect on concrete (Table 2.1). The most
notable points will be presented and the data is from the following research, Felicetti et al. (2000),
Schrefler et al. (2002), Dehn (2004), Klingsch (2014), Way and Wille (2015), Zheng et al. (2015)
and Bei and Zhixiang (2016). At temperatures above 105°C, the physical water is completely
evaporated causing an increase in porosity and mechanical properties to decrease. Whereas in
higher strength concretes the free water can still produce disastrous results in the form of explosive
spalling. The likelihood of explosive spalling is increased at temperatures of 100-300°C upon the
evaporation of the chemically bound water within the cement paste. The decomposition of the
binder at the above-mentioned temperature range triggers strength loss as well.
Further strength loss is attributed to the following, (1) the formation of lime (CaO), from calcium
hydroxide, Ca(OH)2, and the magnitude of strength loss is contingent upon the calcium-oxide
(CaO) to silica-dioxide (SiO2) (C/S) ratio of the concrete mixture, (2) concrete with siliceous

9

aggregate will cause apparent expansions and result in microcracking due to the Îą-quartz
transforming to β-quartz around 573ºC, (3) temperatures above 700°C the decarbonation of CaCO3
produces additional CaO, along with carbon dioxide (CO2), and from 800-1000°C all chemically
bound water within the cement matrix evaporates. Upon cooling of the concrete, CaO rehydrates
from the air and CaCO3 reforms, resulting in drastic widening of the cracks (up to 40% increase).
Lastly, within siliceous aggregate concrete, the β-quartz retransforms to ι-quartz at 573ºC
followed with a steep contraction (Khoury, 2000; Schrefler, 2002; Hertz, 2005). Though research
pertaining to traditional concrete’s fire behavior is ongoing, owing to UHPC’s composition and
construct, grasping its full depth is in its embryonic stage. The following sections will present
research pertaining to concrete’s thermal and mechanical properties, and the occurrence of fire
induced spalling.
2.2 Thermal properties of concrete
2.2.1 General
Concrete is an inert material possessing low thermal conductivity and high specific heat.
Such qualities enable it to prevent the spread of fire because it diminishes the progress of heat.
However, due to the concrete undergoing physiochemical changes with increasing temperatures,
the thermal properties do not remain constant, thereby inflicting change upon the mechanical
behavior of concrete with increasing temperatures. Therefore, understanding the behavior of a
material exposed to high temperatures, enable the designer to fully exercise the material’s full
potential.
2.2.2 Test methods for high temperature thermal properties
There are two methods to determine the thermal conductivity of a material. First is the hot
wire method and second is the guarded heat flow meter method. The guidelines associated with
10

the former are presented in ASTM C1113/C1113M (2013), and for the latter in ASTM E1530-11
(2016). The hot wire method may be utilized to evaluate the thermal conductivity up to 1500°C,
and utilizes an electrical current passing through a wire that is between two blocks of the material
to be tested. Depending upon how fast the heat flows to the material block determines the thermal
conductivity value. The second method, guarded heat flow meter method, can accurately provide
the thermal conductivity up to 325°C. At temperatures beyond 325°C, it is dependent upon the
calibration of the equipment. In this method, the material to be tested, along with a heat flux
transducer (HFT), are placed between two controlled temperature plates. Therefore, dependent
upon the heat transfer from one plate to the other determines the thermal conductivity for the
material.
To determine the specific heat capacity, the guidelines prescribed in ASTM E1269-11 (2011)
recommend using differential scanning calorimetry method. The test is accurate at temperatures
between 100-600°C, and beyond it is dependent upon the equipment utilized. The test can be
performed very quickly and is simple to conduct. Essentially, select a benchmark and measure the
change in heat flow between another selected point and the benchmark.
To evaluate the linear thermal expansion of a solid material, one method is to employ a
thermomechanical analysis as recommended in ASTM E831-14 (2014). The test is effective up to
900°C, and beyond it is dependent upon the equipment utilized. The specimen is exposed to a
consistent heating rate and the expansion of the material is recorded with respect to temperature.
Lastly, to determine the mass loss of a material, ASTM E1868-10 (2015) recommends utilizing
the thermogravimetry method. The test is valid up to 1000C and applicable to a wide range of
materials. The specimen is subjected to a consistent temperature and the mass is measured for

11

either a specified time or once the loss attains a specified value. Following the heating process, the
percent of mass loss is calculated by dividing the mass loss divided by the original mass.
2.2.3 Previous studies on thermal properties
Investigations pertaining to conventional concrete’s thermal properties have been ongoing
for over 4 decades. Moreover, ASCE (1992), Eurocode 2 (2004) and several researchers
formulated constitutive relationships pertaining to both NSC and HSC’s thermal properties with
increasing temperatures. As compared to conventional concretes, UHPC’s thermal properties have
scarcely been researched in the last 2 decades. Furthermore, no current codes dictate the
constitutive relationship of UHPC either. The following two sections will impart the notable points
from literature pertaining to the effect of elevated temperatures on the thermal properties of
conventional concretes and UHPC.
2.2.4 Effect of temperature on thermal properties of conventional concretes
As compared to carbonate aggregate, siliceous aggregate contributes negatively with
respect to fire resistance, and except for the thermal expansion at temperatures above 700°C, NSC
and HSC’s thermal properties are very similar (Lie and Kodur, 1996; Kalifa et al., 2000; Sultan
and Kodur, 2003). Concrete with siliceous aggregate’s specific heat remains steady and only
experiences a bump around 450ÂşC, which is attributed to the formation of CaO and H2O from
Ca(OH)2. Moreover, the coefficient of thermal expansion (CTE), of siliceous aggregate concrete
will abruptly rise around 550°C, due to the quartz transformation. While, the disjoining of CaCO3
will cause two occurrences to transpire in carbonate aggregate concrete. At 700°C, the specific
heat undergoes a drastic spike, and at 800°C the CTE deviates from a steady linear rise (So et al.,
2014). The thermal conductivity experiences a steady decrease, with the siliceous concrete having

12

the higher values and, the mass loss progressively rises linearly but drastically reduces beyond
700°C due to all the chemically bound water evaporating.
2.2.5 Effect of temperature on thermal properties of UHPC
Aside from the thermal diffusivity, RPC’s thermal conductivity, specific heat capacity and
mass loss are lower than HSC. Additionally, if compared at temperature levels, the CTE is about
the same (Ju et al., 2011). The thermal conductivity and thermal diffusivity decrease with
increasing temperatures, while the specific heat capacity and CTE continuously rise (Behloul et
al., 2002; Zheng et al., 2014b). Ju et al. (2011) examined the effects of steel fiber content and
discovered altering the content of steel fiber only influences specific heat capacity and thermal
diffusivity.
Behloul et al. (2002) and Wang et al. (2013) studies reported very similar results pertaining
to the CTE. Both experienced a linear progression (roughly 12e -6/°C) to about 500°C and both
encounter a slight bump between 300-400°C. The bump between 300-400°C is attributed to the
decomposition of Ca(OH)2. Beyond 500°C, the CTE experiences a dramatic increase due to the ιquartz transforming to β-quartz, followed by a decline after 700°C. Finally, Zheng et al. (2014b)
exposed RPC to temperatures of 20 to 800°C and found a smooth and steady mass loss with
increasing temperatures.
2.3 Mechanical properties of concrete
2.3.1 General
The major proponents of concrete’s mechanical properties are compressive and tensile
strength, ductility, stress-strain relationship, and elastic modulus. The preceding properties
degradation dictates whether the fire resistance design is acceptable. Prior to presenting the data
from literature pertaining to concrete’s behavior to increased temperatures, a synopsis regarding
13

concrete’s mechanical behavior under ambient conditions is discussed, and research investigating
UHPC/RPC and ultra-high performance fiber reinforced concrete (UHPFRC) under service
conditions is presented.
At room temperature, the w/c ratio coupled with the percentage of contact amid the binder
with the coarse aggregate, called interface transition zones, play a crucial role regarding the
concrete’s strength (Kodur, 2014; So et al., 2014; Kamen et al., 2007). In very dense concretes,
the strength is directly contributed by the w/c ratio, integrated with granular constituents and silica
fume (Xiao et al., 2014). Specifically, superfluous amounts of silica fume (20-25% the weight of
cement) make up UHPC’s binder, and as cement (C3S) mixes with water, (H2O) the chemical
reaction results in calcium silicate hydrate, C-S-H and Ca(OH)2. Therefore, in addition to
minimizing the voids within UHPC’s matrix, silica fume contributes further by decreasing the C/S
ratio, resulting in less production of Ca(OH)2, and increase the production of C-S-H (Khoury,
2000; Schrefler et al., 2002; Kosmatka and Wilson, 2011; Zdeb, 2013). The C-S-H’s role in dense
concretes is quite significant as it serves as the chief binder directly contributing to its mechanical
properties and houses majority of the chemically bound water (Burke, 2011; Jagannathan, 2013;
Klingsch, 2014). Considering UHPC’s very dense microstructure, it enables UHPC to have
superior mechanical properties as compared to NSC and HSC at ambient conditions.
The current literature pertaining to UHPC/RPC or UHPFRC at ambient conditions predominantly
explores different methods to increase the mechanical properties utilizing various treatments or
with altering the main ingredients. On the contrary, research pertaining to exploring
environmentally friendly and/or sustainable UHPC is quite limited. YazÄącÄą et al. (2010), Aydin and
Baradan (2013), Gesoglu et al. (2015) and Peng et al. (2016) conducted research to develop a
UHPC that maintains its mechanical properties while being environmentally conscience,
14

sustainable or economical. Table 2.2 provides a summation of tests regarding UHPC/RPC and
UHPFRC, conducted in ambient conditions. Unless compressive strengths greater than 300 MPa
are required, developing the appropriate design mix (with no treatment) will result in compressive
strengths as high as 200 MPa, (Wille et al., 2011; Zdeb, 2013; Voit and Kirnbauer, 2014).
2.3.2 Testing methods for high temperature mechanical properties
Extensive research studies at ambient conditions for NSC, HSC and UHPC conducted by
Elwell and Fu (1995) and Graybeal and Davis (2008) provide guidance on choosing the specimens’
shape and state that choosing a cylinder or cube specimen is purely preference and does not alter
the strengths. Furthermore, no ASTM standard exists to evaluate mechanical properties of concrete
at elevated temperatures. Therefore, RILEM Standard (RILEM Technical Committee 200-HTC,
2007) is utilized to standardize the heating process and dimension requirements, while the
respective ASTM standard is implemented for each different mechanical property test.
Consolidated recommendations from the RILEM standard pertaining to the high temperature
mechanical property tests are presented.
RILEM standard recommends three different steady state tests to evaluate mechanical properties
at elevated temperatures, (1) unstressed, (2) stressed, and (3) residual state. The unstressed
condition demands once the specimen attains its testing temperature, to remain in the furnace for
a specified duration to reach complete saturation across the specimen. If a load is to be applied
during the heating process, referred to as a stressed test, it must remain constant and must be a
certain percentage of the at room temperature compressive strength until complete saturation is
attained. Upon saturation of the specimen, the load is increased until failure of the specimen. For
the residual-state, the specimen must completely cool down to the ambient condition prior to
testing. The compressive strength, stress-strain behavior, and elastic modulus may be evaluated
15

using either of the three methods. While the tensile strength is recommended only for unstressed
or residual state.
2.3.3 Previous studies on mechanical properties
Just as each element serves a purpose to contribute to concrete’s facets, the constituent(s)
can undesirably affect the structural member’s performance with increasing temperatures. There
is significant ongoing research evaluating the mechanical properties of conventional concretes.
Owing to HSC’s differing microstructure, literature indicates HSC’s mechanical properties
decrease faster as opposed to NSC (Zheng et al., 2012; Kodur, 2014). Furthermore, HSC is prone
to explosive spalling and mitigation techniques must be utilized. Further discussion pertaining to
fire induced spalling will be presented in section 2.4. Discordant with conventional concretes, the
bulk of UHPC’s studies have been conducted in the last 5 years. Attributed to UHPC’s granulated
elements, and superior strength, its fire performance is dissimilar as compared to the conventional
concretes. UHPC’s assets touted by its advocates prove to be its downfall when exposed to elevated
temperatures. With increasing temperature, the multitude of physiochemical changes, dilapidation,
and the inconsistency of properties between the binder and aggregate undergo, heavily contribute
to UHPC’s mechanical property degradation (Khoury, 2000; Zheng et al., 2015; Li and Liu, 2016).
Unless proper measures are taken (for instance the addition of fibers or employ pozzolanic
resources), UHPC will not be able to endure the elevated temperatures for long (Khoury, 2000; Ye
et al., 2012).
Table 2.3 provides a summation of studies regarding UHPC/RPC and UHPFRC conducted
at elevated temperatures, identifying the strengths, and disadvantages for each investigation.
Summarizing Table 2.3 the current research pertaining to UHPC’s mechanical properties exposed
to increasing temperatures is dominated by residual state testing of customary UHPC design mixes.
16

Less than one-third explore the behavior with varying fiber type and content. Only three studies
evaluate UHPC with a treatment and an additional three investigate the fire behavior of more
economical and environmental design mixes of UHPC. In addition, most studies evaluate the
compressive and tensile strength and only a handful cover the stress-strain relationship along with
the elastic modulus. Finally, there are very few investigations proposing models to predict the
explosive spalling behavior of UHPC. The following sections will impart the notable points from
literature pertaining to the effect of high temperatures on the mechanical properties of conventional
concretes and UHPC.
2.3.4 Compressive strength
2.3.4.1 Effect of temperature on compressive strength of conventional concretes
Investigations pertaining to conventional concretes display compressive strength
decreasing with rising temperatures and HSC experiencing a steeper degradation (Arioz, 2007;
Sarhat and Sherwood, 2013). However, due to the removal of free water and consolidation of
concrete, the denser HSC design mixes underwent an increase in strength between 200-300ÂşC
(Ghandehari et al., 2010). Furthermore, the 300 to 600ÂşC range is crucial for denser concretes, with
great strength loss by 600ºC (Demirel and Keleştemur, 2010; Ghandehari et al., 2010). A handful
of researchers utilized either recycled or locally available aggregate for NSC or HSC, and
discovered the results are comparable to typical aggregates, and quartz dominant concrete will
provide a poorer performance with increasing temperatures (Sarhat and Sherwood, 2013; Hager et
al., 2016). Yet, if a pozzalanic material blend is implemented to reduce the amount of cement, the
concrete’s strength underwent a more rapid deterioration (Demirel and Keleştemur, 2010; Li et al.,
2014). In addition, employing higher contents of cement will provide greater strengths under
ambient conditions, while suffering a quicker decline with the temperature rise (ErgĂźn et al., 2013).
17

A higher content of w/c ratio does not greatly affect the relative strengths, but will allow a more
rapid decrease in strength (Arioz, 2007; Al-Jabri et al., 2016).
2.3.4.2 Effect of temperature on compressive strength of UHPC
UHPC is contrived to deliver high compressive strengths at ambient conditions. However,
unless the proper mixture is formulated, UHPC will experience sudden degradation and/or
explosive spalling. UHPC’s compressive strength at elevated temperatures is dependent upon the
design and Figure 2.1 illustrates the sporadic results amongst the literature for UHPC/RPC studies
from the last three years (Kodur, 2014).
Behloul et al. (2002), Ye et al. (2012) and Zheng et al. (2015) were the only studies
evaluating the hot-state compressive strength and two of those, Behloul et al. (2002) and Ye et al.
(2012) were stressed while being heated. The non-cured UHPC mix studied by Behloul et al.
(2002) performed exceptionally well, attaining 120% of its room temperature strength at 500°C.
The unstressed test degraded the fastest to 50% at only 500°C (Zheng et al., 2015).
The remainder of tests evaluated the residual compressive strengths. For the most part the
tests conducted attain 115-160% of their respective original strength between 100-400°C, soon
followed by a steady decline. Aydin and Baradan (2012) and Bashandy (2013) studies experienced
explosive spalling beyond 300°C. Aydin and Baradan (2012) design contained high volume of
silica fume and cement, while Bashandy (2013) utilized locally available steel fiber and cement to
produce an economic RPC. So et al. (2015), Way and Wille (2015) and Jang et al. (2016) achieved
the greatest residual compressive strengths of 145, 160 and 140% at 400, 300 and 400°C,
respectively. Zheng et al. (2014a) and Zheng et al. (2014b) investigated UHPC with polypropylene
and steel fiber alongside UHPC with polypropylene fiber, nearly lost 40% of the compressive
strength at 100°C and only increased to 80% between 500-600°C.
18

Aydin and Baradan (2012), So et al. (2015) and Jang et al., 2016 studied the residual
compressive strength behavior of UHPC with a ternary pozzalanic material blend. All three studies
conducted tests at and beyond 800°C without experiencing explosive spalling and with
comparable, if not improved fire behavior than their respective typical UHPC formula. Sanchayan
and Foster (2016) tested RPC with a combination of steel fiber and polyvinyl alcohol (PVA) fibers.
The RPC mix with high PVA and low steel fiber content behavior was similar, save the strengths
attained at 300 and 400°C. Additionally, the high PVA fiber content RPC underwent explosive
spalling beyond 500°C.
Approximately 15 of the studies evaluated the effects of varying the fiber content of the
UHPC. Two researchers concluded the steel and polypropylene fibers contribute positively and
0.9% polypropylene fiber content is recommended (Tai et al., 2011; Abdul-Hussain, 2013).
However, Hager et al. (2013) study revealed the content of polypropylene fibers do not provide an
impact to UHPC’s compressive strength. Canbaz (2014), Way and Wille (2015) and Xiong and
Liew (2015) investigated the effects of various curing treatments. Canbaz (2014) evaluated RPC
pressure treated and cured in water at 90°C and Way and Wille (2015) oven dried the UHPFRC
and both achieved improvements with respect to the compressive strength. The RPC in Canbaz
(2014) study experienced an increase up to 400°C, followed by a decline. However, Xiong and
Liew (2015) research did not provide any substantial effect with respect to varying curing
techniques.
2.3.5 Tensile strength
2.3.5.1 Effect of temperature on tensile strength of conventional concretes
The tensile strength of conventional concretes is influenced by the following at elevated
temperatures, (1) enhancement through the addition of steel or hybrid fibers, (2) undergo
19

dilapidation with increasing temperatures, (3) experience a dramatic decrease beyond 300°C, and
(4) HSC’s tensile strength sustains a greater decrease as opposed to its compressive strength
(Ghandehari et al., 2010, Khaliq and Kodur, 2011). Furthermore, the alteration of the cement
content alone does not pose much of a difference with respect to the relative tensile strength, but
employing a ternary pozzolanic blend will result in a drastic reduction to 20% of its original tensile
strength at 500°C (Li et al., 2014). As expected, the quartz aggregate concrete displayed greater
reduction as opposed to limestone aggregate, and increasing the w/c ratio will decrease the tensile
strength as well (Al-Jabri et al., 2016; Hager et al., 2016). Finally, similar to the compressive
strength, very dense HSC’s tensile strength will increase from 100 to 200ºC, followed by a decline
(Ghandehari et al., 2010).
2.3.5.2 Effect of temperature on tensile strength of UHPC
Only two studies evaluated both the hot-state and residual-state tensile (direct) strength of UHPC,
Felicetti et al. (2000) and Behloul et al. (2002). While Felicetti et al. (2000) results displayed very
little difference between the test types, Behloul et al. (2002) study demonstrated the hot-state
tensile strength was slightly higher. The remainder of the residual direct tensile tests behave very
similarly with increasing temperatures and achieve 50% of original strength between 450-550
(Zheng et al., 2014a; Zheng et al., 2014b; Li and Liu, 2016).
Four separate investigations tested the residual splitting tensile strength and Peng et al.
(2012), Abdul-Hussain (2013) and So et al. (2015) results are very similar. Their original strengths
reduce to 50% at about 600°C. While Bashandy (2013) presented drastic degradation, 50% at
300°C, may be attributed to utilizing cheaper steel and cement. Finally, five research studies tested
the residual flexural strength and Aydin and Baradan (2012) and Ali et al. (2016) tested typical
RPC mixture containing silica fume could not withstand temperatures above 300 and 400°C,
20

respectively. Additionally, Aydin and Baradan (2012), So et al. (2015) and Ali et al. (2016) found
the RPC formulation implementing ternary pozzalanic materials performed quite well at and
beyond 600°C. With So et al. (2015) results were exceptionally well, attaining near 150% of
original strength at 400°C, reducing to 50% around 700°C and 22% at 1000°C. Aside from Aydin
and Baradan (2012) ternary pozzalanic RPC mixture, the remaining RPC’s tested up to and beyond
600°C, reached 50% of their respective original strength.
Khoury (2000) claims employing the splitting tensile and flexural strength tests do not
provide an accurate representation and concludes the direct tensile strength tests is supreme.
However, the literature is evenly distributed amongst the 3 types of tensile strength tests. Finally,
increasing the steel fiber content directly contributes to tensile strength (Abdul-Hussain, 2013; Li
and Liu, 2016). Figure 2.2 depicts the tensile strengths (direct tensile, splitting tensile and flexural
strengths) at elevated temperatures from literature.
2.3.6 Stress-strain relationship and elastic modulus
2.3.6.1 Effect of temperature on stress-strain relationship and elastic modulus of
conventional concretes
With increasing temperatures conventional concretes’ peak strain increases, the ascension
portion is near linear, the descent portion becomes progressively recumbent at temperatures of
500ÂşC and above, and the compressive strength and elastic modulus lessen (Chang et al., 2006;
Zheng et al., 2012). Additionally, the elastic modulus degrades the fastest with rising temperatures
amongst the mechanical properties. Finally, the elastic modulus is further reduced utilizing a
ternary pozzolanic mixture to reduce the cement content. A study conducted by Li et al. (2014)
discovered the addition of high contents of GGBFS hastily declines the elastic modulus at 500°C
to nearly a tenth of its original elastic modulus.
21

2.3.6.2 Effect of temperature on stress-strain relationship and elastic modulus of UHPC
Only 3 of the studies evaluated the stress-strain relationship, of which two were residual.
Zheng et al. (2012) demonstrated residual stress-strain curve attains the peak strain at 600°C,
minimal change is seen up to 300°C and beyond 300°C, the curve becomes more flattened. While
Tai et al. (2011) found the residual stress-strain curve’s slope steadily decreases with progressing
temperature. Greater steel fiber content improved the peak strain between 200-800°C for Tai et al.
(2011) but only up to 600°C for Zheng et al. (2015) study. Moreover, Zheng et al. (2015)
determined 2% steel fiber content to be most optimum for increasing ductility.
Three tests evaluated the hot-state elastic modulus, Behloul et al. (2002) research indicated
UHPC decreasing to 50% of its original stiffness beyond 600°C, while Zheng et al. (2015)
experienced a sudden degradation at 200°C. Behloul et al. (2002) conducted residual elastic
modulus tests as well and found the residual elastic modulus to perform exceptionally well. The
UHPC studied maintained within its original stiffness up to 600°C, only to reduce to 80% at 700°C.
Hager et al. (2013) was the only other study to reach 50% decline around 800°C. Aside from
Behloul et al. (2002) and Hager et al. (2013), the bulk of the remainder elastic modulus results,
declined to 50% of its stiffness between 350-500°C.
Xiong and Liew (2015) test consisted of various curing methods and heating rates. Their
findings displayed the curing techniques did not influence the stiffness and the UHPC exposed to
higher heating rates degraded much sooner. Hager et al. (2013) tested various polypropylene fiber
content and concluded the addition of polypropylene fiber does not contribute to RPC’s stiffness.
Figure 2.3 depicts the elastic modulus at elevated temperatures from literature for UHPC/RPC
studies from the last two years.

22

2.4 Fire induced spalling
2.4.1 General
Hertz (1984) encountered explosive spalling and hypothesized ultra-high strength
concretes developed utilizing silica fume, to densify the concrete’s matrix, have a greater risk to
experience explosive spalling. The explosive spalling was attributed to the obstruction of the
vaporized water attempting to evacuate the heated specimen. Following further research, Hertz
(2003) concluded the increased strength does not necessarily correlate to explosive spalling as
much as the exorbitant dense medium attained with silica fume. Although HSC is more susceptible
to explosive spalling than NSC, HSC tested with no silica fume proved to be tolerable of increasing
temperatures and did not explode (Phan, 1996; Hertz, 2003; Kodur, 2014). Additionally, a
comprehensive study conducted by Klingsch (2014) determined UHPC lacking silica fume
exposed to extreme heating rates to survive without exploding either. Silica fume maintains a stout
link with explosive spalling but there are other crucial factors as well. The following sections will
present data from literature pertaining to the foremost causes of spalling and methods to minimize
the occurrence.
2.4.2 Causes of spalling
Minimal permeability, temperature gradients, migration of moisture and pore pressure
build-up, and thermal dilation contribute to explosive spalling (Zheng et al., 2013; Kodur, 2014;
Xiong and Liew, 2015; Jang et al., 2016). Additionally, the threat is further augmented by the type
of aggregate, specimen dimensions, low tensile strength, type of fire exposure and contents of
silica fume, cement (w/c ratio) and limestone filler (Hertz, 2003; Zheng et al., 2013; Klingsch,
2014; Kodur, 2014). Consequently denser concretes experience explosive spalling mainly due to
the development of temperature gradients and pore pressure build-up. Furthermore, UHPC is
23

formulated implementing granular constituents and may not adhere to the same mechanisms of
explosive spalling as HSC, and regrettably UHPC’s design foundation is concocted upon most of
the risks previously listed (Ju et al., 2015; Ali et al., 2016). Therefore, UHPC will experience
explosive spalling at a lower temperature and shorter time period (or much sooner) than HSC
(Phan, 1996). According to literature, several researchers determined explosive spalling can occur
at temperatures of 250°C and above (Khoury, 2000; Burke, 2011; Zheng et al., 2013).
2.4.2.1 Rate of heat and thermal stresses induced by thermal gradient
Attributed to concrete’s low thermal conductivity, fire exposure causes thermal gradients
to develop resulting in a fluctuation of physiochemical variations across the cross section and
thermal stresses (Hertz, 1984). The result is further compounded with applications of excessive
heating rates (type of fire). Hertz (1984) research determined the critical rate of heat to avoid
explosive spalling with immensely dense concretes is as low as 1ÂşC/min. Klingsch and Frangi
(2011) precursory investigation determined the same value (1ÂşC/min) for UHPC. However,
decreasing the heating rate does not necessarily correspond to mitigation of explosive spalling,
because it only minimizes the thermal gradients (Felicetti et al., 2000; Klingsch, 2014). Thermal
gradients vastly impact the development of pore pressure and will be discussed further in the next
section (Phan, 2008; Khaliq, 2012).
2.4.2.2 Pore pressure build-up
Current literature correlates the built-up pore pressure within concrete to be the pivotal
cause leading to explosive spalling (Klingsch, 2014; Sanchayan and Foster, 2016). The
complication is further exacerbated with low permeability and extreme heating rates (Burke,
2011). In addition, various studies conducted found chemically bound water alone can cause
explosive spalling (Hertz, 1984; Hertz, 2003; Sanchayan and Foster, 2016). Therefore, it is no
24

question due to UHPC’s exceptionally dense matrix coupled with high amount of cement,
explosive spalling occurs.
With increasing temperatures, thermal gradients within the concrete’s cross section will
develop and covert the moisture into a vapor. Since concrete maintains a certain amount of
permeability, the water nearest to the surface will evaporate, however some of the moisture will
follow the pressure gradient and travel inwards and condensate due to encountering cooler
temperatures (Ju et al., 2013; Klingsch, 2014). Eventually a saturated barrier will form resulting
in increasing pore pressure and soon followed by explosive spalling (Burke, 2011; Klingsch, 2014;
Kodur, 2014). To withstand such a pressure increase, it is paramount for low permeability concrete
to have great tensile strength (Klingsch, 2014; Kodur, 2014). The process is referred to as moisture
clog and is well documented within the literature (Kalifa et al., 2000; Phan, 2008; Dwaikat and
Kodur, 2009; Felicetti and Lo Monte, 2013).
Explosive spalling can be predicted because the point of explosion will have a much higher
pressure than amongst the remaining specimen (Ju et al., 2013). In addition, the moisture clog will
occur at a greater temperature with decreasing permeability. However, inclusion of polypropylene
fibers within a UHPC design mix reduces the temperature corresponding to the melting point of
the fibers, 170°C (Klingsch, 2014; Ju et al., 2015).
2.4.3 Previous studies on fire induced spalling
Previous studies by researchers resulted in various measures for evaluating the explosive
spalling within HSC and UHPC. The investigations implemented preventive measures such as the
use of organic fibers, diminish the contents of cement and/or silica fume, and numerical models.
Existing research overwhelmingly utilizes organic fibers, such as polypropylene fiber to alleviate
explosive spalling, with only a handful utilizing pozzolanic constituents to reduce the silica fume
25

and cement contents, or the use of locally available material. Moreover, Klingsch (2014) and Ju et
al. (2015) developed numerical models to predict the explosive spalling behavior of UHPC with
compressive strengths greater than 150 MPa. Though, numerical models require sufficient facets
and validation, they provide an immense capability to efficiently determine whether explosive
spalling may occur (Dwaikat and Kodur, 2009; Klingsch and Frangi, 2011).
2.4.4 Methods for minimizing explosive spalling
A simple and cost-effective method to mitigate the occurrence of fire induced spalling is
with the inclusion of polypropylene fibers (Behloul et al., 2002; Bei and Zhixiang, 2016).
Polypropylene fibers melt at a temperature between 150-170°C creating pathways to release the
built-up pore pressure, and alleviate the rise of thermal gradients (Klingsch and Frangi, 2011; Ju
et al., 2015; Xiong and Liew, 2015). Moreover, various studies utilized mercury intrusion
porosimetry (MIP) and scanning electron microscopy (SEM) to confirm pathways created by the
melting of the polypropylene fibers and the total pore volume increasing between 2-7 times
following heating up to 600°C (Hager et al., 2013; Zheng et al., 2014b; Bei and Zhixiang, 2016).
The amount of polypropylene fibers required to avoid explosive spalling is dependent upon the
entire UHPC design mix and literature displayed values as low as 0.9 kg/m 3 to greater than 2.0
kg/m3 (So et al., 2014; Zheng et al., 2014a). Research conducted by Klingsch (2014) and Bei and
Zhixiang (2016) went further to assert the individual polypropylene fibers’ dimension is
paramount. The length must be of appropriate size to allow a continuous path for the vapor pressure
to follow and the diameter must be kept minimal. Lastly, polypropylene fibers diminish the
workability and strength of UHPC and must be implemented proficiently, and researchers advise
their use may prove to be ineffective for encapsulated structures (Behloul et al., 2002; Hertz, 2003;
Klingsch and Frangi, 2011; Aydin and Baradan, 2012).

26

An alternate method alleviate explosive spalling is to increase the tensile strength of the
concrete with the incorporation of steel fibers. Though studies have displayed the effectiveness of
steel fibers with HSC, the literature is not in total agreement with their use in UHPC (Klingsch,
2014; Kodur, 2014; Xiong and Liew, 2015; Bei and Zhixiang, 2016). Furthermore, implementing
a hybrid fiber (steel and polypropylene fibers) UHPC design mix displayed an improved
performance regarding the mitigation of the explosive spalling. Although polypropylene and steel
fibers serve an elementary method to reduce the explosive spalling hazard, researchers are
evaluating other lucrative techniques as well.
To utilize RPC in marine applications to avoid corrosion, while satisfying a tolerance to
elevated temperatures, Sanchayan and Foster (2016) conducted a study to determine if PVA fibers
can be utilized as an effective substitute for steel. By varying the different fiber contents, the
research developed a RPC design mix mitigating the explosive spalling risk through combining
PVA fibers along with a small amount of steel fibers, and oven drying the specimens. Although
inclusion of PVA fibers reduces the compressive strength, the stiffness comparatively remains the
same. Finally, Aydin and Baradan (2012), So et al. (2015) and Jang et al. (2016) successfully
utilized ternary pozzolanic materials consisting of GGBFS, fly ash and silica fume to avoid
explosive spalling.

27

2.5 Research gaps
As per the state of the art literature review, only few studies evaluating the residual-state
mechanical properties of UHPC have been conducted in the past five years. Furthermore, data
supporting the thermal properties of UHPC with increasing temperatures are severely sparse.
Although the residual-state mechanical properties are important, it is quite evident the high
temperature properties are lacking. To facilitate evacuation of occupants and allow the first
responders an opportunity to fight the fire, the structure must remain standing for the duration it
was designed for. Thus, the high temperature mechanical properties are imperative and the
following research gaps remain:
▪

No standard exists to evaluate the properties of a cementitious composite such as
UHPC/RPC or UHPFRC, at elevated temperatures.

▪

Thermal properties of UHPC with increasing temperature are deficient.

▪

Higher temperature mechanical properties of UHPC are gravely scarce.

▪

Detailed data and modelling of the fire induced spalling nature of UHPC is lacking.

▪

Impact to UHPC’s high temperature mechanical properties with varying steel fiber content.

▪

Impact to UHPC’s high temperature mechanical properties with a hybrid (steel and
polypropylene) fiber content.

▪

Impact to sustainable and lucrative design mixes of UHPC’s high temperature and residualstate properties.

28

Table 2.1: Temperature rise and effect on concrete, adapted from Klingsch (2014)
Temperature,
°C
20-100
105
120-150
150
270
>300
374
400-600
535
573
>700
710-720
>800
1200

Resulting change to concrete
Concrete loses free and physical water (begins to dry)
Start of hydrothermal reaction
C-S-H phase decomposes
C-S-H’s first peak
C-S-H’s second peak
Porosity and micro cracking drastically intensify
Complete evaporation of free and physically bound H2O
(critical point)
Detachment of calcium hydroxide, Ca(OH)2  CaO + H2O
Decomposition of Ca(OH)2 maximum point
ι-quartz transforms to β-quartz (reversible), leading to
pronounced expansion
Within cement paste and limestone decarbonation of calcium
carbonate; CaCO3  CaO + CO2
C-S-H’s decomposition, third peak
Full evaporation of all water chemically bound
Melting of concrete ensues

29

Table 2.2: Survey of UHPC/RPC and UHPFRC studies conducted at ambient conditions
Study/Investigation
Richard and Cheyrezy
(1995)
Reda et al. (1999)
Rahman, Molyneaux,
and Patnaikuni (2005)
Kamen et al. (2007)
Graybeal and Davis
(2008)
Yang et al. (2009)

Mechanical Property Tested

Type of Treatment

Strengths Recorded

Compressive strength of RPC

Pressure and thermal curing

Compressive strength 275 MPa

Compressive strength of
UHPC

Hot water bath (HW), ovencured (OV), and pressure applied
to fresh concrete

Compressive strength – 240 MPa
(OV with pressure of 80 MPa), 140
MPa (HW)

None – various mixtures

Compressive strength 155 MPa

Compressive strength of
UHPC
Compressive and bending
strength, and modulus of
elasticity of UHPFRC
Compressive strength of
UHPFRC
Compressive and flexural
strength of UHPFRC

Varying curing temperature

Temperature and RH
Cured in water, 20 and 90°C

YazÄącÄą et al. (2009)

Compressive strength of RPC

Steam and autoclave curing

YazÄącÄą et al. (2010)

Compressive and flexural
strength, and fracture energy of
RPC

Standard, autoclave, and steam
curing, and pressure treatment

Soliman and Nehdi
(2011)

Compressive strength of
UHPC

Wille et al. (2011)

Compressive strength of
UHPC

Varying curing conditions – RH,
and Temperature, and mixtures
to mitigate shrinkage
None – adjust the content of
cement, silica fume and highrange water reducer to maximize
its strength

30

Compressive strength 290 MPa
Bending strength 2 MPa
Modulus of elasticity 48 GPa
240 MPa
Compressive strength 160-180 MPa
Flexural strength 24 MPa
Compressive strength – 234 and 250
MPa
Compressive strength 403 MPa
Flexural strength 33 MPa
Fracture energy 7200 J/m2
Compressive strength 160 MPa
Plain UHPC – 190 MPa
UHPC with steel fiber – 200 MPa

Table 2.2 (cont’d)

Aydin and Baradan
(2013)

Compressive, flexural, and
splitting tensile strength,
modulus of elasticity, and
fracture energy of RPC and
alkali-activated cement
(CRPC)

Steam cured

Hassan et al. (2012)

Compressive, direct tensile
strength and modulus of
elasticity of UHPFRC

None

Magureanu et al.
(2012)

Compressive, flexural, and
splitting tensile strength, and
modulus of elasticity of UHPC

Temperature and RH

Prem et al. (2012)

Compressive, flexural, and
splitting tensile strength, and
modulus of elasticity of UHPC

Water and hot air curing

Zdeb (2013)

Compressive and flexural
strength, and modulus of
elasticity of UHPC

Steam, and water curing,
autoclave

Graybeal (2014)

Compressive, and direct tensile
strength, and modulus of
elasticity of UHPC

Temperature and steam curing

Voit and Kirnbauer
(2014)

Compressive, and splitting
tensile strength and fracture
energy of UHPC

Pressure, and thermal

31

Compressive strength – 215 MPa,
Flexural strength – RPC 42 MPa,
ARPC 35 MPa
Splitting tensile strength – 19 MPa
Modulus of Elasticity – RPC 114
GPa, ARPC 84 GPa
Fracture energy – RPC 14 kJ/m2,
ARPC 16 kJ/m2
Compressive strength – 151 MPa
Direct tensile strength – 9 MPa
Modulus of Elasticity – 46 GPa
Compressive strength – 180 MPa
Flexural strength – 34 MPa
Splitting tensile strength – 13 MPa
Modulus of Elasticity – 52 GPa
Compressive strength – 180 MPa
Flexural strength – 44 MPa
Splitting tensile strength – 24 MPa
Modulus of Elasticity – 40 GPa
Compressive strength – 315 MPa
Flexural strength – 27 MPa
Modulus of Elasticity – 50 GPa
Compressive strength – 182 MPa
Direct tensile strength – 9 MPa
Splitting tensile strength – MPa
Modulus of Elasticity – 53 GPa
Compressive strength – 345 MPa,
Splitting tensile strength – 24 MPa
Fracture energy – 2500 J/m2

Table 2.2 (cont’d)

Xiao et al. (2014)

Compressive, and flexural
strength of UHPC

Steam cured at 90°C, and with
special mixtures

Gesoglu et al. (2015)

Compressive strength of
UHPFRC

Ghafari et al. (2015)

Compressive strength of
UHPC

Water or steam cured, with
varying mixtures
None – adjust the mixture
contents to develop an ecoefficient UHPC while
maximizing its strength

Graybeal (2015)
Kusumawardaningsih
et al. (2015)
Alkaysi (2016)

Peng et al. (2016)

Wang and Gao (2016)
Zdeb (2017)

Compressive strength of
UHPC
Tensile strength of UHPC and
UHPFRC
Compressive and direct tensile
strength of UHPC
Compressive strength, splitting
tensile strength and fracture
energy of UHPC (standard,
and recycled steel fiber)
Compressive and flexural
strength of UHPC
Compressive, and flexural
strength of RPC

Compressive strength
UHPC 1.4% superplasticizer –
151 MPa
UHPC high w/c – 100 MPa
Flexural strength
UHPC 4% steel fiber –
32 MPa
UHPC 1% steel fiber –
16 MPa
Compressive strength 150 MPa

170 MPa

Steam curing

200 MPa

Varying treatment

UHPC 4 MPa, UHPFRC 7 MPa

None – only utilized varying
types of silica fume and powder,
and cement

Compressive strength 186 MPa
Direct tensile strength 9 MPa

None – only utilized varying
types and content of steel fiber
Steam curing at 90°C
Steam and autoclave curing

32

Average – (1) compressive strength
141 MPa, (2) splitting tensile
strength 9 MPa, and (3) fracture
energy 5440 J/m2
Compressive strength 110-230 MPa
Flexural Strength 20-45 MPa
Compressive strength 200 MPa
Flexural strength 19 MPa

Table 2.3: Survey of UHPC/RPC and UHPFRC studies conducted at elevated temperatures
Study/Investigation
Hertz (1984)

Study
Objectives/Detail
Explosive spalling
of ultra high
strength concretes

Features and
Methodology
▪ Observation of
explosive spalling
with high dense
concretes
▪ Heating rate as low
as 1ÂşC/min

Felicetti et al.
(2000)

Mechanical
properties of HPC
and UHPC (hot
and residual) are
investigated

▪ Residual and hotstate direct tension,
Young’s modulus
and fracture energy
▪ 20-600°C, with a rate
of (1) residual,
1ÂşC/min, and (2) hotstate, 2ÂşC/min

Khoury (2000)

Mechanical
properties of HSC,
UHPC are
investigated at
elevated
temperatures

▪ Residual
compressive strength
▪ 20-700°C, with rate
of 2ÂşC/min

33

Observations/Conclusions
▪ At room temperature
compressive strength
170 MPa
▪ Explosive spalling
occurred with lower
moisture content and
rate of heating as
compared to
conventional concretes
▪ UHPC at room
temperature
compressive strength
160 MPa, tensile
strength 10-11 MPa
▪ The hot-state and
residual strengths are
very similar
▪ At room temperature
compressive strength
115-170 MPa

Strengths/Drawbacks
▪ Explosive spalling is solely
due to the low permeability,
i.e. high density, achieved
through addition of silica
fume
▪ Further research will be
conducted to better
understand behavior
▪ The Young’s modulus is
lower with respect to the
hot-state tests
▪ The fracture energy of
UHPC experiences a
dramatic increase up to 250300°C
▪ The test method exhibits an
important role
▪ Mechanical properties fire
performance are dependent
upon the design mix
▪ Be mindful of the ratio
consisting of CaO/SiO2

Table 2.3 (cont’d)
Behloul et al.
(2002)

Mechanical and
thermal properties
of UHPC (stressed
hot-state and
residual) are
investigated at
elevated
temperatures

▪ Compressive, and
direct tensile
strengths, elastic
modulus, thermal
conductivity, specific
heat, and coefficient
of thermal expansion
(CTE)
▪ 20-750°C, with rate
of 2ÂşC/min

Schrefler et al.
(2002)

Investigate the
overall behavior of
UHPC and
conventional
concretes at
elevated
temperatures with
respect to
developing a
model

▪ Model encompasses
the thermal, hydral
and mechanical
behavior
▪ 20-1000°C, with a
rate of 2ÂşC/min

34

▪ At room temperature
compressive strength
140-200 MPa, and
tensile strength 8-12
MPa, elastic modulus
45-50 GPa
▪ Thermal conductivity
and specific heat
steadily decline with
increasing
temperatures, however
experience a sudden
increase at 200°C
▪ CTE increases linearly
up to ~550°C, then a
sudden spike closer to
575 C, followed by a
decline, and then an
increase around 800°C
▪ UHPC at room
temperature
compressive strength
118 MPa (input data
for model)
▪ Mechanical and
chemical
(microstructure)
behavior inputted via
experimental testing

▪ Residual compressive
strength, and elastic
modulus mechanical
properties outperformed the
hot-state values
▪ Hot-state direct tensile
strengths produced slightly
higher values than their
residual counterparts
▪ Based off the formulation of
UHPC, no spalling was
observed

▪ Two examples are provided
within the report to display
the validity of said model

Table 2.3 (cont’d)
Hertz (2003)

Explosive spalling
limits of concrete
at elevated
temperatures

▪ Consolidation of
information and
recommendations
pertaining to
explosive spalling
phenomenon based
off author’s
experience and
knowledge

▪ Concrete will have a
lesser chance of
spalling if silica fume
is not used, coupled
with if the moisture
content is 3-4%

Heinz et al. (2004)

Explosive spalling
behavior of UHPC
of varying content
of steel and PP
fibers, at elevated
temperatures
Mechanical
properties of RPC
are investigated
and compared to
with HPC and
NSC

▪ Cylindrical
specimens
▪ 20-1050°C,
standardized ETK
curve (DIN 4102-2)

▪ All specimens without
PP fiber, exploded

▪ Residual compressive
strength tests
▪ Thermogravimetric
analysis
▪ Fire exposure
(500°C) durations
were altered (0-120
min, 30 min
increments)

▪ RPC at room
temperature
compressive strength
75 MPa
▪ The RPC’s fire
endurance is greater
versus conventional
concretes
▪ The RPC’s displayed a
larger residual
compressive strength

Liu and Huang
(2009)

35

▪ Explosive spalling is
correlated to 374 C,
approximately the pertinent
temperature of steam
▪ Explosive spalling can be
alleviated via either thermal
cracks, and/or PP fiber, if
and only if there is no
immediate external
deterrent
▪ Steel (3.5%) and PP
(0.66%) fiber by volume
produce the best results for
fire resistance
▪ RPC displays greater
resistance when affected by
increasing temperature
▪ RPC experienced explosive
spalling at 790°C

Table 2.3 (cont’d)
Burke (2011)

The compressive
strength of UHPC
is studied with
varying PP fiber
content

▪ Conducted
compressive cylinder
and cube specimen
tests
▪ 20 to 600°C, with
rate of 5ÂşC/min

▪ At room temperature
compressive strength, (1)
cylinder – 137-152 MPa,
and (2) cube – 140-180
MPa
▪ PP fibers alleviated
explosive spalling

Ju et al. (2011)

The
thermophysical
properties of RPC
are evaluated at
elevated
temperatures

▪ Thermophysical
properties, thermal
conductivity, thermal
diffusivity, specific
heat capacity and
linear thermal
expansion with four
varying steel fiber
volume
▪ 20 to 250°C (in 50°C
increments), with rate
of 5ÂşC/min

Tai et al. (2011)

Study the
compressive
stress-strain
relationship of
RPC with varying
steel fiber content
at elevated
temperatures

▪ Residual compressive
strength, and stressstrain relationship
▪ 20-800°C, with rate
of 2ÂşC/min

▪ Minimal mass loss at
250°C
▪ Only specific heat
capacity increases with
higher temperatures
▪ The steel fiber content
effects the specific heat
capacity and thermal
diffusivity values
▪ RPC’s thermal
conductivity, and specific
heat capacity are lower
than HSC
▪ The compressive strength
increases slightly up to
300°C, soon followed by
a decrease with
increasing temperatures
▪ The peak strain increases
with higher temperatures

36

▪ UHPC’s residual
compressive strength
with respect to increasing
temperature is like other
concretes
▪ The exposure duration of
the heat affected the cube
specimens, but not the
cylinder specimens
▪ Due to the differing
microstructure, RPC’s
thermophysical
properties vary as
compared to HSC

▪ RPC with 1% steel fiber
content behave
differently than with 2
and 3% steel fiber
content
▪ The slope decreases of
the stress-strain curve
with increasing
temperatures

Table 2.3 (cont’d)
Aydin and
Baradan (2012)

The mechanical
properties of RPC
with a Portland
cement
replacement at
elevated
temperatures is
investigated

▪ Cement is replaced
with alkali-activatedslag (ARPC), and
results are compared
with cement based
RPC (CRPC)
▪ Residual flexural,
compressive strength
and toughness
▪ 20-800°C, with rate
of 10°C/min

Peng et al. (2012)

The residual
mechanical
properties of RPC
with varying
content of steel
and PP fibers and
the explosive
spalling behavior
were studied

▪ Residual
compressive,
splitting tensile
strength and fracture
energy
▪ 20-600°C, with rate
of 10°C/min

37

▪ The CRPC strength
changes very little up
to 300°C and
undergoes explosive
spalling beyond 300°C
▪ The ARPC’s strength
steadily decreases with
decreasing
temperature
▪ ARPC can withstand
temperatures up to
800°C
▪ At room temperature
compressive strength
70-100 MPa, splitting
tensile strength 13-16
MPa, and fracture
energy 12,000-15000
J/m2
▪ The compressive
strength increases up
to 400°C, while the
splitting tensile
strength and fracture
energy only up to
200°C

▪ CRPC’s increasing
temperature resistance is
greater than ARPC up to
300°C
▪ Beyond 300°C, ARPC is
better
▪ ARPC with 20% silica fume
did not experience explosive
spalling

▪ The use of PP fiber is a
method to avoid explosive
spalling
▪ RPC with water-cement
ratio from 0.16-0.20 has a
higher chance to experience
explosive spalling
▪ Increase of mechanical
properties between 200400°C may be attributed to
hydration of the
cementitious materials, but
further research is required

Table 2.3 (cont’d)
Study the
compressive
strength of UHPC
with varying PP
fiber, and with and
without steel bar
reinforcement at
elevated
temperatures
Zheng et al. (2012) Study the
compressive
stress-strain
relationship of
RPC with varying
steel fiber content
at elevated
temperatures

▪ Stressed at 30% of
the ultimate strength,
hot-state compressive
strength
▪ 20-400°C

▪ Many cracks were
observed with
increasing temperature
up to 300°C
▪ Specimen heated to
400°C experienced
explosive spalling

▪ Adding PP fiber, 2 kg/m3,
alleviated the cracks and
spalling

▪ Residual compressive
strength, stress-strain
relationship and
elastic modulus
▪ 20-900°C, with rate
of 4ÂşC/min
▪ Specimens are heat
treated

▪ Minimum compressive
strength reached at 700°C
▪ Very little changes with
respect to the stress strain
relationship up to 300°C;
following 300°C, the curves
begin to flatten

Abdul-Hussain
(2013)

▪ Residual
compressive, flexural
and splitting tensile
strength on cubes,
cylinders and prisms
▪ 20-600°C, with slow
heating rate

▪ The compressive
strength increase with
initial temperature
increase, however
begin to decrease soon
after
▪ The peak strain
increases with higher
temperatures
▪ At room temperature
(with increasing PP
fiber content)
compressive strength
97-122 MPa, flexural
strength 6-23 MPa,
and splitting tensile
strength 5-15 MPa

Ye et al. (2012)

Various
mechanical
properties of fiber
RPC, and
explosive spalling
were studied

38

▪ All mechanical properties
tested decrease at
temperatures above 400°C
▪ Tensile strengths suffered
the most following exposure
of 600°C

Table 2.3 (cont’d)
Jagannathan
(2013)

To study and
understand the CS-H in concrete at
elevated
temperatures

Zheng et al. (2013) Various
mechanical
properties of steel
fiber RPC were
studied

▪ Utilizing
experimental data,
compare the C-S-H
in UHPC versus
conventional
concretes

▪ Residual compressive
and tensile strength
tests
▪ 20-800°C, with rate
of 5ÂşC/min

39

▪ At the micro-level, the
drastic difference in
stiffness between the
C-S-H phase and steel
fiber, cause the
packing density of the
C-S-H to lessen, and
capillary porosity
within the matrix is
formed
▪ The critical
temperatures for
explosive spalling are
from 260-520°C, and
the risk rises with
larger dimensions
▪ The compressive
strength decreases at
100°C, rises up to
500°C, and followed
by a decline
▪ The tensile strength
decreases at 200°C,
rises only up to 300°C,
and followed by a
decline

▪ UHPC’s C-S-H phase is
quite durable up to 400°C,
and dramatically firmer than
conventional concretes

▪ 2% steel fibers can prevent
RPC to undergo explosive
spalling
▪ Based off study
implementation of codes for
RPC design is conservative
for temperatures between
400-800°C, and not
recommended for
temperatures under 400°C

Table 2.3 (cont’d)
Bashandy (2013)

Various
mechanical
properties of fiber
RPC, with varying
cement and steel
fiber content, and
differing fire
exposure times

Hager et al. (2013)

The compressive
strength and
modulus of
elasticity of RPC
are studied with
varying PP fiber
content

▪ Residual compressive
(cube specimens of
100 x 100 x 100 mm)
and splitting tensile
strength (cylinder
specimens of 200
mm length and 100
mm diameter)
▪ 20-500°C, with a rate
of 10°C/min
▪ Heating durations of
2 and 4 hours
▪ Residual compressive
strength tests on
prism specimens 40 x
40 x 80 mm
▪ 20 to 1000°C, with
rate of 0.5ÂşC/min
▪ Spalling tests to
600°C, with rates of
0.5, 1, 2, 4 and
8ÂşC/min
▪

40

▪ At room temperature
compressive strength
85-150 MPa, splitting
tensile strength 7-11
MPa
▪

▪ Increasing the heating
duration, decreases the
residual strengths
▪ Increase in cement improves
only the at room
temperature strengths
▪ Steel fibers enhance the
strengths only up to 150°C

At room temperature
compressive strength 180220 MPa, and elastic
modulus 48 GPa

▪ 2.0 kg/m3 PP fibers
prevented explosive spalling
▪ PP fibers do not drastically
alter the relative
compressive strength as a
function of temperature
▪ Modulus of elasticity is not
affected by the addition of
PP fibers
▪ Porosity increases from 0.02
to 0.14 cm3/cm3, when
temperature is heated to
600°C

Table 2.3 (cont’d)
Ju et al. (2013)

Wang et al. (2013)

In order to better
understand the
spalling behavior
of RPC, the
microstructure
along with the pore
pressure were
studied at elevated
temperatures
The study
investigated the
thermomechanical
(TMA) analysis of
UHPC with
varying
dimensions and
content of steel
fiber

▪ Plain RPC specimens
▪ The pore
▪ 20-350°C
characteristics
▪ Mercury intrusion
increase with higher
porosimetry (MIP),
temperatures
and scanning electron A model proposed to help
microscopy (SEM)
analyze the pore pressure
were utilized
and explosive spalling

▪ Explosive spalling is mainly
attributed to the lack of
pathways to allow release of
vapor pressure

▪ Coefficient of
thermal expansion
(CTE),
▪ 20-800°C, with rate
of Âą3ÂşC/min

▪ Both UHPC types tested, (1)
extended almost linearly up
to 570°C, (2) severely
contracted above 700°C,
and (3)
▪ Copious amount of cracks
exposed via SEM, in
vicinity of phase interfaces
due to thermal stresses
caused by differences in
CTE

41

▪ α-β quartz
transformation –
570°C, reversible
phase transformation
▪ Dihydroxylation
transition of C-S-H –
above 700°C,
irreversible phase
transformation

Table 2.3 (cont’d)
Canbaz (2014)

Investigate the
compressive strength
of pressurized and
accelerated curing
RPC with varying PP
fiber content, at
elevated temperature

▪ Residual
compressive strength
and UPV tests
▪ 20-900°C, with rate
of 10°C/min

Klingsch (2014) Investigate the
explosive spalling
behavior of UHPC at
elevated temperatures

▪ Vary the silica fume
content and employ
different heating
rates

So et al. (2014)

▪ Residual
compressive and
flexural strength
▪ 20-1095 C, with rate
of 30°C/min (KSF
2257-1, Korean
standard)

Mechanical
properties and
thermogravimetricdifferential thermal
analysis (TG-DTA)
of RPC with varying
cement, silica fume
and PP fiber at
elevated temperatures

42

▪ At room temperature
compressive strength, 200
MPa (0% PP), and 165
MPa (1% PP)
▪ RPC with no PP fiber
relatively well up to
400°C, however poor
performance at high
temperatures up to 900°C
▪ Addition of PP fiber
improved the RPC’s
performance up to high
temperatures
▪ At room temperature
compressive strength 150
MPa
▪ Hydrothermal model is
presented, which is
validated via experimental
testing
▪ At room temperature
compressive strength 106150 MPa, and flexural
strength 17-27 MPa
▪ The relative residual
compressive and flexural
strength at 1095 C was
between 0.15-0.29 and
0.20-0.33, respectively

▪ 1% of PP fibers is
suggested to avoid
explosive spalling and to
maintain its strength at
elevated temperatures up
to 900°C
▪ RPC with no PP fiber
was not able to withstand
temperatures higher than
900°C

▪ Explosive spalling will
occur if the pore pressure
exceeds the concrete’s
tensile strength
▪ Employing a low rate of
heating can still
contribute to explosive
spalling, due to UHPC’s
dense matrix
▪ RPC is more at risk of
explosive spalling with
(1) increasing cement
content and silica fumecement ratio, (2) value of
the pore-volume
proportion, and (3) PP
fiber content of at least
2.0 kg/m3

Table 2.3 (cont’d)
Zheng et al.
(2014a)

Study the
mechanical
properties of RPC
with steel and PP
fiber at elevated
temperatures and
varying fire
exposure durations

▪ Residual compressive
and tensile strength
▪ 20-800°C, with rate
of 5ÂşC/min

43

▪ At room temperature
compressive strength,
135-163 MPa, tensile
strength, 5.5-7.5 MPa
▪ Compressive strength
(PP and steel fiber)
decrease at 100°C, and
increases up to 500°C;
followed by a decrease
at 600°C and above
▪ Compressive strength
increases up to 400°C
with prolonged fire
exposure
▪ Tensile strength (PP
and steel fiber)
decrease at 200°C, and
increase at 300°C;
followed by a decrease
with higher
temperatures

▪ The compressive and tensile
strengths attain an initial
increase, soon followed by a
degradation of strength
▪ PP fibers provide avenues
for the vapor pressure to
release, and avoid explosive
spalling

Table 2.3 (cont’d)
Zheng et al.
(2014b)

Various
mechanical
properties of fiber
RPC, the
microstructure and
explosive spalling
were studied
Objective:
▪ Evaluate the
effects of
differing PP
fiber content
▪ Measure
porosity,
thermal
conductivity

▪ Compressive strength
tests, 70 x 70 x 70
mm cubes and 70 x
70 x 220 mm
prismatic specimens
▪ Tensile strength tests,
dog-bone shaped
▪ 20 to 800°C, with
rate of 5ÂşC/min

44

▪ Room temperature
compressive strength,
(1) cube 107-137
MPa, and (2) 92-138
MPa, (3) tensile
strength 4.5-5.75 MPa
▪

▪ 2.73 kg/m3 PP fibers
prevented explosive spalling
▪ Cube compressive strength
performed the best
▪ Compressive and tensile
strength fluctuated,
decreased at 100 and 200°C,
respectively, increased from
200 to 600°C, and 200 to
300°C, respectively, and
decreased above 600°C, and
400°C, respectively
▪ Thermal conductivity
sharply decreased up to
400°C, and slowly beyond
400°C

Table 2.3 (cont’d)
Ju et al. (2015)

The residual
mechanical
properties and the
spalling behavior
of RPC were
investigated
Objective:
To vary the PP
fiber content (by
volume)

So et al. (2015)

▪ Study the
mechanical
properties of a
more
economical
mixture of
RPC at
elevated
temperatures

▪ Compressive
strength, split tensile
strength, and flexural
strength (up to
350°C), with rate of
3ÂşC/min
▪ The spalling
performance and
details up to 600°C,
with rate of
4.8ÂşC/min

▪ The contents of silica
fume and cement are
drastically reduced
▪ In return, developed
a mixture that
contains silica fume,
blast furnace slag and
fly ash
▪ Residual strength
tests at 20-1000°C
with a rate of
10°C/min

45

▪ Room temperature
compressive strength
135-160 MPa
▪ Thermomechanical
properties increase up
to 300°C, and decrease
following 300°C
▪ 0.9% PP fiber is critical
dosage for ideal
residual strengths
▪ Mass loss variation
with respect to
temperature range
▪ At 280-290°C,
moisture begins to
migrate towards the
outside
▪ Room temperature
compressive strength
140-190 MPa, tensile
strength 15-22 MPa,
and flexural strength
35-42 MPa
▪ Modified RPC residual
compressive, flexural,
and splitting tensile
strengths were 36%,
22%, and 29% of the
ambient strength,
following exposure to
1000°C

▪ 0.9% PP fiber dosage is
optimum for residual
mechanical properties
▪ Above 150°C, mass loss
(evaporation of water)
increases very rapidly
▪ Moisture migration occurs
around 280-290°C

▪ The modified RPC
outperformed the more
traditional RPC mixture
▪ 30% fly ash, 10% blast
furnace slag, and with the
lowest silica fume content
proved to be the best
combination
▪ Substantial strength loss
experienced from 400800°C

Table 2.3 (cont’d)
Way and Wille
(2015)

Xiong and Liew
(2015)

▪ Investigate the
effect of
chemical
degradation
(DSC/TGA) of
the mechanical
properties of
UHPFRC

▪ Residual compressive
strength and stressstrain relationship
▪ 20-900°C with a rate
of 0.1ÂşC/min

▪ At room temperature
206-233 MPa (with
thermal treatment
▪ Oven drying the
specimens provided an
increase in
compressive strength
and elastic modulus
over the control
(90°C)

Investigate
mechanical
properties of
UHPC with
varying fiber type,
dosage, heating
rate and curing
condition at
elevated
temperatures

▪ Residual compressive
strength and elastic
modulus
▪ 20-800°C, with rates
of 5 and 30°C/min

▪ At room compressive
strength 160-185 MPa
▪ Steel fibers are not
effective to avoid
explosive spalling
▪ Compressive strength
experiences an
increase at 200°C –
attributed to water
removal; not seen with
elastic modulus

46

▪ UHPFRC experienced a rise
in strength at 135 and
200°C, followed by a
decrease
▪ The elastic modulus
decreased following 300°C
and above
▪ The peak strain increased
with increasing temperature
▪ The thermomechanical
analysis provided reasoning
to decrease in strength at
respective temperatures
▪ PP fiber with dosages of
0.1% is effective to avoid
explosive spalling
▪ The compressive strength
and elastic modulus reduce
more rapidly with higher
heating rates
▪ Curing conditions do not
effect the compressive
strength and elastic modulus

Table 2.3 (cont’d)
Zheng et al. (2015) Mechanical
properties at hotstate of RPC with
steel fiber at
elevated
temperatures

▪ Hot-state
compressive strength,
stress-strain
relationship, elastic
modulus and energy
absorption
▪ 20-800°C, with rate
of 5ÂşC/min

Ali et al. (2016)

Investigate the use
of waste glass
powder to enhance
the mechanical
properties at
elevated
temperature of
HSC and compare
to RPC

▪ Residual
compressive, and
flexural strengths,
modulus of elasticity
and stress strain
relationship
▪ 20-800°C with rate
of 13ÂşC/min

Bei and Zhixiang
(2016)

Investigate the
explosive spalling
behavior of UHPC
(with varying fiber
mix and content) at
elevated
temperatures

▪ Place UHPC
specimen in a
preheated furnace
(1000°C) for a short
duration, followed by
sudden exposure to a
25 C environment

47

▪ At room temperature
compressive strength
142-165 MPa
▪ With increasing
temperatures, (1) the
thermal expansion,
peak strain, and
energy absorption (up
to 400°C) escalate, (2)
compressive strength,
and Gismodine
progressively
decreases
▪ RPC (HSC with waste
glass powder, no silica
fume, HSC-GP) at
room temperature
compressive strength
142 MPa (81 MPa),
modulus of elasticity
52.8 GPa (40.4 GPa)
and flexural strength
6.7 MPa (5.9 MPa)
▪ At room temperature,
compressive strength
168-198 MPa
▪ Steel fiber merely
postpones the time to
spall

▪ Increasing the steel fiber
content will not improve the
compressive strength,
however will intensity the
ductility
▪ Following temperatures up
to 800C C-S-H decomposes
(pores and cracks appear),
and C2S and C3S formulate
▪ The mechanical properties
dilapidation is due to the
pores and microcracks
▪ RPC experiences a sudden
decline in compressive and
flexural strengths with
increasing temperature,
while HSC-GP experiences
a steady decline
▪ RPC (HSC-GP) experienced
partial spalling at 400°C
(500°C) and full spalling at
and above 500°C (800°C)
▪ PP fiber content greater than
0.20%, and length of 12 and
19 mm prevent explosive
spalling

Table 2.3 (cont’d)
Li and Liu (2016)

Investigates the
tensile properties
of RPC with steel
and PP fibers at
elevated
temperatures

▪ Residual flexural and
direct tensile strength
▪ 20-900°C, with rate
of 4ÂşC/min

Jang et al., 2016

The residual
mechanical
properties and the
spalling behavior
of RPC were
investigated

▪ Compressive,
flexural and splitting
tensile strength
▪ 20 to 1000°C, with a
rate of 10°C/min
▪ Thermogravimetric
and microstructure
analysis

Objective:
▪ To vary the
fiber content
(by volume)
▪ Alter the
admixture with
a combination
of ternary
pozzolanic
materials

48

▪ At room temperature,
flexural strength 22-31
MPa, and direct tensile
strength is between 68 MPa
▪ The flexural and direct
tensile strength
experience a steady
decline with
increasing temperature
▪ Room temperature
compressive strength
150 MPa, and flexural
strength 25-32 MPa,
▪ Modified RPC
contained only 765
kg/m3 of cement, and
an increase in
compressive and
flexural strength as a
function of
temperature
▪ Compressive strength
increased up to 400°C,
and then decreased
▪ Flexural strength
increased up to 200°C,
and then decreased

▪ Only the steel fibers can
increase the tensile strength
and the tensile strength
increases with higher steel
fiber content

▪ A decrease in table flow as
the fiber content was
increased
▪ PP fiber is effective to
inhibit explosive spalling,
so long as the volume is
0.2% of the binder content
▪

Table 2.3 (cont’d)
Peng et al.
(2016)

Study the explosive
spalling behavior and
ambient condition
mechanical properties
of UHPC with varying
types of steel fiber at
elevated temperature

▪ Utilize standard
steel fiber and
sustainable
(recycled) steel
fiber to reinforce
UHPC
▪ 20-800°C, rate of
10°C/min (for
explosive spalling
tests only)

Sanchayan and
Foster (2016)

The residual and hot
state mechanical
properties and the
spalling behavior of
RPC (6 different
mixes) were
investigated

▪ Residual
compressive
strength, and
elastic modulus,
with a rate of
5ÂşC/min
▪ Hot state elastic
modulus, free
thermal strain and
transitional
thermal creep,
with a rate of
5ÂşC/min

Objective:
If PVA fibers would
alleviate the explosive
spalling behavior

49

▪ At room temperature
compressive strength 130-154
MPa, splitting tensile strength
7-12 MPa, and fracture
energy 260-13,828 J/m2
▪ The recycled steel fiber
compressive strength ranged
from the lowest to about the
average of compressive
strengths
▪ While the splitting tensile
strength, and the fracture
energy of the recycled steel
fiber proved to be the most
favorable
▪ Room temperature
compressive strength 134-170
MPa
▪ Utilizing PVA fibers alone
reduced the compressive
strength, elastic modulus
remained constant
▪ Elastic modulus and
ultrasonic pulse velocity drop
following exposure to 300°C
▪ Transitional thermal creep
very high above 250°C
▪ Surfeit amount of mass loss
for tests conducted to 300,
and 400°C

▪ The steel fiber with the
least diameter proved to
be the most effective to
avoid explosive spalling
▪ Additionally, the
recycled steel fiber
UHPC provided
explosive spalling relief
as well

▪ 1% of steel and PVA
fibers design mix
provided the best results
▪ Increase of compressive
strength up to 300°C,
followed by a drop
▪ Hot state elastic
modulus of RPC 1%
steel and PVA fibers
decreased almost
linearly

Table 2.3 (cont’d)
Nazri et al. (2017)

The residual
mechanical
properties and the
spalling behavior
of UHPFRC were
investigated

▪ Residual compressive
strength on cube
specimens of 40 x 40
x 50 mm
▪ 20-600°C with a rate
of 15ÂşC/min

50

▪ At room temperature
compressive strength
157-161 MPa
▪ UHPFRC experienced
explosive spalling at
temperatures above
400°C
▪ Did not contain any
PP fiber

▪ Due to undergoing
explosive spalling only
conducted compressive
strength up to 200°C with
very little reduction in
strength
▪ Fire Propagation Index of
0.0, under BS476 Part 6,
Class 1 fire rating under
Part 7

Figure 2.1: Relative compressive strength versus temperature as per literature from the
past 3 years

(a) Relative direct tensile strength
Figure 2.2 Relative tensile strength versus temperature as per literature
51

Figure 2.2 (cont’d)

(b) Relative splitting tensile strength

(c) Relative flexural strength

52

Figure 2.3 Relative elastic modulus versus temperature as per literature from the past 2
years

53

CHAPTER THREE
3. EXPERIMENTAL STUDIES ON COMPRESSIVE STRENGTH AND EXPLOSIVE
SPALLING OF UHPC AT ELEVATED TEMPERATURES
3.1 General
Data pertaining to the mechanical properties of UHPC at elevated temperatures are needed
to evaluate concrete structural members’ fire response. Thus, grasping the degradation of
compressive strength, stress-strain relationship, and elastic modulus with increasing temperatures,
enable designers to achieve the fire resistance requirements dictated by code provisions. As
presented in the previous chapter, it is apparent data on high temperature mechanical properties
and temperature induced explosive spalling behavior of UHPC are lacking.
To fulfill this integral gap, a comprehensive experimental study was undertaken to investigate
UHPC’s compressive strength, stress-strain relationship, and elastic modulus with increasing
temperatures. From the collected data, an empirical relation was formulated showing the variation
of the 3 mechanical properties of UHPC with respect to temperature (20-700ÂşC), and highlights
pertaining to explosive spalling. The following sections will impart the experimental design and
details, procedures followed, and the findings will be discussed.
3.2 Design of mechanical property experiments
The experimental study evaluated the unstressed mechanical properties of 3 different
design mixes of UHPC. Each UHPC design mix incorporated a varying mixture of fiber(s), to
investigate their influence on the respective mechanical properties, and effect of mitigating the
explosive spalling occurrence. Specifically, two mixes of UHPC with varying content of steel
fibers and one UHPC mix with a combination of steel and polypropylene fibers (hybrid) were

54

tested. The cylinders were tested under compressive loads at various temperatures (20, 100, 200,
300, 400, 600 and 700°C) in a elevated temperature state as opposed to testing under conventional
residual-state regime.
For comparative purposes, in addition to the UHPC 4 other conventional types of concretes, NSC,
NSC with steel fiber (NSC-SF), HSC, and HSC with polypropylene fibers (HSC-PP) were also
tested (at similar temperature ranges). Additionally, in the absence of an ASTM standard dictating
guidelines for material properties at elevated temperatures, to regulate the testing procedure a
combination of published standards was followed. The mechanical property tests were carried out
as outlined in each respective ASTM standard, coupled with the heating regiment delineated in
RILEM Standard (RILEM Technical Committee 200-HTC, 2007) were applied.
3.3 Mechanical properties tests
The mechanical properties investigated as part of this study were compressive strength,
stress-strain relationship, and elastic modulus. Table 3.1 displays the type and number of tests
conducted for each respective type of concrete.
3.3.1 Batch mix proportions
Four batches of UHPC were cast to prepare needed test specimens for evaluating the
mechanical properties. Batch mix proportions for the UHPC mixes have been developed by Metna
Co. Furthermore, 4 conventional concrete design mixes were cast as well, NSC, NSC-SF, HSC,
and HSC-PP. All concrete types evaluated contained Portland cement, Type I, but with varying
quantities. The conventional concrete mixes incorporated 9.5-millimeter (mm) limestone-coarse
aggregate, and natural sand. The HSC mixes contained nearly 2.5 times the cement as the NSC
mixes, and silica fume. All 3 UHPC design mixes contained 2.38 mm limestone-coarse aggregate,
silica and natural sand, over 5 times the silica fume as compared to the HSC mixes, limestone
55

powder, and a superplasticizer, Chryso. The two UHPC design mixes with steel fibers are UHPC
A, B, and D. UHPC A and B (A & B) are the same mixture proportions, however due to logistical
constraints at the time of casting, two concrete batches were made. UHPC D has the same
components but in lesser quantity, as compared to UHPC A & B. Namely, there is less steel fiber
content in UHPC D, vice UHPC A & B. UHPC F is the hybrid design mix with steel and
polypropylene fibers, and the other components are very similar as to design mix UHPC D.
Furthermore, UHPC D and F were casted in Metna Co. laboratory, and therefore in a more
controlled environment. The mix proportions in each type of UHPC design mix, and conventional
concretes tested are given in Table 3.2, and Table 3.3, respectively. Additionally, the conventional
concretes 28-day compressive strength range from 29-110 MPa, while UHPC’s from 132-182
MPa. The 28-day compressive strengths, along with the densities of all concretes tested, are
provided in Table 3.4.
3.3.2 Test specimens
The test specimens comprised of 75 mm in diameter by 150 mm in length cylinders as
shown in Figure 3.1 and were cast using paper molds. Paper molds were used for casting UHPC
cylinders to facilitate the placement of type K thermocouples in the cylinder for temperature
measurement. Prior to casting, each paper mold is drilled from the side at mid-height and
thermocouples are inserted from the side exactly at the center (75 mm from the side) of the
cylinder. However, center thermocouples were not employed with the conventional concretes,
thereby allowing the use of reusable plastic molds. When each respective concrete mixture was
ready, the molds were filled with concrete and placed on a shaking table at high frequency for
about 2 min.

56

The casted specimens are covered with plastic sheets to minimize moisture loss and left to set
slowly for 24 hours. After 24 hours, the hardened UHPC specimens are demolded and immediately
placed in a steam chamber for curing at 90°C for 48 hours. After steam curing, the specimens are
air dried in Michigan State University Civil Infrastructure Laboratory (MSU CIL) for at least 90
days before conducting high-temperature tests.
Since the cylinder’s top and bottom surfaces are originally rough from casting, a table saw (Figure
3.2) is utilized to create a smooth and leveled finish on both end surfaces (Figure 3.1). This smooth
finished surface was performed for high-temperature testing since the available capping material
melt at low temperatures of less than 127ÂşC. However, if an excellent level finish is not attained
using the electrical saw, the results for compressive strength will not be accurate and this can be
observed from the failure mode of the specimen as described in ASTM-C39/C39M (ASTM
C39/C39M, 2016). The smooth leveled surface is achieved through slowly cutting the specimen,
and repeatedly smoothing the surface until it achieves a leveled finish. The finished surface is
finally checked utilizing a magnetic level.
For each heated UHPC specimen in the furnace, three thermocouples are utilized to monitor
temperature rise. Figure 3.1(a) depicts the thermocouple locations when the specimen is placed
within the electrical furnace. Thermocouple (1) monitors the furnace temperature, nearest to the
specimen, (2) the specimen’s surface temperature and (3) specimen’s core temperature. In
addition, the electrical furnace has its own temperature gauges which are compared with the
thermocouple readings. Contrary to UHPC specimens, the four conventional concretes did not
have a center thermocouple, and the surface thermocouple is utilized to monitor the temperature
rise. The arrangement of thermocouples for the conventional concretes is shown in Figure 3.1(b).

57

3.3.3 Test apparatus
The test apparatus for compressive strength tests at various temperatures is composed of
an electrical furnace to heat the specimens to the target temperature, loading machine to conduct
the compressive strength test, and compressometer equipped with two high accuracy linear
variable displacement transducers (LVDT), to track the deformations. The electric furnace shown
in Figure 3.3 is specially designed for simulating high-temperature conditions and can produce
maximum temperatures up to 925°C. It is equipped with internal heating electric elements, a ramp
and a hold temperature controller that can generate different heating rates. Along with attaining
the compressive strength, the 1800 kN load-controlled strength testing machine shown in Figure
3.4, combined with the compressometer shown in Figure 3.5, provide the stress-rate controlled
stress-strain relationship by monitoring the load and deformation with respect to time. The elastic
modulus can be extracted from the stress-strain relation utilizing equation 1, as prescribed in
ASTM-C469/C469M (ASTM C469/C469M, 2014).
𝐸 = (𝑆2 − 𝑆1 )/(𝜀2 − 0.000050)

(1)

Where E, is the chord modulus of elasticity (GPa), S1, is the stress corresponding to a longitudinal
strain, S2, is the stress corresponding to 40% of ultimate load (MPa), 𝜀1 , of 50 millionths, (MPa),
and 𝜀2 , is the longitudinal strain produced by stress, S2, (mm/mm). The LVDTs are part of the
compressometer, thus providing more local and accurate deformation measurements.
3.3.4 Test procedure
3.3.4.1 General procedure
The specimen was evaluated at the elevated target temperature, and the applied load
commenced following the heating of the specimen to the appropriate temperature. RILEM

58

standard considers such a procedure as an unstressed test. Furthermore, a steady-state method is
adopted for determining the compressive strength utilizing the electric furnace (Figure 3.3). In the
steady-state testing method, the specimen is heated to a predetermined target temperature, and then
the specimen is immediately transferred to the strength testing machine for applying stress. Figure
3.6(a) depicts the desired regime of heating in this kind of test and Figure 3.6(b) summarizes the
desired rate of heating for each type of concrete evaluated. The RILEM standard recommends a
heating rate of 2ÂşC/min for specimens with a maximum diameter of 80 mm and a temperature
recording interval of 4 min. For the testing conducted, a rate of heating of 2ÂşC/min was utilized
for the four conventional types of concretes. Unfortunately, due to UHPC’s lower permeability as
compared to conventional concretes, unless a lower heating regime was employed, tests at and
above 300°C would result in explosive spalling. The spalling of this nature reduces the UHPC
specimens into fragments and a compressive strength test cannot be conducted. Therefore, to
improve the understanding of the explosive spalling behavior of UHPC, a heating rate ranging
from 0.25 to 4ÂşC/min were employed. For all tests conducted, temperature recordings were taken
every thirty seconds. Furthermore, due to the constraints of the electric furnace (Figure 3.3),
coupled with the installation of the compressometer (Figure 3.5), the core temperature is taken
30°C beyond the target temperature. As seen in Figure 3.6(a), the increased temperature facilitates
the additional time required to install the compressometer. Plus, as recommended by RILEM
standard the additional 30°C permits the specimen to saturate for 30-45 min before removal from
the furnace. Since the surface thermocouple is utilized to track the heating rate for the conventional
concretes, the specimen is ready once the surface temperature reaches 60°C beyond the target
temperature. The thermocouples are attached to a data acquisition device (Figure 3.7) that
continuously monitor and records the temperature.

59

3.3.4.2 Compressive strength, stress-strain relationship, and elastic modulus
Once concrete specimen attains the targeted temperature, it is removed from the furnace,
Figure 3.6(b), and while the center thermocouple remains intact (UHPC specimens only), the
surface thermocouple is removed prior to weighing the specimen. However, since the conventional
concretes do not have a center thermocouple, the surface thermocouple remains in place to monitor
the temperature while conducting the strength test. Preceding to place the specimen within the
Forney strength machine (Figure 3.4), the compressometer (Figure 3.5) is installed upon the
cylinder. As seen in Figure 3.5, there are two rods that hold the apparatus together, and once the
compressometer is installed upon the concrete specimen, the rods are removed prior to
commencing the compressive strength test. Tests conducted at 20ÂşC follow the steps outlined in
this section, excluding the placement of thermocouples and heating within the furnace.
When the specimen is set up within the Forney strength test machine, as per ASTM C39/C39M, a
load rate of 0.25 Âą 0.05 MPa/sec is applied and must be maintained for the duration of the test.
Once the concrete specimen reaches its maximum load and depending on the temperature at which
the test is being conducted and type of concrete, the specimen may fail in an explosive and brittle
way, thus the test is concluded. Upon hearing the explosion, the test of specimens with steel fiber
is continued briefly to ensure the specimen has in fact completely failed; Further discussion will
be provided in section 3.4.2. The maximum load along with the load versus displacement data
collected is used to develop the stress versus strain curves, along with the modulus of elasticity.
Unless there was a poor failure as per ASTM-C39/C39M (ASTM C39/C39M, 2016), one
conventional concrete specimen was evaluated at each respective temperature. While 3-4
repetition tests were conducted for UHPC A & B, and F up to 400°C, unfortunately, due to limited
specimens cast and explosive spalling, only 1 specimen was tested for UHPC F above 400°C and
60

UHPC D. UHPC A & B and D’s compressive strengths were evaluated at 20, 100, 200, 300, and
400ÂşC, but for UHPC F, tests were conducted similarly as UHPC A & B and D, plus at 600 and
700°C. NSC tests were at 20, 400, 600, 700, and 750 ºC, NSC-SF at 20, 200, 400, 600, and 700ºC,
HSC at 20. 200, 400, and 600ÂşC, and HSC-PP at 20, 200, 400, 600, 700, and 750ÂşC.
3.4 Results and discussion
Results from the compressive strength tests are presented in this section. This section will discuss
each type of concrete’s results individually. Chapter four will discuss the UHPC results as
compared to the conventional concretes tested.
3.4.1 Sectional temperatures
As illustrated in Figure 3.1, thermocouples are mounted in the electrical furnace and on the
cylinders to record the temperature data with respect to time. The thermocouples ensured the
appropriate rate of heat is being applied to the specimen while it is in the electric furnace and the
proper temperature is reached. The temperature data acquisition device (Figure 3.7) records the
temperatures via the thermocouples every 30 seconds. Following the completion of the test, the
data is analyzed and the actual average rate of heat is calculated by plotting the temperature versus
time (Figures 3.8), and enumerating the average slope developed during the heating of the concrete
specimen. In each respective figure, the targeted rate of heat is shown via a dashed dark grey line
as annotated by the legend. Additionally, the temperature versus time data, depicts the thermal
gradients experienced by the UHPC specimen while being heated within the electric furnace. The
thermal gradient is calculated utilizing equation (2).
𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑔𝑟𝑎𝑑𝑖𝑒𝑛𝑡 =

61

𝑇𝑠 −𝑇𝑐
𝑟𝑠𝑝𝑒𝑐

(2)

Where, 𝑇𝑠 , is the surface temperature (C), 𝑇𝑐 , is the center temperature (C), and 𝑟𝑠𝑝𝑒𝑐 , is the radius
of the concrete cylinder, i.e. distance from the surface to the placement of the center thermocouple
within the specimen (centimeter, cm). Therefore, employing equation (2) with the temperature rise
data with respect to time, the thermal gradient is graphed with respect to time as well. The thermal
gradient data is pertinent regarding UHPC’s explosive spalling behavior and will be discussed
further in section 3.4.3.2.
The desired heating rate utilized for the conventional concretes tests was 2ÂşC/min and to
avoid explosive spalling, the rate of heat for UHPC testing depended upon the target temperature,
Figure 3.6(b). Tests conducted at temperatures of 300ÂşC and above employed a heating rate of
0.50°C/min, while the remaining UHPC tests employed the same heating rate as the conventional
concretes. Figure 3.8 illustrates the rise in temperature to 600°C for each type of concrete. Due to
concrete’s low thermal conductivity and high specific heat, as expected the surface temperatures
increase more rapidly as compared to the temperature within the core of the cylinder (Figure 3.8).
Although the equipment is tested prior to each test, at times during a test, a thermocouple would
begin providing erroneous data. While each thermocouple is essential, the surface and center
thermocouples, Figure 3.1(a), are more critical for they provide vital data with respect to explosive
spalling. If the surface thermocouple was reporting erroneous values, the average between the
furnace, and center thermocouples was taken to be the surface values. Additionally, if the center
thermocouple experienced damage, the recorded surface temperatures were utilized for the average
rate of heat and the thermal gradient was not calculated. Furthermore, in case a similar test was
performed previously, i.e. same UHPC type, and furnace input, the respective value from the
preceding test were utilized. Table A1, A2, and A3 in Appendix A will provide further details
associated with each UHPC design mix test.
62

The conventional concretes (Figure 3.8(a) and (b)) temperature rise to 600°C is very
similar. However, due to the lower thermal conductivity of UHPC, as compared to the
conventional concretes, the time required to reach 600°C is over three times greater. There was a
smaller learning curve with the conventional concretes as opposed to UHPC, to calibrate the
furnace to provide the desired rate of heating. Therefore, the same average rate of heat were not
utilized for the conventional concretes depicted in Figure 3.8. However, regarding the tests
conducted at the other temperatures, NSC-SF and HSC both required greater time to reach their
target temperatures as opposed to NSC and HSC-PP, respectively. The additional time for NSCSF is due to the steel fibers and the lack of polypropylene fibers for HSC. Regarding Figure 3.8(c)
the desired rate of heating to allow UHPC F9 to reach 600°C was 0.5ºC/min and as seen in the
figure it took over 17 hours to do so. Furthermore, similar to HSC, the design mixes of UHPC
containing no polypropylene fibers required a greater time versus design mix UHPC F. Moreover,
due to explosive spalling, temperatures above 400°C were not attainable for UHPC A & B and D.
Finally, as seen in Figure 3.8(c) no visible moisture induced plateau was seen in any UHPC F
temperature versus time curve (Klingsch, 2014; Ju et al., 2015). The rate of heating for UHPC will
be discussed in detail in section 3.4.3.
3.4.2 Compressive strength, elastic modulus, and stress-strain relationship
The compressive strength, and modulus of elasticity with respect to temperature are
presented in Figures 3.9 and 3.10, accordingly. All the tests were load-controlled and stopped as
soon as the specimen experienced failure. The stress-strain relationship will be presented in the
following subsequent sections for each respective type of concrete, along with the discussion of
all three mechanical properties. Details pertaining to each UHPC test are given in Tables 3.5, 3.6
and 3.7; While the conventional concretes respective data are in Table 3.8 and 3.9. Chapter Four

63

will compare and discuss the 3 UHPC design mixes, with the 4 conventional concretes and ASCE
(1992) and Eurocode 2 (2004).
3.4.2.1 NSC
Figure 3.11 and 3.13 provide the stress-strain relationship for NSC and NSC-SF,
respectively. Additionally Figure 3.12 and 3.14 display the type of failure modes for NSC and
NSC-SF, following the compressive strength tests. Both types of concretes were evaluated from
20-750°C and the ambient temperature tests for both NSC and NSC-SF were conducted utilizing
rubber caps. Due to the limited capability of the electrical furnace (Figure 3.3), 800°C tests could
not be conducted. The maximum temperature attained were 750 and 700°C for NSC and NSC-SF,
respectively. The NSC-SF specimens’ slightly lower temperature is attributed to the steel fiber
content and the higher specific heat of the concrete at temperatures above 400°C (Kodur and
Khaliq, 2011). The compressive strength tests conducted up to 400°C resulted in a thunderous
explosion. However, the tests conducted at, and above 600°C did not produce any sound upon
failure. The concrete’s hue (Figure 3.12 and 3.14) begins to change following exposure to 600°C,
and above.
The compressive strength (Figure 3.9) of NSC and NSC-SF reduces to about 50% at 600°C.
Such a decrease may be attributed to the severance of calcium-hydroxide between 400 and 600°C
(Schrefler et al., 2002; Klingsch, 2014). Furthermore, the relative compressive strength of NSC at
750°C drastically decreases below 25%. The calcium-carbonate decarbonating at temperatures
greater than 700°C may be the cause of the sudden loss of strength. Moreover, due to the electrical
furnaces’ capability, in an attempt to reach 800°C, the reduction may have been exacerbated
further due to the prolonged exposure to the heat. Overall, up to 700°C, NSC’s compressive
strength (Figure 3.9) degradation is much smoother as compared to NSC-SF. Although the NSC64

SF test conducted at 200°C suffered a diagonal fracture (Figure 3.14), it resulted in a lower
compressive strength than the 400°C test.
For the most part NSC and NSC-SF behave very similarly with respect to their stress-strain
relationship (Figure 14, and 16). At ambient conditions, NSC-SF displayed 33% more ductility
than NSC and with increasing temperatures NSC-SF remained more ductile. Although, steel fibers
are predominantly utilized for tensile strength purposes, the additional ductility during the stressstrain curve is attributed to the fibers as well. As the specimens were exposed to elevated
temperatures, the peak strain increased due to softening of the material. The elastic modulus for
both, NSC, and NSC-SF regress faster as compared to their compressive strengths. NSC elastic
modulus reaches down to 26% at 400°C, and NSC-SF even further to 16% as compared to their
respective ambient condition stiffness. The decaying of the cement paste most likely caused the
rapid stiffness decline (Kalifa et al., 2000). The addition of steel fiber to the concrete did not
contribute towards its stiffness.
3.4.2.2 HSC
Figure 3.15 and 3.17 provide the stress-strain relationship for HSC and HSC-PP,
respectively. Additionally Figure 3.16 and 3.18 display the type of failure modes for HSC and
HSC-PP, following the compressive strength tests. HSC was evaluated from 20-600°C, and HSCPP up to 750°C, and the ambient temperature tests for both HSC and HSC-PP were conducted
utilizing rubber caps. Due to the limited capability of the electrical furnace (Figure 3.3), 800°C
test for HSC-PP could not be conducted. The compressive strength tests sound upon failure for
HSC and HSC-PP is equivalent to NSC and NSC-SF, and the concrete’s hue (Figure 3.16 and
3.18) begins to change following exposure to 400°C, and above.

65

Due to the denser microstructure achieved through the addition of silica fume and high
content of cement, coupled with the reduction in w/c ratio, the compressive strength’s (Figure 3.9)
reduction is much more drastic. The HSC and HSC-PP design mixes tested, employ over two times
the amount of cement as compared to the NSC and NSC-SF. Therefore, at temperatures between
100-300°C, the decomposition of the hydration products of cement contribute greatly to the
strength loss (Dehn, 2004; Klingsch, 2014). As seen in Figure 3.9, at 400°C the compressive
strength is reduced to a third of its original strength. However, HSC and HSC-PP maintain around
30% of their original strength at 600°C. The minimal strength loss is attributed to the reduction of
calcium-hydroxide formation accomplished with the addition of silica fume. Though it is only 7%
of the cement, it is enough to relatively maintain its strength (Khoury, 2000; Zdeb, 2013). The
decreasing strength of HSC-PP beyond 600°C is owed to the analogous factors that influenced
NSC’s strength.
Beyond 200°C, HSC-PP (Figure 3.17) displayed greater ductility as opposed to HSC
(Figure 3.15), and excluding the HSC test at 600°C, the peak strain increased with elevated
temperatures due to softening of the material. The HSC at 20°C and HSC-PP at 200°C, were
repeated due to erroneous results and poor compressive failure mode of the cylinder. Furthermore,
below temperatures of 600°C, the addition of polypropylene fibers improved the stiffness. While,
HSC’s stiffness degraded below 50% at 200°C, it wasn’t until 400°C for HSC-PP. Finally, at
600°C, both concretes stiffness degraded to about 15% of its original elastic modulus. Therefore,
the addition of polypropylene fibers with HSC assists its ductility and stiffness with increasing
temperatures.

66

3.4.2.3 UHPC
Figure 3.19, 3.21 and 3.23 provide the stress-strain relationship for design mixes UHPC A
& B, D, and F, respectively. Additionally Figure 3.20, 3.22 and 3.24 display the type of failure
modes for each UHPC design mix, following the compressive strength tests. Due to UHPC A &
B and D undergoing explosive spalling, both design mixes’ mechanical properties were only
evaluated up to 400°C; while the addition of polypropylene fibers granted UHPC F to be tested up
to 700°C. Moreover, due to explosive spalling and limited resources, only design mixes UHPC A
& B and F allowed repetition of tests (3-4), conducted at temperatures of 20-400°C and tests at
and beyond 300°C could only be performed utilizing an heating rate of less than 1ºC/min. The data
provided for the remaining UHPC tests, consisted of a singular test. There are 4 UHPC A & B
tests in Table 3.5, conducted at 20, 100, 200 and 200°C, that did not adhere to the method according
to ASTM-C469/C469M, but followed ASTM-C39/C39M. Therefore only their respective
compressive strengths were attained, and their corresponding data is incorporated in Figure 3.9.
The compressive strength tests sound upon failure and the concrete’s hue (Figure 3.20, 3.22 and
3.24) are similar to the design mixes of HSC.
The ambient temperature data collected for UHPC A & B and F were conducted utilizing
caps and with no cap. Unfortunately, due to the limited number of specimens of design mix UHPC
D, no 20°C test could be conducted utilizing a cap. The cap removes any imperfections upon the
horizontal surfaces and provides a higher value. Therefore, though UHPC D’s 28-day compressive
strength is 10% greater than UHPC F, from Figure 3.9 UHPC D and F have the same compressive
strength at ambient conditions.
All the UHPC design mixes tested experienced a decrease in strength at 100°C, with UHPC
A & B experiencing the greatest reduction of 18%. The degradation of strength is attributed to the
67

evaporation of the non-chemically bound water within the UHPC, causing increased permeability
and loss of strength (Dehn, 2004; Klingsch, 2014). However, following exposure of 200°C, UHPC
F was the only design mix that did not increase in strength. The increase in strength is owed to
shrinkage of the concrete following the water evaporation (Xiong and Liew, 2015). UHPC F did
not experience an intensification until 300°C. The delayed response of the surge of strength may
be caused by the melting of polypropylene fibers around 150-170°C and further increasing the
permeability (Klingsch and Frangi, 2011; Bei and Zhixiang, 2016).
At 300 and 400°C, UHPC D and F maintain strengths close to their respective original
strength, with UHPC D achieving strengths greater than the original at 110 and 104% at 300 and
400°C, respectively, while UHPC design mix A & B deviate from the group beyond 200°C. The
nonconformity seems to be caused by either the specimen was not properly prepared, resulting in
a poor break, or higher heating rates deteriorated the strength. The former was the cause for one
test and the latter for the remaining four. The greatest compressive strengths attained for UHPC A
& B at 300 and 400°C employed heating rates that were on average 30% lower than the lower
compressive strengths, with on average 16% greater duration of heating. Moreover, the supreme
strength attained from all UHPC design mixes, was by UHPC D8 tested at 300°C with an average
heating rate of 0.25ÂşC/min and total heat exposure duration of 18 hours. Although nearly reducing
the average rate of heating to 50%, UHPC F sustained a 16% increase in compressive strength at
200°C, dramatically increasing the duration in the furnace and employing extremely slow heating
rates, at 300 and 400°C negatively effects UHPC design mix F. When employing a heating rate
like UHPC D8 (0.25ÂşC/min), on average the duration within the furnace was increased and strength
decreased, 2.5 times and 8%, accordingly.

68

UHPC F9 decays to 70% of its original strength at 600°C and experiences a drastic
deterioration at 700°C (UHPC F13). The inordinate amount of silica fume as compared to the
cement content in UHPC F provides a very low C/S ratio, thus resulting in minimal calciumhydroxide (Khoury, 2000; Zdeb, 2013). Therefore, the sudden decrease experienced by UHPC F9
and F13, is dominated by the ι-quartz to β-quartz transformation occurring around 570°C coupled
with the continual dilapidation of the cement paste (Dehn, 2004; Klingsch, 2014).
At ambient conditions (Figure 3.19, 3.21 and 3.23), save the maximum compressive
strengths attained, the three UHPC design mixes tested resulted in similar peak strains. At elevated
temperatures, for the most part UHPC A & B, D, and F behave very similarly, with respect to their
stress-strain relationship, with increasing temperatures the peak strain increases. However, as
explained previously, some of the UHPC A & B tests conducted at 300 and 400°C differed slightly.
The elastic modulus for UHPC behaved quite differently as compared to their respective
compressive strengths. At ambient conditions the elastic modulus for all three UHPC tested is very
similar. At temperatures of 100 and 200°C, UHPC F’s elastic modulus experiences the slowest
degradation. However, all three UHPC tested reach half their elastic modulus around the same
temperature (300°C). At 600 and 700°C design mix UHPC F elastic modulus sustained a rapid
decrease to 10 and 7% from ambient conditions.
The addition of polypropylene fibers along with steel fibers, UHPC F, does improve the
fire performance with respect to compressive strength, elastic modulus below 400°C and reaching
temperatures above 400°C. Finally, utilization of larger volume content of steel fiber (design mix
UHPC A & B) does not necessarily increase the compressive strength.

69

3.4.3 Fire-induced spalling behavior
The following two sections will emphasize the prominent results accumulated as part of
this study corresponding to explosive spalling in UHPC. The data pertaining to the effects of rate
of heating will be discussed first, followed by thermal gradients.
3.4.3.1 Spalling behavior in experiments
The results from this study indicate that employing a low rate of heating does not
necessarily correlate with the mitigation of explosive spalling. Though the rate of heat and
development of thermal gradients are vital, their resultant, pore pressure, seems to be the
predominant trigger (Felicetti et al., 2000). The effects of rate of heat upon the three UHPC design
mixes will be discussed here and information pertaining to thermal gradients will be presented in
detail in the following section.
During this study, the electric furnace experienced fourteen fire induced spalling events;
of those, thirteen were explosive in nature. As seen by Figure 3.25, such an explosion produces an
exorbitant amount of energy, enough to expand the metal frame of the furnace. The numerous
explosive spalling events experienced by the furnace caused a gap of at least 6.35 mm, for most of
the height of the furnace door. Therefore, tie down straps were utilized to compensate for the gap.
The explosion was usually accompanied with a very loud sound, and Figure 3.26 displays
specimens from each design mix that underwent explosive spalling. Prior to utilization of the
temperature acquisition data device (Figure 3.7), a handheld thermocouple reader was used, and
the temperature was measured every 30 to 45 min. Therefore, the first three elevated temperature
tests listed in Table 3.5, the rate of heats listed are estimates, and not exact values. Moreover, the
initial goal was to maintain a rate of heat of 2ÂşC/min as recommended by RILEM standard.
However, there was a bit of a learning curve due to testing a novel material. In the first three
70

elevated temperature tests as listed in Table 3.5, the actual heating rates were far below the desired
2ÂşC/min. Therefore, the electric furnace (Figure 3.3) was calibrated with UHPC A1, in desire to
better understand the furnace behavior with UHPC, coupled with an attempt to complete a 600°C
compressive test. Unfortunately, specimen UHPC A1 experienced explosive spalling with an
average rate of 1.05ÂşC/min.
To better comprehend the explosive spalling involved with UHPC a new testing regime
was devised. In short, rate of heating ranging from 2 to 4ÂşC/min were to be used, with temperature
ranging from 200 to 400°C with design mix UHPC A & B. The data associated with this testing,
along with all UHPC data for tests conducted at, and above 300°C, is given in Table 3.10. To avoid
presenting redundant data, thermal gradient results are presented in Table 3.10 as well, however
as stated previously, thermal gradient results will be discussed further in the following section.
Table A1, A2, and A3 in Appendix A will provide further details associated with each UHPC test
conducted.
As expected, the higher the rate of heat, the sooner the UHPC specimen will experience
explosive spalling. UHPC B3 (Figure 3.26) experienced explosive spalling in the shortest period,
80 min, combined with the lowest center temperature, 191°C. Although specimen UHPC B8’s rate
of heat was higher than UHPC B3, UHPC B8 exploded at 95 min and at a greater core temperature,
248°C. The conflicting results may be attributed to differing furnace input or variance in moisture
within the specimen. As opposed to design mix UHPC A & B, UHPC F contains 2.00 kg/m 3 of
polypropylene fibers as prescribed by Eurocode 2 (2004), and explosive spalling was not expected.
However, UHPC F1 (Figure 3.26) sustained explosive spalling with a rate of heat less than
1.50°C/min. The test was repeated and UHPC F12 underwent explosive spalling around the same
temperature, 295°C. Furthermore, UHPC B3’s test was repeated but with UHPC F14 and sustained
71

explosive spalling at the equivalent time, with an average rate of heat 2.50°C/min and core
temperature slightly higher, 220°C. Item to mention, due to an experiment error, UHPC A7’s test
was abruptly ended and sustained mere surface spalling. However, as the rate was shy of 2ÂşC/min
it is highly expected if the furnace was not shut down too soon, it would have experienced
explosive spalling. The nine specimens to explosively spall as part of the initial testing regiment,
the average rate of heat was 2.70°C/min, with the least slightly higher than 1ºC/min and average
core temperature of 250°C. Therefore, it was concluded to conduct a successful 300°C or higher
compressive strength test, a rate of heat lower than 1ÂşC/min must be utilized.
To further study the behavior of UHPC, all three design mixes were studied with a desired
rate of heat of 0.50°C/min at temperatures of 300°C and above. To achieve such a rate, two
different furnace inputs were utilized, a steady increase with a high target temperature (referred to
as input 1) and rapid increase with a low target temperature (referred to as input 2). The former
produced three (UHPC B1, B2 and D7) out of the nine tests to explode and surprisingly the latter
only two (UHPC D5 and D6) out of fourteen. Employing either of these inputs, design mix UHPC
F did not sustain any explosions.
UHPC D4 (average rate of heating, 0.50°C/min) was the sole specimen from design mix
UHPC D to avoid explosive spalling with input 2 and UHPC D5 and D6 experienced explosive
spalling with a center temperature close to 270°C. Astonishingly, the average rate of heat for the
two tests were very similar to the UHPC A & B tests that did not experience explosive spalling, of
around 0.95ÂşC/min. Following the resulting two explosions, UHPC D7 and D8 were tested
employing input 1 and desired average rate of heating of 0.25ÂşC/min, with 50% success. UHPC
D7 (Figure 3.26) sustained an explosion following a slight increase to the furnace input with core
temperature of 284°C and a time to spall of 850 min. UHPC D8 avoided an explosion with an
72

average rate of heating of 0.25ÂşC/min. Additionally, UHPC B5 was the sole test to avoid explosive
spalling from design mix UHPC A & B with input 1 and the core temperature was taken to 329°C.
Moreover, upon removal from the furnace, there was moisture visibly escaping the specimen in
numerous locations (Figure 3.27). Design mix UHPC F overall average rate of heating to avoid
explosive spalling was 0.75ÂşC/min. The last four UHPC F tests listed in Table 3.7 were not
included as part of the previous average. The tests consisted of altering the average rate of heating
drastically in an attempt to evaluate the compressive strength following prolonged exposure in the
furnace.
The average heating rate, core temperature and time to spall for the specimens to undergo
explosive spalling for each separate design mix are: UHPC A & B, 1.75ºC/min, 250°C and 170
min, UHPC D, 0.75ºC/min, 275°C and 440 min, and UHPC F, 2.00°C/min, 270°C and 140 min.
Overall amongst all UHPC to experience explosive spalling the averages are: 1.50°C/min, 265°C
and 250 min. UHPC D’s time to spall is very high due to UHPC D7 undergoing the explosion at
850 min. If UHPC D7 is not included, UHPC D’s average time to spall is 232 min. Figure 3.28,
displays the average rate of heating with respect to either, the maximum core temperature attained,
or the core temperature for which the specimen experienced explosive spalling, for all the UHPC
tests from Table 3.7. Each specimen is labeled in Figure 3.28 and all 14 specimens to experience
explosive spalling are depicted with a red box along with their respective time to spall. It is
abundantly clear, the rate of heating to avoid explosive spalling for design mixes UHPC A & B,
and F is 0.75ÂşC/min (depicted by a pink dashed line in Figure 3.28) and design mix UHPC D is
0.25ÂşC/min (depicted by a red dashed line in Figure 3.28).
From this experimental study, employing a higher average rate of heating, shows that
UHPC can undergo explosive spalling quicker, with core temperatures and average rate of heating
73

as low as 191°C and 1.50°C/min, accordingly. UHPC A & B contains 10% more steel fiber and
its critical rate of heating to avoid explosive spalling is 3 times versus UHPC D. In addition, steel
fiber alone and the combination with 2.00 kg/m 3 of polypropylene fibers does not mitigate
explosive spalling. Moreover, tests confirmed minimal average rate of heating alone do not
mitigate the explosive spalling behavior of UHPC. As seen with UHPC B5 and D7, moisture is
quite relevant at temperatures above 300°C. Therefore, at lower heating rates, development of
thermal gradients coupled with the release of vapor pressure are more crucial. Effect of thermal
gradients will be discussed further in the following section.
3.4.3.2 Temperature gradient and rate of heat analysis
The thermal gradients development within UHPC directly affect the release of built-up
pore pressure (Phan, 2008: Klingsch, 2014). Furthermore, since majority of the moisture content
is chemically bound within the binder and does not completely evaporate until temperatures above
800°C, the risk of explosive spalling resulting from thermal gradients is continuously present
(Klingsch, 2014). Therefore, results from this research give importance to not only high thermal
gradients, but to minimal values as well (Ju et al., 2013). Table 3.11 displays the key points from
the thermal gradient with respect to time graphs for each UHPC test listed in Table 3.10. Table
3.11 does not include the UHPC tests conducted employing an average heating rate less than
0.50°C/min. Furthermore, the thermal gradient with respect to time graph for each UHPC design
mix (A & B, D, and F) tests, are given in Figure 3.29, 3.30 and 3.31, respectively. Table 3.12 is
given to serve as a legend to Figure 3.29, 3.30 and 3.31. The legend describes the average rate of
heating for each UHPC test.
Referring to Table 3.10 and 3.11, the UHPC tests that underwent high average rate of
heating, of at least 1.25ºC/min, the maximum thermal gradient ranged from 14.55 to 32.00°C/cm.
74

The maximum thermal gradients occurred on average within 82 min of heating duration and the
center temperature of the specimen did not go beyond 240°C. Eight of these tests, UHPC A13, B8,
B3, A4, F1, F12 and F14 (Figure 3.29 and 3.31), underwent explosive spalling, which happened
to occur on average about 40 min following the peak thermal gradient. Only UHPC F14
experienced the maximum thermal gradient and explosive spalling at the same time and UHPC F1
with the longest of 86 min. UHPC A13, B8, F1 and F14 respective thermal gradients experienced
a short-lived sudden decline following the apex, and then a rapid increase which was concluded
abruptly by the occurrence of explosive spalling. UHPC F1’s test was not concluded following the
explosive spalling, because the loud explosion was not heard. Thereby, the specimen was
discovered to have exploded once opening of the furnace. UHPC B3 and F12 did not suddenly
increase following the decline and underwent explosive spalling during their respective decline.
Prior to the test prematurely ending, UHPC A7’s (Figure 3.29) thermal gradient was at the initial
stages of decreasing. Based off the data attained by the three tests that did explode, if the test were
to have continued it too would have experienced explosive spalling.
The UHPC specimens that experienced explosive spalling with the slower rate of heating,
of less than 1.25ÂşC/min, the maximum thermal gradients occurred at a much later time as compared
to the higher average rate of heating. On average, the maximum thermal gradient occurred within
200 min for these five tests, and the time to spall occurred within 70 min thereafter. Three of the
five tests (UHPC A1, B1 and B2) underwent explosive spalling utilizing input 1. Furthermore, the
thermal gradient reached the maximum value during the latter part of the test; with each of the
specimens experiencing an explosion within 20 min following the maximum thermal gradient. The
remaining two tests (UHPC D5, and D6) that underwent explosive spalling, utilized input 2. The
maximum thermal gradient was attained much faster (within 90 min), and the explosion occurred

75

144 min following the maximum thermal gradient. Specimen UHPC D6’s maximum thermal
gradient (20°C/cm) was as high as the specimens that underwent the higher rate of heating (at least
1.25ºC/min). Surprisingly, specimen UHPC D5’s pinnacle value for its thermal gradient was under
5ÂşC/cm, yet it still underwent explosive spalling.
On the other hand, there were thirteen specimens that did not explode implementing the
lower rate of heating. As stated in the previous section UHPC B5 is the sole test not to explode
with input 1. Its maximum thermal gradient was 6.9ÂşC/cm and occurred during the latter part of
the test. Over 90% of the remaining twelve specimens, the maximum thermal gradient occurred
early (within 90 min), ranging from 6.53 to 8.83ºC/min. UHPC F13 (700°C) is the sole test where
the maximum thermal gradient occurred at the end of the heating duration. In an attempt to reach
700°C, the furnace input’s rate was increased towards the latter part of the heating process. Thus,
if the input was not drastically increased, the maximum thermal gradient would have occurred
within the first 100 min of heating.
Based off this study, UHPC inflicted with higher average rate of heating, will develop
greater thermal gradients within the specimen that will occur later in the heating process. Average
rate of heating greater than 1.25ºC/min coupled with thermal gradients of at least 20°C/cm for
design mix UHPC A & B and 5ÂşC/cm for design mix UHPC F, 100% of the specimens experienced
explosive spalling. Plus, the maximum thermal gradient occurred within two-thirds of the heating
duration. Additionally, it makes sense UHPC F would result in a lower thermal gradient because
of the presence of polypropylene fibers (Phan, 2008). However, with lower average rate of heating
of less than 1ÂşC/min, two different scenarios occurred, (1) if the maximum thermal gradient (at
least 10°C/cm) was reached towards the latter portion of the heating process, 67% of the specimens
underwent explosive spalling, and (2) if the maximum thermal gradient (at least 5ÂşC/cm) was
76

reached early during the heating process, 22% of the specimens experienced explosive spalling.
The lower maximum thermal gradient corresponds with UHPC D5 and the higher to UHPC B1.
Given the critical average rate of heating for design mix UHPC D is lower than UHPC A & B, the
values do coincide. Monitoring the thermal gradient while the UHPC specimen is being heated,
can provide an inclination whether the UHPC will in fact, experience explosive spalling.

77

Table 3.1: Testing regime conducted as part of this study
Elevated
temperatures
Cylinder size
75x150 mm
24
7
17
6
3
5
4

Room temperature
Number

Concrete type

1
2
3
4
5
6
7

UHPC A & B
UHPC D
UHPC F
HSC-PP
HSC
NSC-SF
NSC

4
1
4
1
1
1
1

Table 3.2: Mix proportions for UHPC design mixes tested as part of study
Components
Limestone No. 8, kg/m3
Michigan Natural Sand, kg/m3
Silica Sand, kg/m3
Cement Type I, kg/m3
Silica Fume, kg/m3
Slag St. Lawrence, kg/m3
Limestone Powder, kg/m3
Steel Fiber, kg/m3
Polypropylene Fiber, kg/m3
Water, kg/m3
Water Cement Ratio (w/c)
HWRA (28%) Chryso., kg/m3

UHPC A & B

UHPC D

UHPC F

526
553
304
519
228
104
187
130
153
0.14
43

487
513
282
481
212
96
173
118
114
0.14
46

488
514
283
482
212
96
173
118
2.00
115
0.14
46

78

Table 3.3: Mix proportions for conventional concretes tested as part of study
Components

HSC

HSC-PP

NSC

NSC-SF

Cement Type I, kg/m3

561

561

230

230

Fine aggregate, kg/m3

641

641

930

930

Course aggregate 3/8" Limestone, kg/m3

1089

1089

985

985

42

42

-

-

-

-

-

42

-

1.66

-

-

151

151

157

154

0.30

0.30

0.50

0.50

-

-

25

25

-

-

77

77

Silica Fume, kg/m

3

Steel fiber, kg/m3
Polypropylene fibers, kg/m
Water, kg/m

3

3

Water cement ratio (w/c)
Superplastizer, kg/m

3

Fly Ash (Class C), kg/m3

Table 3.4: Density and 28-day compressive strengths of concrete cylinders tested as part of study

Concrete type
UHPC A & B
UHPC D
UHPC F
HSC
HSC-PP
NSC
NSC-SF
* - Denotes an estimate

28-day
compressive
strength, MPa
132
182
165
111
100
29
29

79

Density, kg/m3

Density, g/cm3

2500 *
2456
2436
2297
2358
2256
2231

2.50 *
2.46
2.44
2.30
2.36
2.26
2.23

Table 3.5: Details of compressive strength tests of UHPC A & B cylinders
T

Average rate
of heat

f c'

E

Tmax(core)

Tmax(furnace)

ÂşC

ÂşC/min

MPa

GPa

ÂşC

ÂşC

UHPC B9

20

N/A

135

X

N/A

N/A

UHPC A12

20

N/A

139

39

N/A

N/A

UHPC A2

20

N/A

134

42

N/A

N/A

UHPC A7

20

N/A

138

38

N/A

N/A

UHPC A

100

0.60 *

121

X

X

X

UHPC A

100

0.40

109

30

X

X

UHPC B14

100

1.37

119

35

133

237

UHPC A3

100

0.30*

112

31

105

129

UHPC A9

100

1.00*

103

38

132

174

UHPC A14

100

2.00*

108

35

116

162

UHPC A & B

200

0.70*

118

X

X

X

UHPC A6

200

0.54

138

25

218

262

UHPC A & B

200

0.75*

135

X

X

X

UHPC B12

300

0.70

114

18

354

353

UHPC B19

300

0.65

120

11

X

443

UHPC A12

300

0.42

135

18

301

330

UHPC B4

400

0.60*

100

11

X

X

UHPC A2

400

0.69

117

14

434

339

UHPC A13

400

0.50

134

13

401

425

UHPC B11
400
* – Denotes an estimate

0.60

124

13

X

443

Specimen

X – Denotes data not collected

80

Table 3.6: Details of compressive strength tests of UHPC D cylinders
T

Average rate
of heat

f c'

E

Tmax(core)

Tmax(furnace)

ÂşC

ÂşC/min

MPa

GPa

ÂşC

ÂşC

UHPC D-1

20

N/A

148

36

N/A

N/A

UHPC D-2

100

1.60

134

27

161

244

UHPC D-3

200

1.13

142

19

237

X

UHPC D-8

300

0.27

163

15

330

347

0.53

153

10

455

463

Specimen

UHPC D-4
400
* – Denotes an estimate
X – Denotes data not collected

81

Table 3.7: Details of compressive strength tests of UHPC F cylinders
T

Average rate
of heat

f c'

E

Tmax(core)

Tmax(furnace)

ÂşC

ÂşC/min

MPa

GPa

ÂşC

ÂşC

UHPC F-4

20

N/A

128

38

N/A

N/A

UHPC F-11

20

N/A

170

39

N/A

N/A

UHPC F-19

20

N/A

153

X

N/A

N/A

UHPC F-21

20

N/A

140

X

N/A

N/A

UHPC F-2

100

1.64

120

36

169

229

UHPC F-6

100

2.66

125

X

X

304

UHPC F-20

100

1.65

131

X

170

317

UHPC F-3

200

1.89

115

40

230

330

UHPC F-7

200

1.76

127

24

234

300

UHPC F-10

200

0.91

133

X

238

231

UHPC F-5

300

0.75

150

16

330

360

UHPC F-15

300

0.21

136

13

319

346

UHPC F-17

300

0.25

143

10

329

344

UHPC F-8

400

0.74

151

11

433

X

UHPC F-16

400

0.23

142

13

421

433

UHPC F-18

400

0.24

132

10

421

436

UHPC F-9

600

0.74

100

4

632

629

0.69

38

3

731

768

Specimen

UHPC F-13
700
* – Denotes an estimate
X – Denotes data not collected

82

Table 3.8: Details of compressive strength tests of NSC-SF and NSC cylinders
T

Average rate
of heat

f c'

E

Tmax(core)

Tmax(furnace)

°C

°C/min

MPa

GPa

°C

°C

NSC-SF

20

N/A

42

21

N/A

N/A

NSC-SF

200

2.36

26

10

N/A

364

NSC-SF

400

2.11

25

3

N/A

482

NSC-SF

400

3.66

30

5

N/A

526

NSC-SF

600

2.29

19

2

N/A

692

NSC-SF

700

1.48

11

2

N/A

X

NSC

20

N/A

46

24

N/A

N/A

NSC

400

1.93

30

6

N/A

482

NSC

600

2.15

26

2

N/A

691

NSC

700

1.82

24

3

N/A

742

1.49

11

3

N/A

777

Specimen

NSC
750
* – Denotes an estimate
X – Denotes data not collected

83

Table 3.9: Details of compressive strength tests of HSC-PP and HSC cylinders
T

Average rate
of heat

f c'

E

Tmax(core)

Tmax(furnace)

°C

°C/min

MPa

GPa

°C

°C

HSC-PP

20

N/A

112

32

N/A

N/A

HSC-PP

200

1.04

82

31

N/A

277

HSC-PP

200

2.31

73

13

N/A

401

HSC-PP

400

2.29

41

10

N/A

483

HSC-PP

600

1.94

32

4

N/A

696

HSC-PP

700

1.62

24

3

N/A

741

HSC-PP

750

1.86

16

1

N/A

784

HSC

20

N/A

68

27

N/A

N/A

HSC

20

N/A

120

31

N/A

N/A

HSC

200

2.96

72

14

N/A

432

HSC

400

1.90

41

5

N/A

507

2.40

39

6

N/A

664

Specimen

HSC
600
* – Denotes an estimate
X – Denotes data not collected

84

Table 3.10: Details pertaining to UHPC tests conducted at, and above 300°C

ÂşC/cm
24

Time for
maximum
thermal
gradient
min
245

Heating
rate at
time of
spalling
ÂşC/min
2.80

248

32

83

3.42

95

248

29

72

3.43

1.87

99

185

21

96

3.89

200

2.60

80

191

24

67

5.41

A4

300

2.00

102

227

X

X

2.70

A&B

300

0.42

No spalling

301

X

X

N/A

A&B

400

0.50

No spalling

401

X

X

N/A

B5

300

0.80

No spalling

329

7

290

N/A

B1

400

0.77

281

273

10

279

1.60

B2

300

0.70

290

277

14

293

1.97

B12

300

0.70

No spalling

354

9

76

N/A

B4 *

400

0.60

No spalling

430

X

X

N/A

A2

400

0.69

No spalling

434

8

81

N/A

B19

300

0.65

No spalling

340

7

83

N/A

B11 *

400

0.60

No spalling

430

X

X

N/A

D4

400

0.53

No spalling

455

7

84

N/A

D5 *

600

0.88

228

267

4

87

0.48

D6

300

0.98

236

275

20

90

0.63

D7
D8
F1
F5
F8
F9
F12
F13

300
300
600
300
400
600
600
700

0.39
0.27
1.42
0.75
0.74
0.74
1.99
0.69

849
No spalling
177
No spalling
No spalling
No spalling
159
No spalling

284
330
295
330
433
632
296
731

5
X
15
7
8
8
5
10

849
X
91
90
87
85
83
1104

0.48
N/A
1.51
N/A
N/A
N/A
1.63
N/A

T

Average
rate of
heat

Time to
spall

Tcore

Maximum
thermal
gradient

A1

ÂşC
600

ÂşC/min
1.05

min
293

ÂşC
350

A13

300

2.20

125

B8 *

300

2.83

A7 *

400

B3

Specimen

85

Table 3.10 (cont’d)

Specimen

T

Average
rate of
heat

C
ÂşC/min
F14
300
2.45
F15
300
0.21
F16
400
0.23
F17
300
0.25
F18
400
0.24
* – Denotes an estimate

Time to
spall

Tcore

Maximum
thermal
gradient

min
81
No spalling
No spalling
No spalling
No spalling

ÂşC
220
319
421
329
421

ÂşC/cm
13
2
X
1
3

Time for
maximum
thermal
gradient
min
81
661
X
680
714

Heating
rate at
time of
spalling
ÂşC/min
5.83
N/A
N/A
N/A
N/A

X – Denotes data not collected

Table 3.11: Key points from thermal gradient with respect to time graphs for UHPC tests
conducted at, and above 300°C

Average
rate of heat,
ÂşC/min

Average
rate of
heat,
ÂşC/min

Maximum
average
thermal
gradient,
ÂşC/cm

Average
time for
maximum
average
thermal
gradient,
min

r ≥ 1.25

2.17

20

82

32

100

115

239

r < 1.25

0.71

10

212

24

27

266

389

0.65

8

220

10

N/A

N/A

431

0.88

14

199

24

100

266

288

r < 1.25 (No
spalling)
r < 1.25
(Spalling)

r – Denotes average rate of heat

86

Maximum
thermal
gradient,
ÂşC/cm

% of
tests
to
spall

Average
Tavg.(core),
time to
ÂşC
spall, min

Table 3.12: Color legend for thermal gradient with respect to time graphs
Average rate of heat,
ÂşC/min

Color designator

r ≤ 0.70

Any shade of blue

0.70 < r ≤ 1.25

Any shade of green

1.25 < r ≤ 1.75

Any shade of violet

1.75 < r ≤ 2.25

Any shade of orange

2.25 < r ≤ 2.75

Any shade of red

r > 2.75

Any shade of dark red

r – Denotes average rate of heat

87

(a) UHPC specimens’ thermocouple placement

(b) HSC and NSC specimens’ thermocouple placement
Figure 3.1: Prepared concrete specimens
88

Figure 3.1 (cont’d)

(c) Dimensions of tested cylinders

Figure 3.2: Table saw utilized to cut and prepare the concrete cylinders

89

Figure 3.3: Electrical furnace utilized in heating concrete cylinders

Figure 3.4: Forney load controlled strength testing machine to conduct compressive
strength testing as part of this study

90

Figure 3.5: Compressometer with (2) LVDTs and (2) metal rods utilized to develop the
stress-strain relationship as part of this study

(a) Desired heating regime; Adapted from Khaliq (2012)
Figure 3.6: Desired heating rates for high temperature compression tests
91

Figure 3.6 (cont’d)

(b) Desired rate of heat by type of concrete

(c) Desired loading regime; adapted from Khaliq (2012)
92

Figure 3.7: Temperature data acquisition device utilized as part of this study

93

800
700

Temperature, ÂşC

600
500

400
300
200
100
0
0

50

100

150

200

250

300

350

400

Time, min
NSC Surface

NSC Furnace

NSC-SF Surface

NSC-SF Furnace

2 C/min

(a) NSC and NSC-SF concrete cylinders tested at 600ÂşC

800
700

Temperature, ÂşC

600
500
400
300
200
100
0

0

50

100

150

200

250

300

350

400

Time, min
HSC Surface

HSC Furnace

HSC-PP Surface

HSC-PP Furnace

2 C/min

(b) HSC and HSC-PP concrete cylinders tested at 600ÂşC
Figure 3.8: Temperature versus time development for concrete cylinders tested at 600°C

94

Figure 3.8 (cont’d)
700

Temperature, ÂşC

600
500
400
300
200
100
0
0

100

200

300

400

500

600

700

800

900

1000

1100

Time, min
F9 Surface

F9 Furnace

F9 Center

0.5 C/min

(c) UHPC F9 concrete cylinder tested at 600ÂşC

180

UHPC A & B

UHPC D

160

UHPC F

Compressive Strength, MPa

140

HSC-PP
HSC

120

NSC-SF
100
NSC
80
60
40
20
0
0

100

200

300

400
500
Temperature, ÂşC

600

700

800

Figure 3.9: Compressive strength with respect to temperature of various concrete cylinders
tested as part of research
95

40
UHPC A & B

35

UHPC D
UHPC F

Elastic Modulus, GPa

30

HSC-PP
HSC

25
NSC-SF

NSC

20

15

10

5

0
0

100

200

300

400
500
Temperature, ÂşC

600

700

800

Figure 3.10: Elastic modulus with respect to temperature of various concrete cylinders
tested as part of research

96

50
20°C

45

400°C

40

600°C

35

Stress, MPa

700°C

30
750°C

25
20
15
10
5
0
0.00000

0.00500

0.01000

0.01500

0.02000

0.02500

0.03000

Strain, mm/mm

Figure 3.11: NSC stress-strain relationships with respect to temperature

Figure 3.12: Failure modes of NSC at respective temperatures

97

45
20°C

40

200°C

35
400°C - 1

30

Stress, MPa

400°C - 2

25
600°C

20

700°C

15
10
5
0
0.00000

0.00500

0.01000

0.01500

0.02000

0.02500

0.03000

Strain, mm/mm

Figure 3.13: NSC-SF stress-strain relationships with respect to temperature

Figure 3.14: Failure modes of NSC–SF at respective temperatures

98

140

120

Stress, MPa

100

80

60

40

20

0
0.0000

0.0010

0.0020

0.0030

0.0040

0.0050

0.0060

0.0070

Strain, mm/mm
20°C

20°C - 2

200°C

400°C

600°C

Figure 3.15: HSC stress-strain relationships with respect to temperature

Figure 3.16: Failure modes of HSC at respective temperatures

99

120

100

20°C

200°C - 1

200°C - 2

400°C - 1

400°C - 2

600°C

700°C

750°C

Stress, MPa

80

60

40

20

0
0.00000

0.00500

0.01000

0.01500

0.02000

Strain, mm/mm

Figure 3.17: HSC-PP stress-strain relationships with respect to temperature

100

Figure 3.18: Failure modes of HSC – PP at respective temperatures

101

160
140

Stress, MPa

120
100
80
60
40
20
0
0

0.002

0.004

0.006

0.008

0.01

0.012

Strain, mm/mm
A12 20°C

A2 20°C

A7 20°C

A 100°C

A14 100°C

A3 100°C

A9 100°C

B14 100°C

A4 200°C

B12 300°C

B19 300°C

A12 300°C

B4 400°C

A2 400°C

B11 400°C

A13 400°C

Figure 3.19: UHPC A & B stress-strain relationships with respect to temperature

Figure 3.20: Failure modes of UHPC A & B at respective temperatures

102

180

160
140

Stress, MPa

120
D1 20°C

100

D2 100°C

80
D3 200°C

60
D8 300°C

40
D4 400°C

20
0
0.00000

0.00200

0.00400

0.00600

0.00800

0.01000

0.01200

0.01400

Strain, mm/mm

Figure 3.21: UHPC D stress-strain relationships with respect to temperature

Figure 3.22: Failure modes of UHPC D at respective temperatures

103

180
160
140

Stress, MPa

120

100
80
60
40
20
0
0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

0.016

0.018

Strain, mm/mm
F4 20°C

F11 20°C

F21 20°C

F2 100°C

F6 100°C

F20 100°C

F3 200°C

F7 200°C

F10 200°C

F5 300°C

F15 300°C

F17 300°C

F8 400°C

F16 400°C

F18 400°C

F9 600°C

F13 700°C

Figure 3.23: UHPC F stress-strain relationships with respect to temperature

Figure 3.24: Failure modes of UHPC F at respective temperatures

104

Figure 3.25: Damage to furnace due to explosive spalling

105

Figure 3.26: UHPC specimens following experiencing explosive spalling in electrical furnace

106

Figure 3.27: Specimen UHPC B5 displaying water escaping concrete following 300°C test

107

Figure 3.28: Relation between rate of heating with respect to either, (1) the spalling
temperature, or (2) the maximum center temperature attained

Thermal Gradient, ÂşC/cm

45
40

A 100 C

B14 100 C

35

B3 200 C

A13 300 C

30

B8 300 C

B5 300 C

25

B2 300 C

B12 300 C

20

B19 300 C

A7 400 C

15

B1 400 C

A2 400 C

A1 600 C

B11 400 C

10
5
0
-5

0

100

200

300

400

500

600

700

800

Time, min

Figure 3.29: Thermal gradient with respect to time for UHPC A & B cylinders

108

900

D2 100 C

20

Thermal Gradient, ÂşC/cm

D4 400 C
D5 600 C

15

D6 300 C
D7 300 C
D8 300 C

10

5

0
0

100

200

300

400

-5

500

600

700

800

900

1000

Time, min

Figure 3.30: Thermal gradient with respect to time for UHPC D cylinders

20

F2 100 C
F20 100 C
F3 200 C
F7 200 C
F10 200 C
F5 300 C
F15 300 C
F14 300 C
F8 400 C
F18 400 C
F1 600 C
F9 600 C
F12 600 C
F13 700 C

18

Thermal Gradient, ÂşC/cm

16
14
12
10
8
6

4
2
0
-2

0

100

200

300

400

500

600

700

800

900

1000 1100 1200 1300 1400 1500

Time, min

Figure 3.31: Thermal gradient with respect to time for UHPC F cylinders

109

3.5 Summary
Material tests were conducted to characterize the compressive strength, stiffness, and
stress-strain relationship with increasing temperatures. All concretes evaluated underwent
mechanical degradation with increasing temperatures, and to conduct successful tests of UHPC
above 200°C, lower heating rates were utilized. Variables incorporated within the design mixes
included, (1) addition of fibers (steel or polypropylene), (2) no fiber content, (3) varying the steel
fiber content, and (4) a hybrid fiber mix (steel and polypropylene). The inclusion of fibers does
not increase the strength. However, addition of steel fibers increases the ductility, and addition of
polypropylene fibers influences the stiffness. Finally, all design mixes of UHPC experienced fire
induced spalling, while none for conventional concretes. UHPC can sustain explosive spalling at
temperatures less than 200°C, and minimizing the heating rate does not correspond to complete
mitigation of explosive spalling.

110

CHAPTER FOUR
4. COMPARISON OF PROPERTIES OF UHPC WITH CONVENTIONAL CONCRETE
AND BUILDING CODES
4.1 General
UHPC’s design premise allows for excellent interfacial transition zones because it is
developed upon minimal use of water, fine aggregate (limestone and silica sand), and profuse
amount of silica fume and cement. These developmental qualities allow UHPC to have ambient
condition mechanical properties superior to HSC and NSC. While at elevated temperatures, the
rate of degradation of concrete is proportional to its strength or density. Consequently, the low
permeability of UHPC, hailed for its mechanical properties and durability prowess, produces
calamitous results with increasing temperatures. The compressive strength, stress-strain behavior,
and elastic modulus of the three different design mixes of UHPC, A & B, D, and F, will be
compared to the conventional concretes along with constitutive relationships provided as per
ASCE (1992) and Eurocode 2 (2004). Initially the data will be compared at ambient conditions,
followed by a comparison at elevated temperatures.
4.2 Comparison of room temperature properties
As compared to UHPC, the NSC evaluated as part of this study contains over 3.5 times the
w/c ratio, coarse aggregate and contains no silica fume. While HSC’s w/c ratio is 2 times more, it
utilizes coarse aggregate and has 20% of the silica fume of UHPC. NSC and NSC-SF have the
least compressive strength and elastic modulus due to their high w/c ratio. The two design mixes
of NSC are on average only 30% and 60% of UHPC’s compressive strength and stiffness,
respectively. Concurrently, due to the higher w/c ratio and small amount of silica fume combined
with coarse aggregate, HSC and HSC-PP possess mechanical properties less than the UHPC tested.
111

The two design mixes of HSC are on average 81% of UHPC. Moreover, the peak strain achieved
for all concretes are within a 0.0035 to 0.0050 mm/mm range, with UHPC F displaying the greatest
ductility, closely followed by NSC-SF and HSC. Similar to UHPC F, conventional concretes with
fibers produce the superior ductility. Lastly, regarding stiffness UHPC A & B with the higher
amount of steel fiber achieved the greatest value, and the inclusion of polypropylene fibers with
HSC-PP and UHPC F proved to increase the stiffness as well.
4.3 Comparison of elevated temperature properties
4.3.1 Comparison of UHPC with conventional concretes tested
As seen in Figure 4.1, owing to each type of concrete’s unique design, the rate of
degradation is not consistent. Due to UHPC’s maximized packing density, its compressive strength
with increasing temperatures is quite different versus the conventional concretes tested. Moreover,
due to implementing a heating rate that deviated from RILEM standard’s recommendation, directly
comparing UHPC’s mechanical properties with the conventional concretes at and above
temperatures of 200°C is abstruse. The comparison is further complicated by only UHPC F having
successful tests conducted at 600 and 700°C. Given the above variables, the comparisons are as
follows.
Regarding the relative compressive strength (Figure 4.1), tests conducted through 400°C, all
UHPC design mixes remain relatively close to their respective original strengths. The expected
strength loss from calcium-hydroxide at temperatures of 400°C is immensely reduced due to high
contents of silica fume in each UHPC design mix. While the conventional concretes experience a
steady degradation, HSC design mixes suffer a steeper decline. Beyond 500°C the quartz within
UHPC will further exacerbate the degradation, and at temperatures greater than 700°C UHPC will
encounter strength loss due to the decomposition of the cement paste. The conventional concretes’
112

strength suffers similarly to with the latter, however, UHPC’s will be much higher due to the higher
contents of cement. Another similarity seen in Figure 4.1 is that the inclusion or increasing the
contents of steel fiber does not correlate to greater compressive strength. Lastly, as seen by UHPC
F and HSC-PP, the addition of polypropylene fibers slightly compromises the compressive
strength.
Contrary to the differences between the compressive strength, the relative modulus of
elasticity (Figure 4.2) for all concretes tested, degrade in a similar fashion with increasing
temperatures. Each type of concrete’s elastic modulus drops to 50%, ranging from 200 to 250°C.
Moreover, with respect to the concretes tested up to 300°C, the addition of steel fibers did not
contribute to NSC’s stiffness, while the inclusion of steel (UHPC A & B and D), polypropylene
(HSC-PP), or polypropylene and steel fibers (UHPC F) did assist the modulus of elasticity.
Although the relative elastic modulus for the various concrete degraded similarly, the relative
stress-strain relationships (Figure 4.3) demonstrated differences.
At temperatures up to 400°C, the ductility of the tested conventional concretes was
comparable to the UHPC design mixes (twice the ambient condition peak strain). UHPC D4
(400°C) withstood the greatest ductility amongst all design mixes of UHPC with 3.25 times its
original peak strain. The addition of steel or polypropylene fibers with the UHPC and conventional
concretes improved the ductility with increasing temperatures. Though UHPC F9 (600°C)
displayed greater peak strain as compared to the conventional concretes, it was 20% less than
UHPC D4. Furthermore, UHPC F13’s (700°C) relative peak strain was 30% less than the tested
conventional concretes. The drastic reduction in peak strain of UHPC F at 600 and 700°C may be
attributed to the lengthy duration of fire exposure involved to reach the targeted temperatures.

113

4.3.2 Comparison of UHPC with building codes
Through 400°C, the relative compressive strengths (Figure 4.1) of the code conventional
concretes display minimal degradation and behave comparably as the tested UHPC. Furthermore,
UHPC F’s relative strengths at 600 and 700°C are analogous to the constitutive relationships
dictated by the codes as well. Along with the strengths, the elastic modulus (Figure 4.2), with
increasing temperatures, displays minimal variation between the UHPC and code conventional
concretes. However, the relative peak strain recommended by the codes is nearly 4.5 times the
tested UHPC cylinders at a temperature range of 400 to 700°C.
4.4 Comparison of spalling
4.4.1 Comparison of UHPC with conventional concretes tested
Though the HSC investigated as part of this study contains 7% silica fume, as compared to
cement weight, and no polypropylene fibers, it did not result in explosive spalling. In fact, no
conventional concrete evaluated encountered fire induced spalling. Whereas all UHPC design
mixes considered as part of this research experienced explosive spalling with heating rates as low
as 0.40°C/min. The occurrence of explosive spalling at such minimal heating rates implies the
principle cause is the pore pressure build-up due to the dense microstructure of UHPC.
4.4.2 Comparison of UHPC with building codes
Design mix UHPC F contains nearly 45% silica fume to cement weight ratio and as
prescribed by Eurocode 2 (2004) for HSC, 2 kg/m3 of polypropylene fiber were incorporated to
mitigate fire induced spalling. However, three UHPC F cylinders suffered an explosion with
heating rates as low as 1.40°C/min. Though the high content of silica fume lowered the C/S ratio,
it gravely reduced the permeability and increased the main binder for UHPC, C-S-H.

114

1.25

UHPC A & B
UHPC D
UHPC F

1.00

HSC-PP

Relative Compressive Strength

HSC
NSC-SF
NSC

0.75

ASCE
EC-Siliceous
EC Carbonate

0.50

0.25

0.00
0

100

200

300

400
500
Temperature, ÂşC

600

700

800

900

Figure 4.1: Relative compressive strength with respect to temperature of various concrete
cylinders tested as part of research, and Eurocode 2 (2004) and ASCE (1992)

115

1.00
UHPC A & B

UHPC D
UHPC F
HSC-PP

0.75

Relative Elastic Modulus

HSC
NSC-SF
NSC
ASCE

0.50

EC-Siliceous
EC-Carbonate

0.25

0.00
0

100

200

300

400
500
Temperature, ÂşC

600

700

800

900

Figure 4.2: Relative elastic modulus with respect to temperature of various concrete
cylinders tested as part of research, and Eurocode 2 (2004) and ASCE (1992)

116

1.25
NSC 20°C
NSC 400°C
NSC 600°C
NSC 700°C

1.00

NSC 750°C

Relative Stress, fc',T/fc'

NSC-SF 20°C

NSC-SF 200°C
0.75

NSC-SF 400°C
NSC-SF 600°C

NSC-SF 700°C
0.50

0.25

0.00
0

1

2

3

4
5
Relative Strain, ÎľT/Îľ

6

7

8

(a) Relative stress-strain relationship for NSC and NSC-SF concrete cylinders tested
Figure 4.3: Relative stress-strain relationships for concrete cylinders tested

117

Figure 4.3 (cont’d)
1.25
HSC 20°C
HSC 200°C
HSC 400°C
1.00

HSC 600°C
HSC-PP 20°C

Relative Stress, fc',T/fc'

HSC-PP 200°C
HSC-PP 400°C
0.75

HSC-PP 600°C
HSC-PP 700°C
HSC-PP 750°C

0.50

0.25

0.00
0

1

2

3

4
5
Relative Strain, ÎľT/Îľ

6

7

8

(b) Relative stress-strain relationship for HSC and HSC-PP concrete cylinders tested

118

Figure 4.3 (cont’d)
1.25

Relative Stress, fc',T/fc'

1.00

0.75

0.50

0.25

0.00
0

1

2

3

4
5
Relative Strain, ÎľT/Îľ

6

UHPC A & B 20°C

UHPC B14 100°C

UHPC A4 200°C

UHPC A12 300°C

UHPC A13 400°C

UHPC D1 20°C

UHPC D2 100°C

UHPC D3 200°C

UHPC D8 300°C

7

8

UHPC D4 400°C

(c) Relative stress-strain relationship for UHPC A & B and D concrete cylinders tested

119

Figure 4.3 (cont’d)
1.25
UHPC F4 20 C
UHPC F11 20°C
UHPC F2 100°C

1.00

UHPC F3 200°C

Relative Stress, fc',T/fc'

UHPC F5 300°C
UHPC F8 400°C

0.75

UHPC F9 600°C
UHPC F13 700°C
0.50

0.25

0.00
0

1

2

3

4
5
Relative Strain, ÎľT/Îľ

6

7

(d) Relative stress-strain relationship for UHPC F concrete cylinders tested

120

8

Figure 4.3 (cont’d)
1.25

Relative Stress, fc',T/fc'

1.00

0.75

0.50

0.25

0.00
0

1

2

3

4

5

6

7

8

9

10

11

Relative Strain, ÎľT/Îľ
20°C ASCE - NSC

100°C ASCE - NSC

200°C ASCE - NSC

300°C ASCE - NSC

400°C ASCE - NSC

500°C ASCE - NSC

600°C ASCE - NSC

700°C ASCE - NSC

800°C ASCE - NSC

(e) Relative stress-strain relationship for NSC from ASCE (1992)

121

12

13

Figure 4.3 (cont’d)
1.25

Relative Stress, fc',T/fc'

1.00

0.75

0.50

0.25

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

Relative Strain, ÎľT/Îľ
20°C EC

100°C EC

200°C EC

300°C EC

400°C EC - Sil.

400°C EC - Calc.

500°C EC - Sil.

500°C EC - Calc.

600°C EC - Sil.

600°C EC - Calc.

700°C EC - Sil.

700°C EC - Calc.

800°C EC - Sil.

800°C EC - Calc.

(e) Relative stress-strain relationship for siliceous and carbonate concrete from Eurocode 2
(2004)

122

4.5 Summary
Comparisons were presented from the experimental testing of UHPC and conventional
concretes, and UHPC with the constitutive relationships recommended by ASCE (1992) and
Eurocode 2 (2004). As expected due to UHPC’s contrasting composition, there are differences
regarding the properties of UHPC and conventional concretes. Furthermore, UHPC’s strength
degradation was comparable to the conventional concretes as prescribed by the codes. However,
there is a vast variation of the peak strains between UHPC and code conventional concretes. Lastly,
following the recommendations as prescribed by Eurocode 2 (2004) to mitigate fire induced
spalling was not successful.

123

CHAPTER FIVE
5. CONCLUSIONS AND RECOMMENDATIONS
5.1 General
The experimental investigation conducted as part of this thesis provides data pertaining to
the high temperature behavior of UHPC. Thereby providing the necessary data to evaluate the fire
response of UHPC structural members. The mechanical properties tested from 20-700°C were
compressive strength, stress-strain relationship, and elastic modulus of three different design mixes
of UHPC consisting of varying fiber type and content, and four different conventional types of
concrete were tested for comparative purposes. Furthermore, differing heating rates were
employed to study the explosive spalling nature of UHPC. Empirical relations depicting the
mechanical behavior with respect to temperature, and prominent data of fire induced spalling are
presented.
5.2 Key findings
Based on the experimental research accumulated as part of this thesis, the following
conclusions can be derived:
1. To conduct successful testing above 200°C of all UHPC design mixes heating rates less
than 1°C/min were utilized, otherwise explosive spalling occurred. Furthermore, UHPC F
sustained three explosive spalling occurrences, deeming 2.00 kg/m 3 of polypropylene
fibers not sufficient to mitigate the abovementioned.
2. Fire induced spalling occurred with heating rates less than 0.50°C/min, implying the root
cause of explosive spalling is the built-up pore pressure owing to UHPC’s dense medium.

124

3. UHPC’s compressive strength varies closely to its original strength up to 400°C, thereby
degrading differently with temperature as compared to conventional concretes.
4. The steel fiber content does not influence the compressive strength of UHPC, but it does
improve the ductility. Additionally, the combination of steel and polypropylene fibers
(UHPC F) improves the stiffness up to 200°C
5. There is a vast disconnect for experimental data and recommendations prescribed in codes.
Specifically, mechanical properties such as compressive strength and stress-strain
behavior, displayed poorer fire performance as compared to their respective codes
counterpart.
5.3 Recommendations for future research
Although all three design mixes of UHPC suffered explosive spalling, the literature
contains various studies that prove contrary. Investigations have employed heating rates of
10°C/min and evaluated UHPC specimens up to 1000°C. Furthermore, UHPC mechanical
properties considered at the respective elevated temperatures, as opposed to the residual-state, is
quite scarce. Therefore, refinement of the design mix to be tested at the elevated temperature state
is of the utmost importance. Below are the author’s recommendations for future research:
▪

Formulate a design mix that would utilize a combination of silica fume, cement, and fly
ash or GGBFS and evaluate the mechanical properties at elevated temperatures with
varying heating rates

▪

Evaluate the mechanical properties at elevated temperatures with varying heating rates,
utilizing a design mix with the above mentioned basic formula and vary the fiber content
as follows:
i.

Steel fiber

125

▪

ii.

Steel and polypropylene fibers

iii.

Steel and polypropylene fibers and vary the polypropylene fibers dimension

Evaluate the thermal properties at elevated temperatures utilizing a design mix with the
above mentioned basic formula

▪

Further investigation of fire induced spalling at differing heating rates

5.4 Research impact
It is evident that the data generated as part of this study is vital, and will provoke a swift
impression upon the implementation of UHPC for infrastructure applications. By virtue of UHPC’s
microstructure, it possesses consummate mechanical and durability properties. However, the
preceding properties cause UHPC to be susceptible to fire induced spalling, and such an occurrence
can inhibit a structure reaching its fire resistance requirement. Through the course of testing, this
limiting facet could not be minimized by additional steel fiber or inclusion of polypropylene fiber
as recommended in codes, and results as such emphasize the continual evaluation of UHPC’s
properties at time of fire exposure. Therefore, to avoid further loss, supplementary research must
be conducted evaluating high temperature properties of UHPC.

126

APPENDIX

127

Appendix High temperature properties of UHPC
Table A1: Additional notes pertaining to UHPC A & B specimen tests

T
Specimen

Average
rate of
heating

Comments

°C

°C/min

UHPC B9

20

N/A

None

UHPC A12

20

N/A

None

UHPC A2

20

N/A

None

UHPC A7

20

N/A

None

UHPC A

100

0.60 *

UHPC A

100

0.40

None

UHPC B14

100

1.37

None

UHPC A3

100

0.30*

None

UHPC A9

100

1.00*

None

UHPC A14

100

2.00*

None

UHPC A & B

200

0.70*

No data acquisition utilized. Average rate of heat is an
estimation based off time

UHPC A6

200

0.54

None

UHPC A & B

200

0.75*

UHPC B12

300

0.70

UHPC B19

300

0.65

None

UHPC A12

300

0.42

None

UHPC B4

400

0.60*

Data acquisition turned off; average rate of heat is an
estimate based off a similar test

UHPC A2

400

0.69

None

UHPC A13

400

0.50

None

UHPC B11

400

0.60

Center thermocouple was not working

No data acquisition utilized. Average rate of heat is an
estimation based off time

No data acquisition utilized. Average rate of heat is an
estimation based off time
Surface thermocouple readings were lower than center
thermocouple; therefore for therm grad, utilized the
average of the center and furnace for the surface temp

* – Denotes an estimate
128

Table A2: Additional notes pertaining to UHPC D specimen tests

T
Specimen

Average
rate of
heating

Comments

°C

°C/min

UHPC D-1

20

N/A

None

UHPC D-2

100

1.60

None

UHPC D-3

200

1.13

Surface and furnace thermocouples providing erroneous
values

UHPC D-8

300

0.27

None

UHPC D-4

400

0.53

None

* – Denotes an estimate

129

Table A3: Additional notes pertaining to UHPC F specimen tests
T
Specimen

Average rate
of heating

Comments

°C

°C/min

UHPC F-4

20

N/A

None

UHPC F-11

20

N/A

None

UHPC F-19

20

N/A

None

UHPC F-21

20

N/A

None

UHPC F-2

100

1.64

UHPC F-6

100

2.66

UHPC F-20

100

1.65

UHPC F-3

200

1.89

UHPC F-7

200

1.76

None
Center thermocouple values were void; data is regarding
the surface thermocouple
None
Surface thermocouple readings were lower than
expected; therefore, for thermal grad, utilized the average
of the center and furnace for the surface temp
None

UHPC F-10

200

0.91

None

UHPC F-5

300

0.75

UHPC F-15

300

0.21

UHPC F-17

300

0.25

UHPC F-8

400

0.74

UHPC F-16

400

0.23

UHPC F-18

400

0.24

UHPC F-9

600

0.74

None
Testing the compressive strength with reduced rate of
heating
Testing the compressive strength with reduced rate of
heating
None
Testing the compressive strength with reduced rate of
heating
Testing the compressive strength with reduced rate of
heating
Surface thermocouple readings were lower than
expected; therefore, for thermal grad, utilized the average
of the center and furnace for the surface temp
None

UHPC F-13 700
* – Denotes an estimate

0.69

130

REFERENCES

131

REFERENCES

1. Abdul-Hussain, S.T. (2013). "Effect of Elevated Temperatures on Compressive and Tensile
Strengths of Reactive Powder Concrete." Journal of Engineering and Development, vol. 17.4:
pp. 259-278.
2. Abrams, D.A. (1919). “Design of concrete mixtures.” Structural Materials Research Laboratory,
Lewis Institute.
3. Al-Jabri, K.S., Waris, M.B., Al-Saidy, A.H. (2016). "Effect of aggregate and water to cement
ratio on concrete properties at elevated temperature: Effect of Aggregate and Water to Cement
Ratio on Concrete Properties." Fire and Materials, vol. 40(7): pp. 913-925.
4. Ali, M.H., Dinkha, Y.Z., Haido J.H. (2016). "Mechanical properties and spalling at elevated
temperature of high performance concrete made with reactive and waste inert powders."
Engineering Science
and Technology, an International Journal, vol. 2: pp. 536-541.
5. Alkaysi, M. (2016). “Strength and Durability of Ultra-High Performance Concrete Materials
and Structures.” Doctoral dissertation, University of Michigan.
6. Arioz, O. (2007). "Effects of elevated temperatures on properties of concrete." Fire Safety
Journal, vol. 42(8): pp. 516-522.
7. ASCE (1992). “Structural fire protection.” American Society of Civil Engineers Manuals and
Reports on Engineering Practice, 78, New York.
8. Aslani, F., Bastami, M. (2011). "Constitutive Relationships for Normal- and High-Strength
Concrete at Elevated Temperatures." ACI Materials Journal, vol. 108(4): pp, 355.
9. ASTM C39/C39M (2016). “Standard Test Method for Compressive Strength of Cylindrical
Concrete Specimens.” American Society for Testing and Materials, West Conshohocken, PA.
10. ASTM C293/C293M (2016). “Standard Test Method for Flexural Strength of Concrete (Using
Simple Beam with Center-point Loading.” American Society for Testing and Materials, West
Conshohocken, PA.
11. ASTM C469/C469M (2014). “Standard Test Method for Static Modulus of Elasticity and
Poisson’s Ratio of Concrete in Compression.” American Society for Testing and Materials,
West Conshohocken, PA.

132

12. ASTM C496/C496M (2011). “Standard Test Method for Splitting Tensile Strength of
Cylindrical Concrete Specimens.” American Society for Testing and Materials, West
Conshohocken, PA.
13. ASTM C1113/C1113M (2013). “Standard Test Method for Thermal Conductivity of
Refractories by Hot Wire (Platinum Resistance Thermometer Technique).” American Society
for Testing and Materials, West Conshohocken, PA.
14. ASTM E119a (2008). “Standard Methods of Fire Test of Building Construction and Materials.
” American Society for Testing and Materials, West Conshohocken, PA.
15. ASTM E831 (2014). “Standard Test Method for Linear Thermal Expansion of Solid Materials
by Thermomechanical Analysis.” American Society for Testing and Materials, West
Conshohocken, PA.
16. ASTM E1269 (2011). “Standard Test for Determining Specific Heat Capacity by Differential
Scanning Calorimetry.” American Society for Testing and Materials, West Conshohocken, PA.
17. ASTM E1530 (2016). “Standard Test Method for Evaluating the Resistance to Thermal
Transmission of Materials by the Guarded Heat Flow Meter Technique.” American Society
for Testing and Materials, West Conshohocken, PA.
18. ASTM E1868 (2015). “Standard Test Methods for Loss-On-Drying by Thermogravimetry.”
American Society for Testing and Materials, West Conshohocken, PA.
19. Aydin, S., Baradan, B. (2013). “Engineering Properties of Reactive Powder Concrete without
Portland Cement." ACI Materials Journal, vol. 110(6): pp. 619-627.
20. Aydin, S., Baradan, B. (2012). "High Temperature Resistance of Alkali-Activated Slag- and
Portland Cement Based Reactive Powder Concrete." ACI Materials Journal, vol. 109(4): pp.
463-470.
21. Aziz, E.M. (2015). “Response of fire exposed steel bridge girders.” Doctoral dissertation,
Michigan State University.
22. Bashandy, A.A. (2013). "Influence of elevated temperatures on the behavior of economical
reactive powder concrete." Journal of Civil Engineering Research, vol. 3.3: pp. 89-97.
23. Behloul, M., Chanvillard, G., Casanova, P., Orange, G., France F.F. (2002). "Fire resistance
of DuctalÂŽ ultra high performance concrete." 1st fib Congress, vol. 1: pp. 421-430.

133

24. Bei, S., Zhixiang, L. (2016). "Investigation on spalling resistance of ultra-high-strength
concrete under rapid heating and rapid cooling." Case Studies in Construction Materials, vol.
4: pp. 146-153.
25. Buchanan, A.H., Abu, A.K. (2017). “Structural Design for Fire Safety.” John Wiley & Sons,
Chichester, West Sussex, United Kingdom.
26. Burke, B.T. (2011). "Residual Strength of Ultra-High Performance Concrete After Exposure
to Elevated Temperatures." Master thesis, University of Connecticut.
27. Canbaz, M. (2014). "The Effect of High Temperature on Reactive Powder Concrete."
Construction and Building Materials, vol. 70: pp. 508-513.
28. Chang, Y.F., Chen, Y.H., Sheu, M.S., Yao, G.C. (2006). "Residual stress–strain relationship
for concrete after exposure to high temperatures." Cement and Concrete Research, vol. 36(10):
pp. 1999-2005.
29. Chen, G.M., Yang, H., Lin, C.J., Chen, J.F., He, Y.H., Zhang, H.Z. (2016). "Fracture behaviour
of steel fibre reinforced recycled aggregate concrete after exposure to elevated temperatures."
Construction and Building Materials, vol. 128: pp. 272-286.
30. Chen, X., Chen, W., Rougelot, T., Davy, C.A., Agostini, F., Skoczylas, F., Bourbon, X. (2009).
"Experimental Evidence of a Moisture Clog Effect in Cement-Based Materials Under
Temperature." Cement and Concrete Research, vol. 39(12): pp. 1139-1148.
31. Demirel, B., Keleştemur, O. (2010). "Effect of elevated temperature on the mechanical
properties of concrete produced with finely ground pumice and silica fume." Fire Safety
Journal, vol. 45(6): pp. 385-391.
32. Dehn, F. (2004). “Fire Resistance of Ultra High Performance Concrete (UHPC) – Testing of
Laboratory Samples and Columns
under Load.” International Symposium on Ultra
High Performance Concrete, Kassel, Germany, September 13-15.
33. Dwaikat, M. B., Kodur, V.K.R. (2009). "Hydrothermal Model for Predicting Fire-Induced
Spalling in Concrete Structural Systems." Fire Safety Journal, vol. 44(3): pp. 425-434.
34. Elwell, D.J., Fu, G. (1995). “Compression testing of concrete: cylinders vs. cubes”. No.
FHWA/NY/SR-95/119.Albany, New York, NY, USA: Transportation Research and
Development Bureau, New York State Department of Transportation.
35. ErgĂźn, A., KĂźrklĂź, G., Serhat, M.B., Mansour, M.Y. (2013). "The effect of cement dosage on
mechanical properties of concrete exposed to high temperatures." Fire Safety Journal, vol. 55:
pp. 160-167.
134

36. EC2. (2004). “EN 1992-1-2: Design of concrete structures, Part 1-2: General Rules-Structural
Fire Design.” European Committee for Standardization, Brussels, Belgium.
37. Felicetti R., Gambarova, P.G., Sora, M.N., Khoury, G.A. (2000). "Mechanical behaviour of
HPC and UHPC in direct tension at high temperature and after cooling." Fifth RILEM
symposium on fibre-reinforced concretes, vol. N: pp. 749-758.
38. Felicetti, R., Lo Monte, F. (2013). "Concrete Spalling: Interaction between Tensile Behaviour
and Pore Pressure during Heating." MATEC Web of Conferences, vol. 6: pp. 3001.
39. Gesoglu, M., GĂźneyisi, E., Nahhab, A.H., YazÄącÄą, H. (2015). "Properties of Ultra-High
Performance Fiber Reinforced Cementitious Composites made with Gypsum-Contaminated
Aggregates and Cured at Normal and Elevated Temperatures." Construction and Building
Materials, vol. 93: pp. 427-438.
40. Ghafari, E., Costa, H. and JĂşlio, E. (2015). "Statistical mixture design approach for ecoefficient UHPC." Cement and
Concrete Composites, vol. 55: pp. 17-25.
41. Ghandehari, M., Behnood, A., Khanzadi, M. (2010). "Residual Mechanical Properties of HighStrength Concretes after Exposure to Elevated Temperatures." Journal of Materials in Civil
Engineering, vol. 22(1): pp. 59-64.
42. Graybeal, B.A. (2015). "Compression Testing of Ultra-High-Performance Concrete."
Advances in Civil Engineering Materials, vol. 4.2: pp. 102-112.
43. Graybeal, B.A. (2014). "Tensile Mechanical Response of Ultra-High-Performance
Concrete." Advances in Civil Engineering Materials, vol. 4.2: pp. 62-74.
44. Graybeal, B., Davis, M. (2008). "Cylinder Or Cube: Strength Testing of 80 to 200 MPa (11.6
to 29 Ksi) Ultra-High-Performance Fiber-Reinforced Concrete." ACI Materials Journal, vol.
105(6): pp. 603-609.
45. Gu, C.P., Ye, G., Sun, W. (2015). "Ultrahigh Performance Concrete-Properties, Applications
and Perspectives." Science China Technological Sciences, vol. 58(4): pp. 587-599.
46. Hager, I., Tracz, T., Śliwiński, J., Krzemień, K. (2016). "The influence of aggregate type on
the physical and mechanical properties of high-performance concrete subjected to high
temperature: Influence of aggregate type on HPC subjected to high temperature." Fire and
Materials, vol. 40(5): pp. 668-682.
47. Hager, I., Zdeb, T., Krzemień, K. (2013). "The Impact of the Amount of Polypropylene Fibres
on Spalling Behaviour and Residual
Mechanical Properties of Reactive Powder
Concretes." MATEC Web of Conferences, vol. 6: pp. 2003.
135

48. Hassan, A.M.T., Jones, S.W., Mahmud, G.H. (2012). "Experimental test methods to determine
the uniaxial tensile and compressive behaviour of ultra high performance fibre reinforced
concrete (UHPFRC).” Construction and Building Materials, vol. 37: pp. 874-882.
49. Heinz, D., Dehn, F., Urbonas, L. (2004). “Fire Resistance of Ultra High Performance Concrete
(UHPC) – Testing of Laboratory Samples and Columns under Load.” International
Symposium on Ultra High Performance Concrete, Kassel, Germany, September 13-15.
50. Hertz, K.D. (2005). "Concrete Strength for Fire Safety Design." Magazine of Concrete
Research, vol. 57(8): pp. 445-453.
51. Hertz, K.D. (2003). "Limits of Spalling of Fire-Exposed Concrete." Fire Safety Journal, vol.
38(2): pp. 103-116.
52. Hertz, K.D., Sørensen, L.S. (2005). "Test Method for Spalling of Fire Exposed Concrete." Fire
Safety Journal, vol. 40(5): pp. 466-476.
53. Hertz, K. (1984). “Explosion of silica-fume concrete.” Fire Safety Journal, vol. 8(1): pp. 77.
54. Ingason, H. (2004) "Fire development in catastrophic tunnel fires (CTF)." SP RAPPORTSTATENS PROVNINGSANSTALT, International Symposium on Catastrophic Tunnel Fires
(CTF), Boras, Sweden, November 20-21, 2003: pp. 31-48.
55. Jagannathan,
D.
(2013). “Calcium-Silicate-Hydrate
in
cementitious
systems:
chemomechanical correlations, extreme temperature behavior, and kinetics and morphology
of in-situ formation.” Doctoral dissertation, Massachusetts Institute of Technology.
56. Jang, H., So, H., So, S. (2016). The properties of reactive powder concrete using PP fiber and
pozzolanic materials at elevated temperature.” Journal of Building Engineering, vol. 8: pp.
225-230.
57. Ju, Y., Wang, L., Liu, H., Tian, K. (2015). "An Experimental Investigation of the Thermal
Spalling of Polypropylene-Fibered Reactive Powder Concrete Exposed to Elevated
Temperatures." Science China Technological Sciences, vol. 23: pp. 2022-2053.
58. Ju, Y. Liu, H., Liu, J., Tian, K., Wei, S., Hao, S. (2011). "Investigation on Thermophysical
Properties of Reactive Powder Concrete." Science China Technological Sciences, vol. 54(12):
pp. 3382-3403.
59. Ju, Y., Liu, H., Tian, K., Wang, L., Ge, Z. (2013). "An Investigation on Micro Pore Structures
and the Vapor Pressure Mechanism of Explosive Spalling of RPC Exposed to High
Temperature." Science China Technological Sciences, vol. 56(2): pp. 458-470.

136

60. Kalifa, P., Menneteau, F.D., Quenard, D. (2000). "Spalling and Pore Pressure in HPC at High
Temperatures." Cement and Concrete Research, vol. 30(12): pp. 1915-1927.
61. Kamen, A., DenariĂŠ, E., BrĂźhwiler, E. (2007). "Thermal Effects on Physico-Mechanical
Properties of Ultra-High-Performance Fiber-Reinforced Concrete." ACI Materials Journal,
vol. 104(4): pp. 415-423.
62. Khaliq, W. (2012). “Performance characterization of high performance concretes under fire
conditions.” Doctoral dissertation, Michigan State University.
63. Khoury, G.A. (2000). "Effect of fire on concrete and concrete structures." Progress in
Structural Engineering and Materials, vol. 2.4: pp. 429-447.
64. Kim, Y., Kim, G., Lee, T. (2009). "Mechanical properties of high-strength concrete subjected
to high temperature by stressed test." Transactions of Nonferrous Metals Society of China, vol.
19: pp. s128-s133.
65. Klingsch, E. (2014). "Explosive spalling of concrete in fire." Doctoral dissertation, ETH
Zurich.
66. Klingsch, E., Frangi, F. (2011). "Experimental analysis on changes of the porosity of ultrahigh
performance concrete at elevated temperatures." 2nd International RILEM Workshop on
Concrete Spalling due to Fire Exposure. RILEM Publications SARL.
67. Ko, J., Ryu, D., Noguchi, T. (2011). "The Spalling Mechanism of High-Strength Concrete
Under Fire." Magazine of Concrete Research, vol. 63(5): pp. 357-370.
68. Kodur, V. (2014). “Properties of Concrete at Elevated Temperatures.” ISRN Civ. Eng. pp. 115.
69. Kodur, V.K.R., Phan, L. (2007). "Critical Factors Governing the Fire Performance of High
Strength Concrete Systems." Fire Safety Journal, vol. 42(6): pp. 482-488.
70. Khaliq, W., Kodur, V.K.R. (2011). "Effect of High Temperature on Tensile Strength of
Different Types of High-Strength Concrete." ACI Materials Journal, vol. 108(4): pp. 394.
71. Kodur, V., Khaliq, W. (2011). "Effect of Temperature on Thermal Properties of Different
Types of High-Strength Concrete." Journal of Materials in Civil Engineering, vol. 23(6): pp.
793-801.
72. Kosmatka, S.H. Wilson, M.L. (2011). “Design and Control of Concrete Mixtures.” EB001,
15th edition, Portland Cement Association, Skokie, Illinois, USA, 460 pages.

137

73. Kusumawardaningsih, Y., Fehling, E., Ismail, M., Aboubakr, A.A.M. (2015). Tensile strength
behavior of UHPC and UHPFRC. Procedia Engineering, vol. 125: pp. 1081-1086.
74. Li, H., Liu, G. (2016). "Tensile Properties of Hybrid Fiber-Reinforced Reactive Powder
Concrete After Exposure to Elevated Temperatures." International Journal of Concrete
Structures and Materials, vol. 10(1): pp. 29-37.
75. Li, Q., Yuan, G., Xu, Z., Dou, T. (2014). "Effect of elevated temperature on the mechanical
properties of high-volume GGBS concrete." Magazine of Concrete Research, vol. 66(24): pp.
1277-1285.
76. Li, Z., Venkata, H.K., Rangaraju, P.R. (2015). "Influence of Silica flour–silica Fume
Combination on the Properties of
High Performance Cementitious Mixtures at
Ambient Temperature Curing." Construction and Building Materials, vol. 100: pp. 225-233.
77. Lie, T.T., Kodur, V.K.R. (1996). "Thermal and mechanical properties of steel-fibre-reinforced
concrete at elevated temperatures." Canadian Journal of Civil Engineering, vol. 23(2): pp. 511517.
78. Lim, J., Moon, K., Falchetto, A.C., Jeong, J. (2016). "Testing and Modelling of HygroThermal Expansion Properties of Concrete." KSCE Journal of Civil Engineering, vol. 20(2):
pp. 709-717.
79. Liu, C.T., Huang, J.S. (2009). "Fire performance of highly flowable reactive powder concrete."
Construction and Building Materials, vol. 23.5: pp. 2072-2079.
80. Magureanu, C., Sosa, I., Negrutiu, C., Heghes, B. (2012). "Mechanical Properties and
Durability of Ultra-High-Performance Concrete." ACI Materials Journal, vol. 109(2): pp. 177.
81. Mindeguia, J., CarrĂŠ, H., Pimienta, P., La Borderie, C. (2015). "Experimental Discussion on
the Mechanisms Behind the Fire Spalling of Concrete." Fire and Materials, vol. 39(7): pp. 619635.
82. Naser, M. Z., and Kodur, V.K.R. (2015). “A probabilistic assessment for classification of
bridges against fire hazard.” Fire Safety Journal, vol. 76: pp. 65-73.
83. National Fire Data Center (U.S.) (2017). “Fire in the United States, 2005-2014.” Emmitsburg,
MD: United States Fire Administration, National Fire Data Center.
84. Nazri, F.M., Jaya, R.P., Abu Bakar, B.H., Ahmadi, R. (2017). "Fire resistance of ultra-high
performance fibre reinforced concrete due to heating and cooling." MATEC Web of
Conferences. 87. EDP Sciences.

138

85. Nilson, A.H., Darwin, D., Dolan, C.W. (2010). “Design of Concrete Structures.” McGrawHill, New York, New York.
86. Peng, G.F., Yang, J., Long, Q.Q., Nin, X.J., Zeng, Q.P. (2016). “Comparison between UltraHigh-Performance Concretes with Recycled Steel Fiber and Normal Industrial Steel Fiber.”
Fourth International Conference on Sustainable Construction Materials and Technologies,
August 7-11, 2016.
87. Peng, G., Kang, Y., Huang, Y., Liu, X. & Chen, Q. (2012). "Experimental Research on Fire
Resistance of Reactive Powder Concrete." Advances in Materials Science and Engineering,
vol. 2012: pp. 1-6.
88. Phan, L.T. (1996). “Fire performance of high-strength concrete: A report of the state-of-the
art.” US Department of Commerce, Technology Administration, National Institute of
Standards and Technology, Office of Applied Economics, Building
and Fire Research
Laboratory, 1996.
89. Phan, L.T. (2008). "Pore Pressure and Explosive Spalling in Concrete." Materials and
Structures, vol. 41(10): pp. 1623-1632.
90. Prem, P.R., Bharatkumar, B.H., Iyer, N.R. (2013). "Influence of Curing Regimes on
Compressive Strength of Ultra High Performance Concrete." Sadhana, vol. 38(6): pp. 14211431.
91. Prem, P.R., Bharatkumar, B.H., Iyer, N.R. (2012). “Mechanical properties of Ultra High
Performance Concrete.” International Journal of Civil, Environmental, Structural,
Construction and Architectural Engineering, vol. 6(8): pp. 676-685.
92. Rahman, S., Molyneaux, T., Patnaikuni, I. (2005). "Ultra High Performance Concrete: Recent
Applications and Research." Australian Journal of Civil Engineering, vol. 2(1): pp. 13.
93. Reda, M.M., Shrive, N.G., Gillott, J.E. (1999). "Microstructural investigation of innovative
UHPC." Cement and Concrete Research, vol. 29(3): pp. 323-329.
94. Ren, W., Xu, J., Su, H. (2016). "Dynamic compressive behaviour of concrete after exposure
to elevated temperatures." Materials and Structures, vol. 49(8): pp. 3321-3334.
95. Richard, P., Cheyrezy, M. (1995). "Composition of Reactive Powder Concretes." Cement and
Concrete Research, vol. 25(7): pp. 1501-1511.
96. RILEM Technical Committee 200-HTC. (2007). “Recommendation of RILEM TC 200-HTC:
mechanical concrete properties at high temperatures – modelling and applications: Part 1:
Introduction-General presentation.” Mater. Struct., vol. 40: pp. 841-853.
139

97. RILEM Technical Committee 200-HTC. (2007). “Recommendation of RILEM TC 200-HTC:
mechanical concrete properties at high temperatures—modelling and applications: Part 2:
Stress–strain relation.” Mater. Struct., vol. 40: pp. 855–864.
98. Russell, H.G., Graybeal, B.A. (2013). “Ultra-High Performance Concrete: A State-Of-The-Art
Report for the Bridge Community.” U.S. Department of Transportation, Federal Highway
Administration, Research, Development, and Technology, Turner-Fairbank Highway
Research Center, McLean, VA, 2013.
99. Sanchayan, S., Foster, S.J. (2016). "High Temperature Behaviour of Hybrid steel–PVA Fibre
Reinforced Reactive Powder Concrete." Materials and Structures, vol. 49(3): pp. 769-782.
100. Sarhat, S.R., Sherwood, E.G. (2013). "Residual Mechanical Response of Recycled Aggregate
Concrete after Exposure to Elevated Temperatures." Journal of Materials in Civil Engineering,
vol. 25(11): pp. 1721-1730.
101. Schrefler, B.A., Majorana, C.E., Khoury, G.A., Gawin, D. (2002). "Thermo-hydromechanical modelling of high performance concrete at high temperatures." Engineering
Computations, vol. 19.7: pp. 787-819.
102. So, H., Jang, H., Khulgadai, J., So, S. (2015). "Mechanical Properties and Microstructure of
Reactive Powder Concrete using Ternary Pozzolanics Materials at Elevated Temperature."
KSCE Journal of Civil Engineering, vol. 19(4): pp. 1050-1057.
103. So, H., Yi, J., Khulgadai, J., So, S. (2014). "Properties of Strength and Pore Structure of
Reactive Powder Concrete Exposed to High Temperature." ACI Materials Journal, vol. 111(3):
pp. 335.
104. Soliman, A.M., Nehdi, M.L. (2011). "Effect of Drying Conditions on Autogenous Shrinkage
in Ultra-High Performance Concrete at Early-Age." Materials and Structures, vol. 44(5): pp.
879-899.
105. Sultan, M.A., Kodur, V.K.R. (2003). "Effect of Temperature on Thermal Properties of HighStrength Concrete." Journal of Materials in Civil Engineering, vol. 15(2): pp. 101-107.
106. Tai, Y.S., Pan, H.H., Kung, Y.N. (2011). "Mechanical Properties of Steel Fiber Reinforced
Reactive Powder Concrete Following Exposure to High Temperature Reaching 800 °C."
Nuclear Engineering and Design, vol. 241(7): pp. 2416-2424.
107. Ulm, F.J. (2008). “What's the matter with concrete?” Atti del XX Convegno Nazionale del
Gruppo Italiano Frattura, vol. 3.

140

108. Voit, K., Kirnbauer, J. (2014). "Tensile Characteristics and Fracture Energy of Fiber
Reinforced and Non-Reinforced Ultra High Performance Concrete (UHPC)." International
Journal of Fracture, vol. 188(2): pp. 147-157.
109. Wang, D., Shi, C., Wu, Z., Xiao, J., Huang, Z., Fang, Z. (2015). "A review on ultra high
performance concrete: Part II. Hydration, microstructure and properties." Construction and
Building Materials, vol. 96: pp. 368-377.
110. Wang, J.J., Ren, F., DiPaolo, B.P. (2013). "Thermal Expansion Study and Microstructural
Characterization of High-Performance Concretes." Journal of Materials in Civil
Engineering, vol. 25.10: pp. 1574-8.
111. Wang, R., Gao, X. (2016). "Relationship between Flowability, Entrapped Air Content and
Strength of UHPC Mixtures Containing Different Dosage of Steel Fiber." Applied Sciences,
vol. 6(8): pp. 216.
112. Watanabe, K., Bangi, M.R., Horiguchi, T. (2013). "The effect of testing conditions (hot and
residual) on fracture toughness of fiber reinforced high-strength concrete subjected to high
temperatures." Cement and Concrete Research, vol. 51: pp. 6-13.
113. Way, R.T., Wille, K. (2015). "Effect of Heat-Induced Chemical Degradation on the Residual
Mechanical Properties of Ultrahigh-Performance Fiber-Reinforced Concrete." Journal of
Materials in Civil Engineering, vol. 28.4.
114. Wille, K., Naaman, A.E., Parra-Montesinos, G.J. (2011). "Ultra-High Performance Concrete
with Compressive Strength Exceeding 150 MPa (22 ksi): A Simpler Way." ACI Materials
Journal, vol. 108(1): pp. 46-54.
115. Xiao, R., Deng, Z., Shen, C. (2014). “Properties of Ultra High Performance Concrete
Containing Superfine Cement and without Silica Fume." Journal of Advanced Concrete
Technology, vol. 12(2): pp. 73-81.
116. Xiong, M.X., Liew, R.Y.J. (2015). “Spalling behavior and residual resistance of fibre
reinforced Ultra-High
performance concrete after exposure to high temperatures.” Mater.
Construction, vol. 65: pp. e071.
117. Yang, S.L., Millard, S.G., Soutsos, M.N., Barnett, S.J., Le, T.T. (2009). "Influence of
aggregate and curing regime on the mechanical properties of ultra-high performance fibre
reinforced concrete (UHPFRC)." Construction and Building Materials, vol. 23(6): pp. 22912298.

141

118. Yazıcı, H., Yardımcı, M.Y., Yiğiter, H., Aydın, S., Türkel, S. (2010). "Mechanical Properties
of Reactive Powder Concrete Containing High Volumes of Ground Granulated Blast Furnace
Slag." Cement and Concrete Composites, vol. 32(8): pp. 639-648.
119. Yazıcı, H., Yardımcı, M.Y., Aydın, S., Karabulut, A.Ş. (2009). “Mechanical Properties of
Reactive Powder Concrete Containing Mineral Admixtures Under Different Curing Regimes."
Construction and Building Materials, vol. 23(3): pp. 1223-1231.
120. Ye, H., Feng, N., Ling-hu, Y., Ran, Z., Lin, L., Qi, S., Dong, Y. (2012). "Research on Fire
Resistance of Ultra-High-Performance Concrete." Advances in Materials Science and
Engineering, vol. 2012: pp. 1-7.
121. Youssef, M.A., Moftah, M. (2007). "General stress–strain relationship for concrete at
elevated temperatures." Engineering Structures, vol. 29(10): pp. 2618-2634.
122. Yu, K., Yu, J., Lu, Z., Chen, Q. (2016) "Fracture properties of high-strength/highperformance concrete (HSC/HPC) exposed to high temperature." Materials and Structures, vol.
49(11): pp. 4517-4532.
123. Zdeb, T. (2013). “Ultra-high performance concrete – properties and technology.” Bull. Pol.
Acad. Sci. Tech. Sci., vol. 61: pp. 183–193.
124. Zdeb, T. (2017). "An Analysis of the Steam Curing and Autoclaving Process Parameters for
Reactive Powder Concretes." Construction and Building Materials, vol. 131: pp. 758-766.
125. Zheng, W., Luo, B., Lu, S. (2014a). "Rezistenta La Compresiune Si Încovoiere a Betoanelor
Cu Continut De Pulberi Reactive Si Fibre Hibride La Temperaturi ĂŽnalte/compressive and
Tensile Strengths of Reactive Powder Concrete with Hybrid Fibres at Elevated Temperatures."
Revista Romana de Materiale, vol. 44(1): pp. 36.
126. Zheng, W., Luo, B., Wang, Y. (2013). "Compressive and Tensile Properties of Reactive
Powder Concrete with Steel Fibres at Elevated Temperatures." Construction and Building
Materials, vol. 41: pp. 844-851.
127. Zheng, W., Luo, B., Wang, Y. (2014b). "Microstructure and Mechanical Properties of RPC
Containing PP Fibres at Elevated Temperatures." Magazine of Concrete Research, vol. 66(8):
pp. 397-408.
128. Zheng, W., Luo, B., Wang, Y. (2015). "Stress–strain Relationship of Steel-Fibre Reinforced
Reactive Powder Concrete at Elevated Temperatures." Materials and Structures, vol. 48(7):
pp. 2299-2314.

142

129. Zheng, W., Li, H., Wang, Y. (2012). "Compressive stress–strain Relationship of Steel FiberReinforced Reactive Powder Concrete After Exposure to Elevated Temperatures."
Construction and Building Materials, vol. 35: pp. 931-940.

143