International Journal of Concrete Structures and Materials

Vol.11, No.3, pp.447–457, September 2017

DOI 10.1007/s40069-017-0211-x
ISSN 1976-0485 / eISSN 2234-1315

Evaluation of Fire-Damaged Concrete: An Experimental Analysis

based on Destructive and Nondestructive Methods
Gang-Kyu Park1), and Hong Jae Yim2),*

(Received June 20, 2016, Accepted June 17, 2017, Published online September 12, 2017)

Abstract: Fire damage to concrete causes contact-type defects that degrade its durability through impaired mechanical properties.
Various nondestructive tests are used to evaluate defects induced by fire damage. Recently, nonlinear ultrasonic methods such as
the nonlinear resonance vibration method and nonlinear modulation method have been introduced. These nonlinear methods are
more sensitive to fire-induced contact-type defects than the linear ultrasonic method. This study involved an experimental analysis
of the residual material properties of fire-damaged concrete, specifically, compressive strength, splitting tensile strength, and static
elastic modulus. The residual material properties of 116 cylindrical concrete samples with various mix proportions and subjected to
various heating temperatures were measured by a destructive method, and their nonlinearity parameters were measured by two
nonlinear ultrasonic methods. Through regression analysis, correlated relationships that can facilitate the prediction of residual
material properties of fire-damaged concrete using measured nonlinearity parameters were identified. In addition, the effect of fire
damage on the mechanical strength of concrete was investigated by comparison with the relationships for undamaged concrete,
and relationships for the evaluation of fire-damaged concrete were identified through regression analysis.
Keywords: nonlinear ultrasonic method, fire-damaged concrete, mechanical property, correlation study.

1. Introduction including strength (Al-Nimry and Ghanem 2017; Dos Santos

et al. 2002; Li and Liu 2016). To evaluate the durability and
Although concrete is popular as a nonflammable material reusability of fire-damaged concrete, destructive and non-
with low thermal conductivity, thermophysical and thermo- destructive tests are performed in this study with various
chemical alterations induced in concrete at high temperatures proposed methods.
can degrade its performance (Bazant and Kaplan 1996). As An increase in contact-type defects within fire-damaged
the constituent materials of concrete have varying thermal concrete results in the degradation of mechanical properties
expansion coefficients, fire-damaged concrete has distributed including stiffness and strength. Among the nondestructive
contact-type defects between the individual materials. These methods, ultrasonic measurement methods demonstrate high
defects manifest as openings and pores within the concrete, potential and are applicable for evaluating damaged concrete
and the nondestructive assessment of these defects can be (Dilek 2007; Dilek and Leming 2007; Ham and Oh 2013).
used to evaluate the extent of fire damage (Yim et al. 2014). These methods can be divided into two categories depending
Researchers have conducted experimental studies to identify on the measurement techniques and target defects: linear and
the effect of fire damage on the material properties of con- nonlinear ultrasonic methods. Linear ultrasonic methods,
crete with various mix proportions under various fire sce- which measure wave velocity, wave attenuation, and impact
narios (Chang et al. 2006; Handoo et al. 2002; Lee et al. echo, are conventional ultrasonic methods and have been
2008; Tufail et al. 2016). It has been reported that high widely used to evaluate damaged concrete, particularly fire-
temperatures degrade the material properties of concrete, damaged concrete (Chaix et al. 2003; Colombo and Felicetti
2007; Dilek and Leming 2007; Epasto et al. 2010; Kee and
Nam 2015; Yang et al. 2009). However, these linear methods
1)
Department of Civil and Environmental Engineering, exhibit lower sensitivity to distributed defects and contact-
Korea Advanced Institute of Science and Technology, type defects at the micro-scale (Jhang 2009; Park et al. 2015)
291 Daehak-ro, Yuseong-gu, Daejeon 34141, compared to nonlinear ultrasonic methods, which are more
Republic of Korea.
sensitive to early-stage micro-scale defects as these defects
2)
Department of Construction and Disaster Prevention cause nonlinear behavior of an incident wave (Zaitsev et al.
Engineering, Kyungpook National University, 2559
2006). To investigate this phenomenon in concrete, Chen
Gyeongsang-daero, Sangju, Gyeongsangbuk-do 742-711,
et al. (2010) performed an experiment to characterize micro-
Republic of Korea.
scale defects induced by an alkali-silica reaction by using a
*Corresponding Author; E-mail: [email protected]
nonlinear impact resonance acoustic spectroscopy technique.
Copyright Ó The Author(s) 2017. This article is an open
In addition, Payan et al. (2007) conducted nonlinear resonant
access publication

447
ultrasound spectroscopy to assess the damage to concrete For measurements using the nonlinear modulation method
caused by exposure to high temperatures. Moreover, and compressive strength test, 16 cylindrical samples,
recently, defects in fire-damaged concrete with various mix including four reference specimens, were used for each mix.
proportions and under various temperature scenarios have On the other hand, for the nonlinear resonance vibration
been evaluated by various nonlinear ultrasonic methods method and splitting tensile strength test, 13 cylindrical
(Park et al. 2015; Yim et al. 2014). samples, including one reference specimen, were used for
Park et al. (2015) measured the hysteretic nonlinearity each mix. Moreover, each cylindrical sample was divided
parameter (HNP) of fire-damaged concrete with various mix into five disk samples that were 25 mm in height.
proportions and under various fire scenarios by using a Prior to exposure to high temperatures, the prepared
nonlinear resonance vibration method (Van Den Abeele et al. concrete samples were placed in a drying oven at 100 °C for
2000a). The splitting tensile strengths of the samples were approximately 24 h to avoid hygrothermal spalling, i.e.,
also evaluated, and the correlation between both the mea- explosive spalling, during the fire experiment. The concrete
surement results was reported. On the other hand, Yim et al. samples, apart from the reference samples, were then
(2014) measured the nonlinearity parameters of concrete exposed to various peak temperatures (200, 400, and
samples fabricated with mix proportions same as to those in 600 °C). For the homogenization of temperature in the entire
a previously mentioned study, using a nonlinear modulation volume of each sample, exposure to peak temperatures was
method (Van Den Abeele et al. 2000b). Moreover, mea- maintained for 2 h using an electric muffle furnace as it was
surements of residual material properties such as compres- reported that heat conduction up to the center of a sample is
sive strength, static elastic modulus, and peak strain were complete when it is subjected to 2 h of exposure (Yim et al.
also performed to obtain correlations with the nonlinear 2014). After exposure, all the samples were cooled in water
parameter. Notwithstanding the numerous correlation studies at 20 °C for 5 min to avoid unexpected recovery phenomena
of fire-damaged concrete that have been reported to date, during cooling, such as refilling of the fire-induced defects
several correlations remain to be analyzed. In this context, due to rehydration under a high-humidity condition, and
the present study involves an experimental analysis to were subsequently kept under an air-curing condition.
evaluate the residual material properties of fire-damaged The degree of thermal damage in fire-damaged concrete is
concrete based on nonlinearity parameters measured using influenced by several factors including exposure tempera-
two dissimilar nonlinear ultrasonic methods. To examine the ture, exposure time, after-fire curing periods, and sample
sensitivity of the two nonlinear ultrasound methods in sizes. Among these factors, it has been reported that the
evaluating fire-damaged concrete of various mix proportions after-fire curing period insignificantly affects the residual
at various temperatures, their results are compared, and material properties under an air-cured condition, while fire-
correlated relationships are proposed to assess residual damaged concrete under a completely saturated curing
mechanical strengths using measured nonlinearity parame- condition appears to recover its mechanical strength (Park
ters. In addition, the relationship between the compressive et al. 2015). Accordingly, for a correlation analysis, the
strength and tensile strength of fire-damaged concrete is concrete samples in this study were air-cured after heating to
proposed and compared with the corresponding relationships ignore the curing condition. Additionally, sample size
for undamaged concrete. influences the degree of thermal damage; however, it has
been reported that varying the sample size negligibly influ-
ences the measured nonlinearity parameters (Yim et al.
2. Sample Preparation: Fire-Damaged 2014). Therefore, this study attempts to correlate the mea-
Concrete sured HNP of thin disk samples with the measured nonlin-
earity parameter of the cylindrical samples.
Four types of concrete samples were prepared for
destructive and nondestructive tests after fire damage. Each
type of concrete sample had a unique mix proportion regu-
lated by varying the water-to-cement weight ratios (w/cm) 3. Nonlinear Ultrasonic Methods
and fine-to-coarse aggregate weight ratios for normal
strength concrete. Type I Portland cement was used to pro- 3.1 Nonlinear Resonance Vibration
duce all the samples with crushed gravel as coarse aggregate Measurement
(maximum size 19 mm) and fine aggregate (maximum size The nonlinear resonance vibration method is a nonlinear
4 mm). Additional admixtures or materials were not used in ultrasonic technique based on the measurement of the res-
any of the concrete samples. According to the various mix onance frequency shift. This implies that it measures the
proportions, the samples were labeled from C1 to C4, as frequency variation between the input and output signals,
presented in Table 1. The concrete samples were cast into a which is a function of the degree of damage. This can be
cylindrical shape with a height of 200 mm and diameter of described by the phenomenological model for hysteretic
100 mm. Totally, 116 cylindrical concrete samples (29 nonlinearity proposed by Van den Abeele et al. (2000a). The
samples for each mix) were then cast and cured for 28 days constitutive relationship for the elastic modulus K can be
prior to high temperature exposure. expressed as follows (Van Den Abeele et al. 2000a):

448 | International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017)
 Table 1 Mix proportions of concrete samples.
3
Mix Water (kg/m ) Cement (kg/m3) w/cm Fine aggregate Coarse aggregate Fine-to-coarse
ratio (kg/m3) (kg/m3) aggregate ratio

C1 160 320 0.5 744 1100 0.68

C2 171 285 0.6 744 1100 0.68
C3 160 320 0.5 922 922 1
C4 171 285 0.6 922 922 1

Kðe; e_ Þ ¼ K0 f1  a½De þ eðtÞsignð_eÞ þ   g ð1Þ National Instruments Corp.), the analog response signal was
converted to a digital signal with a sampling rate of 100 kHz
where a is the HNP, e_ is the strain rate, De is the strain and duration of 50 ms. The vibration signal was then con-
amplitude variation over the previous period, and signð_eÞ ¼ verted to the frequency domain via fast Fourier transform
1 if e_ [ 0 or signð_eÞ ¼ 1 if e_ \0. The hysteretic (FFT); a representative vibration signal and the corre-
nonlinearity of cement-based materials can be measured by sponding FFT result are presented in Fig. 2. The linear
the hysteretic nonlinear characteristics, namely nonlinear resonance frequency (f0 ) is determined by linear regression
attenuation, harmonic generation, and amplitude-dependent analysis, as the x-axis intercept of the FFT result. This is
resonance frequency shift, which are sensitive to contact- because the linear resonance frequency, which is dependent
type defects. Therefore, these characteristics have been used on the amplitude, is challenging to obtain by an impact test.
as damage indicators (Chen et al. 2011; Leśnicki et al. 2011; The HNP is then determined by measuring the amount of
Van Den Abeele et al. 2000a). Park et al. (2015) measured resonance frequency shift according to the intensity of
the HNP to evaluate the degree of fire damage by using the impact based on the linear resonance frequency; further
amplitude-dependent resonance frequency shift. The details regarding the experimental procedure are available in
resonance frequency shift with variation in the input the study by Park et al. (2015).
amplitude can be expressed as follows (Van Den Abeele
et al. 2000a): 3.2 Nonlinear Modulation Measurement
The constitutive law of concrete materials follows a non-
f0  f linear behavior and is expressed up to second-order non-
¼ aDe ð2Þ
f0 linearity in a straightforward manner as follows (Van Den
Abeele 1996; Van Den Abeele et al. 2000b):
where f0 is the linear resonance frequency, and f is the
measured resonance frequency according to the magnitude r ¼ E0 eð1 þ beÞ ð3Þ
of the input amplitude. Moreover, HNP (a: obtained from the
amplitude-dependent resonance frequency shift) can be used where r is the stress, E0 is the Young’s modulus, e is the
as the damage factor. The resonance frequency varies lin- strain, and b is the second-order nonlinear coefficient.
early with increasing input amplitude, and the degree of shift Nonlinear ultrasonic phenomena such as higher harmonics
increases with damage. Therefore, the extent of fire damage mode or mixed frequency response are represented by wave
can be evaluated through an analysis of the amplitude-de- propagation through a solid medium (Payan et al. 2010; Van
pendent resonance frequency shift; further details are avail- Den Abeele et al. 2000b). The mixed frequency response can
able in the study by Van Den Abeele et al. (2000a). be obtained using two types of generating waves with
For an experiment based on this relationship, thin, disk- unequal frequencies and nonlinear wave modulation spec-
shaped concrete samples (25 mm in height) were prepared troscopy. A low-frequency vibration (fl ) and a high-fre-
and placed on a soft mat, which provides a soft boundary quency stress wave (fh ) are generated by the shaker and
condition that maximizes the capability of the disks for free ultrasonic transducer, respectively. When simultaneously
vibration and minimizes noise from the external experi- applying both low and high frequencies to a sample, the low-
mental conditions. The experimental setup is illustrated in frequency wave causes modulation of the high-frequency
Fig. 1. To generate impact excitation in the sample, a steel wave at the contact-type defects in the sample. This results
bead (diameter 15 mm, 13.8 g) was dropped 20 times from in an additional spectrum of the modulated wave at the
various heights onto the center of the sample; the intensity of sideband range in the frequency domain. The sideband is
impact of the steel bead is to be unique for each attempt to generally located at the sum and difference of both fre-
obtain the amplitude-dependent resonance frequency shift. quencies (fl  fh ), and the spectral component of the mod-
The vibration response induced by an impact was measured ulated frequency can represent the nonlinear behavior
by a shear piezoelectric accelerometer (PCB 353B15, PCB induced by contact-type defects.
Piezotronics Inc.) attached on the opposite side of the sam- The amplitude of the sideband is proportional to the
ple. Through an analog-to-digital converter (NI PXI 4472-B, amplitude of the low- and high-frequency generated signals

International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017) | 449
Fig. 1 Experimental setup of nonlinear resonance vibration method. a Schematic of experimental system and b experimental setup.

Fig. 2 Representative results of vibration signal and the amplitude-dependent resonance frequency shift of damaged concrete.
a Vibration signal and b FFT.

and the area of contact-type defects in the sample (Donskoy Ps

¼ DPl ð4Þ
et al. 2001). This implies that the nonlinearity parameter (D) Ph
can be evaluated by the energy relationship of both high (Ph )
and low frequencies (Pl ) and the sideband component (Ps ). Yim et al. (2014) proposed an impact-modulation method
Accordingly, the nonlinearity parameter that reflects the to identify the degree of thermal damage to concrete. The
degree of contact-type defects in the sample can be obtained experimental setup of this method is illustrated in Fig. 3. A
by the following expression (Van Den Abeele et al. 2000b; low-frequency vibration (f0), instead of a low-frequency
Warnemuende and Wu 2004): ultrasonic longitudinal wave, was generated using an impact

Fig. 3 Experimental setup of nonlinear modulation method. a Schematic of experimental system and b experimental setup.

450 | International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017)
hammer with a soft tip (086C03; PCB Piezotronics, Inc.) to concrete. A detailed description of the phenomena related to
create a resonance vibration mode in the sample (Donskoy this experiment is available in the study by Yim et al. (2014).
et al. 2001; Warnemuende and Wu 2004). High-frequency
sinusoidal signals were generated using a function generator
(NI PXI-5421; National Instruments Corp.) with a sample 4. Mechanical Strength Measurements
rate of 100 MS/s, and a power amplifier (BA4825; NF
Corp.) was used to amplify the generated signal. Two lon- After obtaining the fire damage and nondestructive mea-
gitudinal narrow-band transducers (Panametrics X1019; surements of the fabricated concrete samples, two types of
Olympus NDT, Inc.) with a center frequency of 46.1 kHz mechanical strength measurements were performed. The
were used for transmitting and receiving the high-frequency thin-disk samples were used to obtain the splitting tensile
ultrasonic wave passing through the sample, and a tri-axial strength according to the degree of fire damage. Based on
accelerometer (356A33; PCB Piezotronics, Inc.) placed the measured dimensions and mass of the samples, the
opposite to the impact region measured the impact vibration splitting tensile strength test was performed following the
(low frequency) in the three orthogonal directions. The procedure of ASTM C 496 (2011). As illustrated in Fig. 5,
measured high-frequency signal was digitized (NI PXI- the concrete sample was placed in the direction of the
5105; National Instruments Corp.) with a 60 MS/s sampling diameter at the center of the bearing plates, and two wood
rate (12 bit resolution), and the dynamic vibration was straps were placed between the bearing plates and the sample
measured (NI PXI-4472B; National Instruments Corp.) with to avoid interface cracking and to apply a distributed load
a sampling rate of 102.4 kS/s (24 bit resolution). along the length of the disks. The splitting tensile strength
The spectral energy of the impact vibration (El ) was can then be obtained by the following expression:
determined by the integrated power spectral density of the
2F
measured signal of frequency up to 5 kHz. The spectral Ts ¼ ð5Þ; Þ
energy of the sideband was obtained by integrating the pld
power spectral density of the ultrasonic signal of frequency
where Ts is the splitting tensile strength, F is the maximum
between 41.1 and 51.5 kHz, excluding the range of the high-
applied load until failure of each sample, and l and d are the
frequency waves (46.1 kHz). The nonlinearity parameter
length and diameter of the sample, respectively. Park et al.
was then obtained from the slope of the relationship between
(2015) reported that the measured tensile strength decreased
the modulated energy (Es =Eh ) and the impact vibration (El ).
with increasing fire damage to the concrete sample. Fol-
Thirty measurements per test for each sample were con-
lowing ASTM C39 (2001), the compressive strength was
ducted at an identical position to obtain a linear curve, and
also obtained using cylindrical concrete samples subjected to
ten independent tests were performed for reproducibility at
varying degrees of fire damage, and the axial deformation
other arbitrary locations of impact and accelerometer
and uniaxial stress–strain curve were measured using dis-
placement. The representative results of frequency modula-
placement transducers attached on the sides of the cylinder.
tion achieved by simultaneously applying the low and high
From this experiment, Yim et al. (2014) measured the
frequency waves are presented in Fig. 4. The nonlinearity
residual mechanical properties of fire-damaged concrete,
parameter can then be obtained from the average of the
such as compressive strength, static elastic modulus (calcu-
values from the 10 tests. It has been reported that the ratio of
lated as the secant elastic modulus), and peak strain, with
increase of the nonlinearity parameter represents contact-
respect to the degree of thermal damage. The reported results
type defects developed because of thermal damage to the
indicated that increasing the fire damage to concrete

Fig. 4 The modulated high frequency wave of samples. a Modulation signal in time domain and b frequency spectra.

International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017) | 451
 Fig. 5 Schematic diagram of splitting tensile strength test. a Schematic diagram and b experimental setup (Park et al. 2015).

degrades the mechanical properties and induces a linear and nonlinearity parameter); the initial and post-fire-damage
stress–strain curve. material properties are presented in Table 2. Figure 6 pre-
sents the increased ratios of the nonlinearity parameters as
measured by the two methods with various mix proportions
5. Results and Discussion and under various heating temperatures; the results were
calculated as the ratio of the increase to the reference result
Based on previous studies (Park et al. 2015; Yim et al. (20 °C). To ascertain their experimental adequacy, the sen-
2014), this study performed an experimental analysis using sitivity of the two nonlinearity parameters, as measured by
selected parameters of mechanical strength (compressive and the nonlinear resonance vibration method and nonlinear
tensile strength) and measured nonlinearity parameters (HNP modulation method, were analyzed by comparing the results.

Table 2 The average results of destructive and non-destructive tests.

Mix Temp. Compressive Splitting tensile Secant elastic HNP Nonlinearity
strength (MPa) strength (MPa) modulus parameter (D)
C1 Initial (20 °C) 52.11 4.33 33.25 1.82 (10-6) 0.046
200 °C 37.17 3.96 22.89 5.98 (10-6) 0.225
-5
400 °C 30.37 2.59 8.169 1.12 (10 ) 4.645
-5
600 °C 23.96 1.00 2.911 2.84 (10 ) 21.25
C2 Initial (20 °C) 39.03 3.56 28.61 9.02 (10-7) 0.022
-6
200 °C 27.08 3.34 18.74 4.73 (10 ) 0.240
-5
400 °C 25.67 2.15 6.651 1.54 (10 ) 3.093
600 °C 18.29 0.84 2.154 2.97 (10-5) 20.536
-7
C3 Initial (20 °C) 46.71 4.11 30.82 7.68 (10 ) 0.030
-6
200 °C 34.78 3.62 18.88 4.30 (10 ) 0.224
400 °C 27.68 2.46 7.190 1.15 (10-5) 2.330
600 °C 21.81 1.07 3.002 2.04 (10-5) 14.26
-7
C4 Initial (20 °C) 38.70 3.78 27.32 6.82 (10 ) 0.021
-6
200 °C 28.42 2.93 16.33 3.19 (10 ) 0.153
400 °C 24.35 2.10 5.964 9.03 (10-6) 1.513
-5
600 °C 17.95 0.70 1.976 2.1 (10 ) 16.70

452 | International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017)
Fig. 6 Comparison of ratio of increase of the nonlinearity parameters at various temperatures. a Nonlinear resonance vibration
method and b nonlinear modulation method.

The ratios of increase measured by both the methods are significant reduction in compressive strength at a low heat-
similar at approximately 200 °C for all the mix proportions; ing temperature indicate that these reactions in concrete have
however, from 200 to 600 °C, the ratio of increase measured a marginally higher effect on the compressive strength of the
by the modulation method is substantially higher than that concrete material than on is splitting tensile strength. On the
measured by the resonance method. These trends are pre- other hand, above 300 °C, the cement paste undergoes
sented by comparing the results for all the mix proportions. morphological alterations, and the skeleton does not appear
The ratio of increase as measured by the nonlinear modu- to be a continuous medium owing to its chemical dehydra-
lation method increases noticeably with temperature com- tion caused by the continuously increasing temperature (Yim
pared to that for the nonlinear resonance vibration method. et al. 2012). These phenomena are likely to exert a higher
Based on the comparison, it can be concluded that although influence on the reduction of splitting tensile strength than
the results of the nonlinear modulation method are more the reduction of compressive strength at a high heating
scattered than those of the nonlinear resonance vibration temperature. Accordingly, the various ratios of decrease of
method, the nonlinear modulation method can better reflect mechanical strength can be used as a guideline to evaluate
fire damage (contact-type defects) at high heating tempera- the heating temperature during a fire.
tures than the nonlinear resonance vibration method. The experimental results of previous studies have revealed
In addition, a comparative analysis of compressive the relationship between the residual material properties and
strength and splitting tensile strength was performed at the parameters measured by a linear ultrasonic method (Yang
various temperatures, as illustrated in Fig. 7. The trends of et al. 2009) and that between the residual material properties
the ratios of decrease of the compressive strength and and the nonlinearity parameters obtained from nonlinear
splitting tensile strength according to temperature appear to ultrasonic methods (Park et al. 2015; Yim et al. 2014). The
be highly similar regardless of mix proportions. The ratio of nonlinearity parameters measured by both the nonlinear
decrease of compressive strength reduces to approximately methods were also correlated with the measured mechanical
30% at 200 °C, which exhibits a noticeable variation from strength of the fire-damaged concrete: the nonlinearity
that of splitting tensile strength, except for the C4 mix parameter versus compressive strength; and the HNP versus
proportion; however, the ratio of decrease of both the tensile strength. As a departure from previous studies, this
strength measurements are similar at 400 °C (approximately study presents new correlated relationships between the
60% of undamaged strength). On the other hand, the ratio of residual material properties and the measured nonlinearity
decrease of the splitting tensile strength reduces to approx- parameters; this area is being addressed for the first time.
imately 20% of the undamaged strength, at 600 °C; the Figure 8 presents the correlation between the ratio of
degree of decrease is substantial when compared to the decrease of the residual material properties, specifically
decrease in compressive strength (approximately 50%). This compressive strength, splitting tensile strength, and static
implies that the splitting tensile strength marginally elastic modulus, and the ratio of increase of the nonlinearity
decreased compared to the compressive strength until parameters as measured by each of the nonlinear ultrasonic
200 °C and then reduced remarkably at 600 °C. While the methods. These correlations are expressed after excluding
compressive strength significantly decreased at a low tem- the effect of mix proportions and exposure times. The
perature (200 °C), this decrease weakened with increasing dashed lines in Fig. 8 represent the relationships between the
temperature. It is determined that reactions in concrete, such two variables (nonlinearity parameter and mechanical
as water evaporation in concrete and dehydration of the property) determined by regression analysis; the results
cement gel, are initiated up to 200 °C (Bazant and Kaplan exhibit a consistent tendency notwithstanding the omission
1996). The experimental results that exhibited a relatively of exposure time and mix proportion. The obtained

International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017) | 453
 Fig. 7 Comparison of ratios of decrease of strength with temperature. a C1, b C2, c C3 and d C4.

regression results are expressed as Eqs. (6)–(8), which cor- the fire-damage to the concrete when subjected to a tem-
respond to Fig. 8a–c, respectively: perature of approximately 600 °C. This implies that the
 0:183 variations in the nonlinearity parameter ratios increase with
fcn0 an temperature to a significantly higher degree than the varia-
¼ 0:93 ð6Þ
fc00 a0 tions in the strength ratios and static elastic modulus ratio. In
 0:239 addition, it can be observed that the compressive strength
ftn Dn ratio, splitting tensile strength ratio, and static elastic mod-
¼ 1:34 ð7Þ
ft0 D0 ulus ratio converged to approximately 0.5, 0.25, and 0.1,
 0:75 respectively. This implies that among the three parameters,
Kcn an the static elastic modulus is most sensitive to elevation in
¼ 1:35 ð8Þ
Kc0 a0 temperature, followed by the splitting tensile strength and
compressive strength.
where the subscript n corresponds to each experiment under It may also be feasible to correlate the measured nonlin-
various thermal conditions, and the subscript 0 corresponds earity parameters with the mechanical properties regardless
to the experiment under the initial condition without dam- of the mix proportions used in this study. Therefore, the
0
age. Moreover, fc , ft , a, and D are the compressive strength, residual material properties of concrete induced by fire
splitting tensile strength, hysteretic nonlinearity parameter, damage can be evaluated using these criteria under the
and nonlinearity parameter, respectively. assumption of the used mix proportions. For example, the
These relationships for all the thermal damage cases are residual compressive strength of fire-damaged concrete can
adequately described by a negative power function. In the be estimated by measuring the HNP via the nonlinear res-
initial phase, the ratios of the mechanical properties exhib- onance vibration method using thin concrete disks.
ited a remarkable decrease until the nonlinearity parameter An additional correlation study based on the ratios of
ratios increased to 10 for the HNP (a) and 200 for the decrease of the compressive strength and splitting tensile
nonlinearity parameter (D) (approximately before 400 °C); strength (measured mechanical strengths) of the fire-dam-
this is illustrated in Fig. 8. In the following phase, relatively aged concrete was performed, as illustrated in Fig. 9a. The
widespread values of the nonlinearity parameters are ratio of compressive strength is proportional to that of the
apparent with marginal variations in the mechanical prop- splitting tensile strength, and the optimized regression result
erties. Generally, this widespread section can be attributed to is expressed as follows:

454 | International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017)
Fig. 8 Correlation between ratio of decrease of the residual material properties and ratio of increase of the nonlinearity
parameters.

 
fcn0 ftn various investigators. Oluokun (1991) proposed the rela-
0 ¼ 0:47 þ 0:3532: ð9Þ tionship between compressive strength and tensile strength
fc0 ft0
solid line in Fig. 9b); however, it was obtained through
This study also investigated the effect of fire damage on regression analysis of measurements of undamaged concrete
the relationship between the residual compressive strength samples, and the relationship for fire-damaged concrete
and splitting tensile strength of concrete (dashed line in samples has not been reported yet. In the case of concrete
Fig. 9b) through a comparison with the relationship collated with low fire damage, the relationship follows the trend of
and provided by Oluokun (1991), which has been tested by the result for undamaged concrete proposed by Oluokun

International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017) | 455
 Fig. 9 Relationship between compressive strength and splitting tensile strength.

(1991). : On the other hand, as fire damage in the concrete relationships, and relationships determined through regres-
sample increase, the fire-damaged relationship represented a sion analysis were proposed for fire-damaged concrete.
steeper decline compared to undamaged relationship, which
yields the following expression:
Acknowledgement
ftn ¼ 3:78 lnðfcn0 Þ  10:07: ð10Þ
This work was supported by the National Research Foun-
The regressed relationships illustrate that fire damage
dation of Korea (NRF) grant funded by the Korea govern-
degrades compressive strength more than it degrades tensile
ment (MSIP) (No. 2015R1C1A1A01055474).
strength. Furthermore, based on Eqs. (9) and (10), the
residual compressive strength of fire-damaged concrete can
be estimated from the measured value of splitting tensile
Open Access
strength, and vice versa; thus, it is advantageous to select the
more efficient method based on the circumstances.
This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits un
6. Conclusion
restricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s)
This study involved an experimental analysis of the non-
and the source, provide a link to the Creative Commons
linearity parameters measured by two nonlinear ultrasonic
license, and indicate if changes were made.
methods. For the evaluation of fire-damaged concrete, the
nonlinearity parameter measured by the nonlinear modula-
tion method provided higher sensitivity than that measured
by the nonlinear resonance vibration method. The measured
References
nonlinearity parameter can be affected by experimental
conditions such as the sample size, the impact type and
Al-Nimry, H. S., & Ghanem, A. M. (2017). FRP confinement of
location of impact that is generated, wave reflection, and
heat-damaged circular RC columns. Int. J. Concr. Struct.
boundary condition of the sample. Accordingly, the experi-
Mater., 11, 115–133.
mental results of this study, rather than representing general
ASTM. (2001). ASTM C39: Standard test method for com-
relationships, can provide a guideline on the relationship
pressive strength of cylindrical concrete specimens. West
between the mechanical properties of fire-damaged concrete
Conshohocken, PA, USA: ASTM Int.
and nonlinearity parameters measured by the two nonde-
ASTM. (2011). ASTM C 496 M-11: Standard test method for
structive methods. Based on the results of these experimental
splitting tensile strength of cylindrical concrete specimens.
methods, correlated relationships were proposed to evaluate
West Conshohocken, PA, USA: ASTM Int.
the residual properties of fire-damaged concrete using the
Bazant, Z. P., & Kaplan, M. F. (1996). Concrete at high tem-
measured nonlinearity parameter. From these relationships,
peratures: material properties and mathematical models.
the residual compressive strength, tensile strength, and static
Harlow: Longman.
elastic modulus after being exposed to fire can be estimated
Chaix, J. F., Garnier, V., & Corneloup, G. (2003). Concrete
using the measured nonlinearity parameters without con-
damage evolution analysis by backscattered ultrasonic
sidering the mix proportions of the concrete. In addition, the
waves. NDT E Int., 36(7), 461–469.
effect of fire damage on the mechanical strength of concrete
Chang, Y. F., Chen, Y. H., Sheu, M. S., & Yao, G. C. (2006).
was investigated by a comparison of undamaged
Residual stress-strain relationship for concrete after

456 | International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017)
 exposure to high temperatures. Cement and Concrete elevated temperatures. International Journal of Concrete
Research, 36(10), 1999–2005. Structures and Materials, 10(1), 29–37.
Chen, J., Jayapalan, A. R., Kim, J. Y., Kurtis, K. E., & Jacobs, Oluokun, F. A. (1991). Prediction of concrete tensile strength
L. J. (2010). Rapid evaluation of alkali-silica reactivity of from its compressive strength. Evaluation of existing rela-
aggregates using a nonlinear resonance spectroscopy tech- tions for normal weight concrete. ACI Materials Journal,
nique. Cement and Concrete Research, 40(6), 914–923. 88(3), 302–309.
Chen, J., Kim, J.-Y., Kurtis, K. E., & Jacobs, L. J. (2011). Park, S.-J., Park, G.-K., Yim, H. J., & Kwak, H.-G. (2015).
Theoretical and experimental study of the nonlinear reso- Evaluation of residual tensile strength of fire-damaged
nance vibration of cementitious materials with an applica- concrete using a non-linear resonance vibration method.
tion to damage characterization. Journal of the Acoustical Magazine of Concrete Research, 67(5), 235–246.
Society of America, 130(5), 2728–2737. Payan, C., Garnier, V., Moysan, J., & Johnson, P. A. (2007).
Colombo, M., & Felicetti, R. (2007). New NDT techniques for Applying nonlinear resonant ultrasound spectroscopy to
the assessment of fire—damaged concrete structures. Fire improving thermal damage assessment in concrete. The
Safety Journal, 42(6), 461–472. Journal of the Acoustical Society of America, 121(4),
Dilek, U. (2007). Ultrasonic pulse velocity in nondestructive EL125.
evaluation of low quality and damaged concrete and Payan, C., Garnier, V., & Moysan, J. (2010). Potential of non-
masonry construction. J. Perform. Constr. Facil., 21(5), linear ultrasonic indicators for nondestructive testing of
337–344. concrete. Advances in Civil Engineering, 2010, 1–8.
Dilek, U., & Leming, M. L. (2007). Comparison of pulse Tufail, M., Shahzada, K., Gencturk, B., & Wei, J. (2016). Effect
velocity and impact-echo findings to properties of thin of elevated temperature on mechanical properties of lime-
disks from a fire damaged slab. J. Perform. Constr. Facil., stone, quartzite and granite concrete. International Journal
21(1), 13–21. of Concrete Structures and Materials., 11, 1–12.
Donskoy, D., Sutin, A., & Ekimov, A. (2001). Nonlinear Van Den Abeele, K. E. A. (1996). Elastic pulsed wave propa-
acoustic interaction on contact interfaces and its use for gation in media with second- or higher-order nonlinearity.
nondestructive testing. NDT E Int., 34(4), 231–238. Part I. Theoretical framework. Journal of the Acoustical
Dos Santos, J. R., Branco, F. A., & De Brito, J. (2002). Society of America, 99(6), 3334–3345.
Assessment of concrete structures subjected to fire–The Van Den Abeele, K. E. A., Carmeliet, J., Ten Cate, J. A., &
FBTest. Mag. Concr. Res., 54(3), 203–208. Johnson, P. A. (2000a). Nonlinear elastic wave spec-
Epasto, G., Proverbio, E., & Venturi, V. (2010). Evaluation of troscopy (NEWS) techniques to discern material damage,
fire-damaged concrete using impact-echo method. Materi- part I: Nonlinear wave modulation spectroscopy (NWMS).
als and Structures, 43(1–2), 235–245. Research in Nondestructive Evaluation, 12(1), 17–30.
Ham, S., & Oh, T. (2013). Effect of mixing and placing in hot Van Den Abeele, K. E. A., Johnson, P. A., & Sutin, A. (2000b).
weather on hardened concrete properties. Int. J. Concr. Nonlinear elastic wave spectroscopy (NEWS) techniques to
Struct. Mater., 7(2), 165–174. discern material damage, part II: Single-Mode nonlinear
Handoo, S. K. K., Agarwal, S. K. K., & Agarwal, S. K. K. resonance acoustic spectroscopy. Research in nondestruc-
(2002). Physicochemical, mineralogical, and morphologi- tive evaluation, 12(1), 31–42.
cal characteristics of concrete exposed to elevated temper- Warnemuende, K., & Wu, H. C. (2004). Actively modulated
atures. Cement and Concrete Research, 32(7), 1009–1018. acoustic nondestructive evaluation of concrete. Cement and
Jhang, K. Y. (2009). Nonlinear ultrasonic techniques for non- Concrete Research, 34(4), 563–570.
destructive assessment of micro damage in material: A Yang, H., Lin, Y., Hsiao, C., & Liu, J. (2009). Evaluating
Review. Int. J. Precis. Eng. Manuf., 10(1), 123–135. residual compressive strength of concrete at elevated tem-
Kee, S. H., & Nam, B. (2015). Automated surface wave mea- peratures using ultrasonic pulse velocity. Fire Safety
surements for evaluating the depth of surface-breaking Journal, 44(1), 121–130.
cracks in concrete. Int. J. Concr. Struct. Mater., 9(3), Yim, H. J., Kim, J. H., Park, S.-J., & Kwak, H.-G. (2012).
307–321. Characterization of thermally damaged concrete using a
Lee, J., Xi, Y., & Willam, K. (2008). Properties of concrete after nonlinear ultrasonic method. Cement and Concrete
high-temperatures heating and cooling. ACI Materials Research, 42(11), 1438–1446.
Journal, 105(4), 334–341. Yim, H. J., Park, S.-J., Kim, J. H., & Kwak, H.-G. (2014).
Leśnicki, K. J., Kim, J.-Y., Kurtis, K. E., & Jacobs, L. J. (2011). Nonlinear ultrasonic method to evaluate residual mechan-
Characterization of ASR damage in concrete using non- ical properties of thermally damaged concrete. ACI Mate-
linear impact resonance acoustic spectroscopy technique. rials Journal, 111(4), 399–409.
NDT E International, 44(8), 721–727. Zaitsev, V., Nazarov, V., Gusev, V., & Castagnede, B. (2006).
Li, H., & Liu, G. (2016). Tensile properties of hybrid fiber- Novel nonlinear-modulation acoustic technique for crack
reinforced reactive powder concrete after exposure to detection. NDT E Int., 39(3), 184–194.

International Journal of Concrete Structures and Materials (Vol.11, No.3, September 2017) | 457