You are on page 1of 16

Engineering Fracture Mechanics 202 (2018) 87–102

Contents lists available at ScienceDirect

Engineering Fracture Mechanics


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

Determining tensile strength of concrete based on experimental


T
loads in fracture test

Longbang Qing , Xinyu Shi, Ru Mu, Yuehua Cheng
College of Civil Engineering and Transportation, Hebei University of Technology, Tianjin 300401, China

A R T IC LE I N F O ABS TRA CT

Keywords: A theoretical method for determining the tensile strength ft of concrete was proposed based on
Concrete the experimental loads in fracture tests. Using the fracture extreme method, only the initial
Tensile strength cracking load Pini and the peak load Pmax were required to be measured. The experimental data
Initial cracking load from three-point bending notched beams and wedge splitting specimens were used to calculate ft
Peak load
with different initial crack length versus specimen depth ratios α0 and specimen dimensions. The
Fracture extreme method
results show that the ft obtained by the proposed method agree with the test values, and the
method is insensitive to α0 and specimen dimensions.

1. Introduction

Concrete is a multiphase composite composed of aggregates and mortar. Some microcracks are generated between aggregates and
mortar, sand and cement paste, or unhydrated cement particles during the forming stage of concrete. The formation of cracks cause
weak performance of concrete in tension and may result in structural safety issues. In addition, temperature, age, and curing con-
dition can also affect the tensile property of concrete. The tensile strength ft of concrete is a basic mechanical parameter and an
important inherent material coefficient in structural designing. It is also an important parameter to describe the softening relation
which has been widely used in simulation of crack propagation in concrete.
The uniaxial tensile test is generally considered as the most direct method for determining tensile strength ft [1], which is carried
out not under any theoretical hypothesis and is more reliable than other methods [2]. However, many tests conducted in the past
failed because of unexpected crushing which occurred as a result of local stress concentration and eccentricity [3,4]. The accuracy of
the tests might be affected by the loading facilities, the shapes and sizes of specimens, etc.
Splitting tests [5], flexural tests [6] and fracture tests are also typically used to determine the ft of concrete as indirect approaches.
However, the values measured by different test methods might be different using various formulas and predetermined assumptions
[7] and test facilities, such as the heel block implemented for the cube-splitting test [8]. Different empirical formulas have been
developed to describe the relations among the direct tensile strength ft, the splitting tensile strength fts, the flexural tensile strength fr
and the compressive strength fc [9–14]. Compared with the uniaxial tensile test, these indirect tension test methods are widely used
for studying the properties of concrete due to their simplicity. For instance, the three-point bending (TPB) [15–17], the wedge
splitting (WS) [18], the compact tension [19], the split-tension [20], the Brazilian disc [21], the semi-circular bend [22] and many
other types of specimen along with the methods in [23–25] have been used to study the mechanics characteristics. Among those
methods, TPB test is adopted by RILEM [26,27] as a recommended method for determining fracture parameters.
Many numerical methods have been developed to characterise the tensile strength ft based on the fracture tests. Zhang et al. [28]


Corresponding author.
E-mail address: qing@hebut.edu.cn (L. Qing).

https://doi.org/10.1016/j.engfracmech.2018.09.017
Received 8 April 2018; Received in revised form 12 August 2018; Accepted 10 September 2018
Available online 11 September 2018
0013-7944/ © 2018 Elsevier Ltd. All rights reserved.
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

Nomenclature H half depth of wedge splitting specimen


KICini initial fracture toughness
ɑ effective crack length KICun unstable fracture toughness
ɑ0 initial crack length M1, M2, M3 parameters of weight function
ɑc critical effective crack length at peak load P external load
B specimen width Pini initial cracking load
c1, c2 material constants for nonlinear softening function Pmax peak load
CMOD crack mouth opening displacement S specimen span of three-point bending beam
CTOD crack tip opening displacement W specimen weight
CTODc critical crack tip opening displacement α crack length versus specimen depth ratio
D specimen depth α0 initial crack length versus specimen depth ratio
E Young’s modulus αc critical crack length versus specimen depth ratio
fc compressive strength of concrete ω0 critical value of crack opening displacement at
fr flexural tensile strength of concrete which the σ = 0
ft tensile strength of concrete σ cohesive stress
fts splitting tensile strength of concrete

proposed an inverse analysis approach to determine the ft, stress–crack opening relationship, cracking and fracture energy with
multiple specimens and integral calculations. Slowik et al. [29] proposed a new optimisation method for the fitting process that used
repeated numerical simulations to approximate experimental results. And the method requires updating the assumed softening curve
and re-analysing in several iterations. Tin-Loi and Que [30] investigated a number of numerical algorithms to characterize the
parameters based on solving a series of standard nonlinear programming problems.
Recently, Qing and Li [31] proposed an extreme method using the experimental peak load in the fracture test to calculate the
fracture parameters. The fracture extreme method regards that when the external load reaches the peak load, the partial derivative of
the external load to the effective crack length is zero. Subsequently, combining with the weight function method [32], a simplified
extreme method was proposed for TPB concrete beams [33] and central notched split-tension specimens [34]. This method requires
only one specimen and avoids measuring the values of the critical crack mouth opening displacement CMODc.
In this study, a theoretical method was proposed based on the fracture extreme method, that only uses peak load Pmax and initial
cracking load Pini from fracture tests to determine the ft of concrete. The method was verified by comparing the calculated values and
the existing experimental data. Furthermore, the influences of different initial crack length versus specimen depth ratios α0 and
specimen dimensions on the method in TPB and WS specimens were investigated.

2. Theoretical method to determine tensile strength

2.1. Fracture extreme method

The typical fracture characteristic of concrete is that a fracture process zone (FPZ) exists at the crack tip [35]. Several techniques
have been developed to measure the size and shape of the FPZ [36–39]. These studies indicate that the FPZ length at the peak load
and the maximum FPZ length increase with an increase in specimen height, but decrease by increasing α0 [36]. And both the length
and the width of FPZ are influenced by the maximum aggregate size [37]. Hillerborg et al. [40] proposed the fictitious crack model
(FCM) to describe the physical characteristics around the crack tip. In the FCM, the FPZ is equivalent to a fictitious crack, which is
assumed to be loaded with cohesive stress due to the aggregate interlock, the friction and the remaining bonding between aggregates
and mortar.
Fig. 1 shows the stress statuses of two different criteria around the crack tip. Several criteria have been used to evaluate the crack
propagation in concrete, e.g., stress criterion [40], toughness-based criterion [41,42], strain-based criterion [43–47], stress-based
criterion [48–50] and energy-based criterion [51].

Fig. 1. Stress status at the crack tip based on different criteria.

88
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

Fig. 2 is a typical P − a/D (where P is the external load, a is the effective crack length, and D is the depth of specimen) curve of
concrete fracture. Pini can be determined by utilising several test technologies, e.g., laser speckle interferometry, acoustic emission,
photoelastic coating and strain gauges [52]. For instance, the strain gauges method utilises a pair of electric resistance strain gauges
symmetrically pasted on both sides of crack tip and obtains a load–strain curve of the specimen. The distinct turning point on the
load–strain curve corresponds to the onset of stable crack propagation and the load at this point is taken as initial cracking load Pini of
specimen [53]. When P reaches Pini, the crack begins to propagate. Then P increases nonlinearly with a. When P reaches the Pmax, a
equals to its critical value ac. Subsequently, P decreases with a. Assuming the partial derivation of P to a is continuous at P = Pmax, the
following equation can be used to describe the extreme point on the P − a/D curve [31]:
∂P
=0
∂a a = ac (1)
Based on the fracture extreme method [31], the assumption of homogeneity and the initial fracture toughness criterion were
adopted in this study. The crack begins to propagate when the stress intensity factor at the effective crack tip KI reaches the initial
cracking toughness Kini. Fig. 3 illustrates a superposition scheme for calculating KI caused by the external load P and the cohesive
stress on the fictitious crack zone.

2.2. Theoretical method of determining tensile strength

According to RILEM [26,27], the TPB specimen is adopted as an example to illustrate the determining process of ft by using
fracture extreme method in this section. Fig. 4 shows the TPB specimen dimensions and related variables used in the calculation. B
and S are the width and span of the specimens, respectively. a0 is the initial crack length and ac is the critical crack length. σ(x) is the
cohesive stress, and the specimen is regarded as the standard TPB specimen when S/D = 4.
According to the initial fracture toughness criterion and the weight function method [32], the external load P can be expressed as
Eq. (2) [33]:
2BD 2 ⎡ 2 ⎤ W
P= g (a) + KIini
C −
3S a k (α ) ⎢
⎣ 2πa ⎥
⎦ 2 (2)
where
3(2Pini + W ) S
KIini
C = a 0 k (α )
4BD 2 (3)

1.99 - α (1 - α )[2.15 - 3.93α + 2.7α 2]


k (α ) =
(1 - α )3 2 (1 + 2α ) (4)
α = a/D, k(α) has a 0.5% accuracy for any α, g(a) can be expressed analytically by using four-term or five-term weight function [32]:

2 1 4 M1 2 4 M3 ⎡ a 3 a
g (a) = A1 a ⎛2s1 2 + M1 s + M2 s 3 2 + M3 s 2⎞ + A2 a2 ⎧ s 3 2 + s + M2 s5 2 + 1−⎛ 0 ⎞ −3s 0 ⎤ ⎫
3 2 ⎨ ⎢ a⎥ ⎦⎬
⎝ ⎠ ⎩3 2 15 6 ⎣ ⎝a⎠ ⎭ (5)
ft − σs (CTOD) a
where A1 = σs (CTOD) , A2 = a − a0
, and s = 1− a0 . The accuracy of g(a) is greater than 3%. M1, M2, and M3 can be expressed as a
function of α:

1
For i = 1 or 3: Mi = [ai + bi α + ci α 2 + di α3 + ei α 4 + fi α5]
(1 − α )3 2
For i = 2: Mi = ai + bi α

Pmax

Pini

a0/D ac /D a/D
Fig. 2. P − a/D curve [31].

89
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

P P
KI KIP KIı

ı(x)
= + ı(x)
ıs ıs

Fig. 3. The stress intensity factor at the crack tip KI (KI = KIP + KIσ).

Fig. 4. The configuration of TPB specimen.

The coefficients in M1, M2, and M3 are shown in Table 1.


The softening curve proposed by Reinhardt et al. [54] and the expression of the CTOD proposed by Jenq and Shah [55] are
expressed as Eqs. (6) and (7), respectively.
3
⎧ c CTOD ⎞ ⎤ −c CTOD ⎞ CTOD ⎫
σs (CTOD) = ft ⎡1+⎛ 1 exp ⎛ 2
⎜ −⎟ (1 + c13) exp(−c2)
⎜ ⎟

⎨⎢ ⎝ ω 0 ⎠⎦

⎝ ω0 ⎠ ω0 ⎬ (6)
⎩⎣ ⎭
where c1, and c2 are the concrete material parameters, ω0 is the maximum crack opening displacement at the crack tip.
1
a 2 a a a 2 2
CTOD = CMOD ⎧ ⎛1− 0 ⎞ + ⎛−1.149 + 1.081⎞ ⎡ 0 −⎛ 0 ⎞ ⎤ ⎫

⎩⎝ a⎠ ⎝ D ⎠⎢⎣a ⎝a⎠⎦ ⎥⎬
⎭ (7)
in which CMOD can be solved by LEFM [56].

6PSa ⎡ 0.66 ⎤
CMOD = 0.76−2.28α + 3.87α 2−2.04α3 +
BD 2E ⎢
⎣ (1−α )2 ⎥
⎦ (8)
For common concrete, the material parameters in Eq. (6) are taken as c1 = 3, c2 = 7, and ω0 = 160 μm [28]. Eq. (8) is valid for
0.2 ≤ α ≤ 0.7 with the maximum error of less than 0.5% and empirical formula with 1% accuracy for any α. According to the
hypothesis of the fracture extreme method, the partial derivation of P to a at P = Pmax is continuous when P reaches the Pmax and
correspondingly effective crack length a reaches the critical value ac.
Based on Eqs. (1) and (2),
∂P
= ζ ′ (ac ) × ft + η′ (ac ) = 0
∂a a = ac (9)

where
4BD 2 g ′ (ac ) × k (α c ) ac −g (ac )[k ′ (α c ) ac + k (α c )] 1
ζ ′ (ac ) =
3 2π S k 2 (α c ) ac 2 ft (10)

Table 1
Coefficients of four-term weight function parameters M1, M2, and M3.
i ai bi ci di ei fi

1 0.057201 −0.87416 4.046567 −7.89442 7.85497 −3.18832


2 0.493546 4.436494 – – – –
3 0.340417 −3.95341 16.19039 −16.0959 14.63025 −6.13065

90
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

BD 2 ac −1 2k (α c ) + 2ac1 2k (α c ) ini
η′ (ac ) = − KI C
3S ac k 2 (α c ) (11)

k′(α) and g′(α) are provided in Appendix A. Thus, the theoretical expression of ft can be obtained:

ζ ′ (ac )
ft = −
η′ (ac ) a = ac (12)

By substituting Eq. (2) into Eq. (1), an equation with only one unknown parameter ac, can be obtained. Consequently, the values
of tensile strength ft, critical crack tip opening displacement CTODc, and critical crack length versus specimen depth ac/D can be
determined using the obtained ac. A flowchart for determining ft is shown in Fig. 5 to illustrate the proposed method.

3. Results and discussions

3.1. Three-point bending notched beam

In this section, the TPB tests data of series B and C [57] were used to determine the tensile strength as well as the fracture
parameters. The ac/D, CTODc, and ft calculated by the proposed method are listed in Tables 2 and 3, respectively, to compare with ac/
D and CTODc calculated by Kumar and Barai [58] and ft obtained from the test [57]. The a0/D of small-size TPB beams have been
controlled in a certain range considering the influence of back boundary on the FPZ [59], i.e., a0/D ≤ 0.6. The empirical equation

Start

Input specimen
dimensions and
materials parameters

Determine CMOD by
solving Eq. (8).

Substitute CMOD
into Eq. (7), then get CTOD.

Calculate k(Į) by Determine ıs(CTOD) Calculate s by formula


solving Eq. (4). by Eq. (6). s = 1-a0/a.

Determine A1, A2 by formulas:


A1 = ıs(CTOD);
A2 = (ft-ıs(CTOD))/(a-a0).

Substitute A1, A2 into Eq. (5)


to obtain g(a).

Calculate Kini by Substituting k(Į) and g(a) into Eqs.


calculating Eq. (3) . (10) and (11) to obtain ȟ', Ș'.

Combining Eqs. (2) and (12)


to obtain ft

End

Fig. 5. The flowchart for determining ft.

91
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

Table 2
Comparison of calculated values of ft, CTODc, and ac for series B. (B × D × S = 76 mm × 203 mm × 762 mm, fc = 53.1 MPa, ft = 3.631 Mpa,
E = 38.4 GPa).
Specimen no. Kini/ a0/D Pini/N Pmax/N CTODc/μm ac/mm ft/MPa
MPa·m1/2
Results in Calculated (1)/(2) Results in Calculated (3)/(4) Calculated Relative
Ref. [58] results (2) Ref. [58] results (4) results error/%
(1) (3)

B01 0.789 0.383 3364 5523 12.68 16.46 0.77 88.51 93.35 0.95 3.46 4.7
B03 0.706 0.442 2480 4365 12.98 15.96 0.81 101.30 105.23 0.96 3.48 4.2
B04 0.494 0.319 2437 5612 17.89 18.45 0.97 85.67 87.06 0.98 3.59 1.2
B05 0.879 0.558 2179 3207 21.08 13.03 1.62 130.12 123.27 1.06 3.41 6.1
B15 0.614 0.376 2601 5033 13.22 17.02 0.78 89.52 94.93 0.94 3.45 4.9
B16 0.494 0.309 2511 5790 18.89 18.57 1.02 85.06 85.23 1.00 3.59 1.1
B17 0.552 0.362 2406 5166 20.97 18.01 1.16 96.63 93.79 1.03 3.60 0.9
B18 0.759 0.478 2397 4053 17.54 15.59 1.13 112.46 111.03 1.01 3.58 1.4
B19 0.819 0.495 2483 3919 11.82 14.68 0.81 109.42 112.99 0.97 3.47 4.4
B20 0.709 0.459 2357 4187 17.30 16.05 1.08 109.21 108.38 1.01 3.57 1.8
B31 0.934 0.45 3335 4855 27.50 14.68 1.87 114.49 103.15 1.11 3.35 7.7
B33 0.488 0.369 2061 4832 27.48 18.06 1.52 105.76 96.21 1.10 3.60 0.8
B34 0.617 0.575 1358 2494 22.46 13.54 1.66 137.63 129.73 1.06 3.45 4.9
B36 0.644 0.431 2310 4409 20.78 16.77 1.24 108.40 104.49 1.04 3.59 1.2
B37 0.872 0.455 3025 4676 22.71 15.43 1.47 111.85 105.27 1.06 3.52 3.1
B39 0.715 0.588 1534 2539 19.47 12.99 1.50 136.42 130.66 1.04 3.47 4.4
B40 0.758 0.49 2310 3830 24.42 14.97 1.63 121.60 112.97 1.08 3.49 4.0
B44 1.038 0.594 2321 3073 18.46 11.80 1.56 133.37 128.08 1.04 3.39 6.6
B45 0.972 0.55 2508 3563 16.42 13.46 1.22 123.42 121.15 1.02 3.55 2.1

Avg. – – – – 19.16 15.55 1.25 110.57 107.73 1.02 3.51 3.4

Table 3
Comparison of calculated values of ft, CTODc, and ac for series C. (B × D × S = 76 mm × 305 mm × 1143 mm, fc = 54.4 MPa, ft = 3.675 MPa,
E = 39.3 GPa).
Specimen no. Kini/ a0/D Pini/N Pmax/N CTODc/μm ac/mm ft/MPa
MPa·m1/2
Results in Calculated (1)/(2) Results in Calculated (3)/(4) Calculated Relative
Ref. [58] results (2) Ref. [58] results (4) results error/%
(1) (3)

C01 0.692 0.403 1123 6547 40.27 23.50 1.71 168.36 150.52 1.12 3.66 0.5
C02 0.761 0.437 1711 6057 39.14 22.31 1.75 175.38 157.99 1.11 3.63 1.1
C03 0.709 0.44 1802 5879 33.43 22.83 1.46 171.41 160.10 1.07 3.72 1.1
C04 0.857 0.478 3397 5612 29.15 21.36 1.36 175.38 167.19 1.05 3.70 0.8
C05 0.932 0.546 2292 4543 22.04 19.30 1.14 186.66 184.03 1.01 3.77 2.4
C06 0.739 0.515 3683 4543 31.15 20.45 1.52 190.02 178.98 1.06 3.69 0.4
C07 0.812 0.597 2082 3385 21.12 17.73 1.19 202.52 199.19 1.02 3.76 2.2
C15 0.469 0.453 2941 4899 35.98 23.40 1.54 182.39 168.77 1.08 3.76 2.3
C16 0.901 0.507 1930 5077 35.92 19.63 1.83 189.71 173.64 1.09 3.57 2.8
C17 0.897 0.556 4449 4276 20.44 19.05 1.07 188.19 187.18 1.01 3.76 2.3
C19 0.935 0.549 2081 4498 22.07 19.17 1.15 187.58 184.75 1.02 3.76 2.2
C20 0.705 0.504 2915 4676 26.95 21.17 1.27 183.00 176.92 1.03 3.74 1.8
C21 0.853 0.465 3394 5879 22.14 21.87 1.01 163.79 163.93 1.00 3.73 1.4
C22 0.839 0.381 1584 7660 21.93 23.74 0.92 138.78 141.71 0.98 3.68 0.2
C23 0.407 0.379 3309 6013 23.89 25.02 0.95 148.54 150.90 0.98 3.71 0.9
C24 0.725 0.426 2713 6124 17.14 22.50 0.76 147.93 155.73 0.95 3.61 1.8
C26 0.75 0.529 726 4276 14.80 19.60 0.75 175.68 182.25 0.96 3.63 1.3
C32 0.647 0.352 1838 7571 29.09 25.10 1.16 142.74 138.16 1.03 3.72 1.3
C33 0.628 0.551 3103 3697 24.56 19.92 1.23 195.51 190.56 1.03 3.77 2.5

Avg. – – – – 26.91 21.46 1.25 174.40 169.08 1.03 3.70 1.5

ft = 0.4983 fc [60] has been used to determine the experimental values of ft.
Tables 1 and 2 provide the information regarding the error rates of ft calculated by proposed method with different a0/D. As
shown, the calculated results are close to the experimental values. The maximum relative error is 7.7% and 2.8% for series B and C,
respectively, indicating that the accuracy of the calculated ft is insensitive to the variation in initial crack length ratio a0/D.
The relation between a0/D and fracture parameter CTODc are plotted in Fig. 6. The dashed red lines represent the linear fitting for
the calculated values in [58]. The CTODc obtained by the double-K method fluctuated with a0/D, while the CTODc calculated by the
proposed method is much steadier. The discrepancy may result from the measurement errors of the CMODc during the fracture tests.

92
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

30 45

40
25
35

CTOD c / ȝm
CTOD c / ȝm
20 30

25
15 20

15
10 Calculated values by proposed method Calculated values by proposed method
Calculated values by Kumar and Barai [58] 10 Calculated values by Kumar and Barai [58]
Linear fitting for Kumar and Barai [58] Linear fitting for Kumar and Barai [58]
5 5
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.35 0.40 0.45 0.50 0.55 0.60
a0/D a0/D
(a) B series (b) C series
Fig. 6. Comparison of calculated CTODc for series B (a) and C (b).

The proposed method does not require the measurement of the CMODc, which is inevitable for the double-K method. Kumar and Barai
[58] used the experimental values of CMODc to calculate the CTODc based on the double-K method. The CMODc is generally mea-
sured by the clip gauge method, which will lead to errors if gauge and specimen are improperly connected. As a0/D increases, the
CMODc is more difficult to be measured accurately.
Fig. 7 shows the calculated ac/D by two different methods with increasing a0/D. The results calculated by the proposed method
increased moderately in the range of 0.3 ≤ a0/D ≤ 0.6, while the results in [58] reflect an identical tendency but fluctuate within a
wider range. However, we can still conclude that the calculated ac/D of the proposed method agrees well with that of Kumar and
Barai [58].

3.2. Wedge splitting specimen

The WS test [18] is a convenient testing method for heavy concrete specimens since the weight of specimens can be supported,
and the crack growth in WS specimens is relatively stable and easy to control. Therefore, the WS test has also been widely used to
determine the fracture parameters of concrete [61–63]. In this section, the test data on WS specimens [61–63] were used to verify the
proposed method. The formulas developed for WS specimens can refer to the Appendix B. Figs. 8–11 and Tables 4–6 show the
comparisons of tensile strengths calculated by proposed method and those obtained by tests [61–63].
The experiments conducted by Zhao et al. [61] and Zhu [62] were used to investigate the influences of the specimen dimension
and the initial crack length. As shown in Tables 4 and 5, the calculated ft are slightly smaller than the existing experimental data but
within a allowable error range. The explanation for the errors between the experimental values and calculated results might be the
measurement errors of the CMODc during the fracture tests. Figs. 8 and 9 show a similar tendency for the curves obtained with
different Kini, indicating that the influence of Kini on the method is not evident.
Kim et al. [63] conducted tests with the specimen of different water–cement ratios and types of aggregate. Hence, different
softening curve parameters [32] (listed in Table 6) were adopted to determine the tensile strengths by the proposed method. Fig. 10
exhibits that the calculated ft are close to the experimental values, except for the slight discrepancies in specimens L20 and R20. As
shown in Table 6, the error rate is insensitive to the aggregate type which indicates that the aggregate type does not impact the

0.70 0.70

0.65 0.65
0.60
0.60
0.55
0.55
ac /D
ac /D

0.50
0.50
0.45
0.45
0.40
calculated values by proposed method 0.40 Calculated values by proposed method
0.35
calculated values by Kumar and Barai [58] Calculated values by Kumar and Barai [58]
0.30 0.35
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.35 0.40 0.45 0.50 0.55 0.60
a0/D a0/D
(a) B series (b) C series
Fig. 7. Comparison of calculated ac/D for series B (a) and C (b).

93
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

3.5 3.5
Calculated values based on Kini in Xu and Reinhard [64] Calculated values based on Kini in Xu and Reinhard [64]
Calculated values based on Kini in Kumar and Barai [32] Calculated values based on Kini in Kumar and Barai [32]
3.0 3.0
f t / MPa

f t / MPa
2.5 2.5

2.0 2.0

1.5 1.5
0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50

Specimen width B / m Specimen depth D / m

Fig. 8. Calculated ft with different specimen widths (a) and sizes (b) for the tests of Zhao et al. [61].

4.00

3.75 Calculated values based on Kini in Xu and Reinhard [64]


Calculated values based on Kini in Kumar and Barai [32]

3.50
ft / MPa

3.25

3.00

2.75

2.50

2.25
0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65
a0/D

Fig. 9. Calculated ft with different a0/D for the tests of Zhu [62].

5.0

4.5

4.0

3.5
ft / MPa

3.0

2.5

2.0
Calculated values by proposed method
1.5 Experimental values tested by Kim et al. [63]

1.0
L60 G60 R60 G40 L40 R40 R30 L20 G20 R20
The number of specimens
Fig. 10. Calculated ft for the tests of Kim et al. [63].

accuracy of the proposed method.


R20 and L20 specimens exhibit much higher values of Kini/Kun than other specimens (Kini/Kun = 0.89 and 0.82 for R20 and L20
specimens, respectively). Fig. 11 compares the values of Kini/Kun of all specimens (> 70) used in this study. It can be seen that Kini/
Kun are much higher for R20 and L20 specimens. In addition, Kini/Kun is generally less than 0.6 in other related studies [42,53,68–70].
Specimens R20, L20 and G20 have a water–cement ratio of 20% and the cylinder compressive strengths are over 80 MPa. For such

94
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

1.0

specimen R20
0.8 L20

Raito of Kini/Kun
0.6

0.4

0.2

0.0
B series C series Zhao [61] Zhu [62] Kim [63]
in [58] in [58]

Fig. 11. Kini/Kun of specimens in the tests.

high strength concrete specimens, the cracks tend to propagate through the aggregates rather than around them [32]. This may result
in a more brittle failure and specimens tend to have small FPZ and low values of the stress intensity factor caused by cohesive stress
KIσ (=Kun - Kini). In other words, brittle failure may occur when the value of Kini/Kun is great. Accurate measurements of Pini during
brittle failure can be challenging, and measuring errors of Pini may cause the discrepancy of the calculated tensile strength obtained
by the proposed method.

3.3. Verification by numerical simulation

In this section, the crack propagation processes of series C specimens [57] were simulated by finite element analysis (FEA). The
tensile strengths obtained by the proposed method in Section 3.1 were used as parameters in simulation. And the corresponding peak
loads Pmax were obtained to compare with the experimental peak loads. In addition, the cohesive elements technique was used to
simulate the propagation of cracks. The Mode I fracture energy GF, can be determined using the following expressions from Xu [65]
and the CEB-FIP model code [66].
0.95
GF = (0.0204 + 0.0053d max 8)(fcm fcmo )0.7 (13)

where fcm is the mean compressive strength, MPa, fcmo = 10 MPa, dmax = maximum size of aggregates, fcko = 10 MPa. According to
the CEB-FIP model code [66], the relation fcm = fck + 8 has been used and fck = the characteristics strength of concrete, MPa. As
mentioned by Xu and Reinhardt [67], dmax is 19 mm and GF can be obtained (112.6 N/m). Table 7 includes the values of the model
parameters used to simulate the concrete beam behaviour.
Fig. 12 shows the three-dimensional mesh adopted for modelling the fracture propagation of TPB specimen of concrete. In the
present mesh, 43,350 Lagrangian eight-node solid elements with different dimensions were used. The loading and boundary con-
ditions of the model were applied according to the particularities of the test set-up, as shown in Fig. 12. To localise the crack
formation at the notch region, the cohesive elements were assigned to the area above the notched zone (at the mid-span of the
specimen), while the remaining solid elements were assigned a linear elastic behaviour. The numerical analyses were performed
under displacement control with 0.004 mm increment at mid-span vertically.
Fig. 13 shows the P–CMOD curves calculated by FEA for series C specimens. Table 8 and Fig. 14 compare the values of peak loads
calculated by FEA and those obtained by tests. The good agreement indicates the accuracy of the proposed method (see Fig. 15).

4. Conclusions

A theoretical method was proposed to determine the tensile strength ft of concrete based on experimental loads in the fracture
tests. With the proposed method, the tensile strength ft, the critical effective fracture length ac, and the fracture parameter critical
crack tip opening displacement CTODc of standard TPB and WS concrete specimens were calculated. and following conclusions can be
drawn:

(1) Only two parameters Pini and Pmax need to be measured in the fracture tests. The results of ft calculated by proposed method agree
well with those obtained by the fracture test.
(2) The theoretical results of the method are insensitive to the α0 and specimen dimensions.
(3) Different parameters of softening curve, types of aggregate and values of Kini do not affect the accuracy of the proposed method.

The proposed method could be applied to other fracture tests, or it could be used to determine the dynamic tensile strength of
concrete in further studies. In addition, this study implies that theoretical relationship between fracture parameters and mechanical

95
L. Qing et al.

Table 4
Comparison of ft for the tests of Zhao et al. [61]. (fc = 20.47 MPa, ft = 2.75 MPa, E = 30.6 GPa).
Specimen no. D × 2H × B/m a0/D Pmax/N Kini/MPa·m1/2 ac/D ft/MPa

Results in Results in Results in Calculated (1)/(2) Results in Calculated (3)/(4) Calculated Relative Calculated Relative
Ref. [64] Ref. [32] Ref. [64] results with Ref. Ref. [32] results with Ref. results with error/% results with error/%
(1) [64] (2) (3) [32] (4) Ref. [64] Ref. [32]

T1-WS 0.2 × 0.2 × 0.15 0.5 4685 0.456 0.408 0.691 0.578 1.20 0.707 0.577 1.23 2.373 13.7 2.328 15.3
T2-WS 0.2 × 0.2 × 0.2 0.5 6003 0.387 0.331 0.686 0.587 1.17 0.702 0.588 1.19 2.487 9.6 2.511 8.7

96
T3-WS 0.2 × 0.2 × 0.3 0.5 10,519 0.592 0.539 0.662 0.562 1.18 0.681 0.562 1.21 2.274 17.3 2.232 18.8
T4-WS 0.2 × 0.2 × 0.45 0.5 11,889 0.216 0.163 0.712 0.609 1.17 0.725 0.609 1.19 2.629 4.4 2.692 2.1
V1-WS 0.15 × 0.15 × 0.15 0.5 3488 0.330 0.277 0.680 0.593 1.15 0.697 0.595 1.17 2.533 7.9 2.599 5.5
V2-WS 0.2 × 0.2 × 0.2 0.5 6003 0.387 0.331 0.686 0.587 1.17 0.702 0.588 1.19 2.487 9.6 2.511 8.7
V3-WS 0.3 × 0.3 × 0.3 0.5 16,008 0.762 0.693 0.641 0.555 1.15 0.661 0.555 1.19 2.321 15.6 2.296 16.5
V4-WS 0.45 × 0.45 × 0.45 0.5 36,137 0.913 0.802 0.594 0.560 1.06 0.619 0.562 1.10 2.569 6.6 2.656 3.4

Avg. – – – – – – – 1.16 – – 1.18 – 10.6 – 9.9


Engineering Fracture Mechanics 202 (2018) 87–102
L. Qing et al.

Table 5
Comparison of ft for the tests of Zhu [62]. (fc = 41.8 MPa, ft = 3.22 MPa, D × 2H × B = 0.2 m × 0.2 m × 0.1 m).
Specimen no. E/MPa a0/D Pmax/N Kini/MPa·m1/2 ac/D ft/MPa

Results in Results in Results in Calculated results (1)/(2) Results in Calculated results (3)/(4) Calculated Relative Calculated Relative
Ref. [64] Ref. [32] Ref. [64] with Ref. [64] (2) Ref. [32] with Ref. [32] (4) results with Ref. error/% results with Ref. error/%
(1) (3) [64] [32]

KZP-7 35,378 0.30 6500 0.471 0.294 0.472 0.425 1.11 0.506 0.435 1.16 3.080 4.3 3.300 2.4
KZP-11 35,268 0.30 6300 0.428 0.245 0.452 0.430 1.05 0.487 0.441 1.10 3.103 3.6 3.297 2.3
KZP-19 35,665 0.30 6250 0.410 0.234 0.496 0.432 1.15 0.529 0.442 1.20 3.104 3.6 3.336 3.5
KZP-16 30,258 0.35 6500 0.711 0.572 0.533 0.437 1.22 0.563 0.443 1.27 2.878 10.6 2.983 7.4
KZP-1 35,749 0.40 5750 0.669 0.527 0.510 0.487 1.05 0.542 0.495 1.09 3.062 4.9 3.211 0.3

97
KZP-4 32,801 0.40 5200 0.545 0.412 0.556 0.497 1.12 0.584 0.505 1.16 3.033 5.8 3.188 1.0
KZP-9 32,070 0.40 5700 0.669 0.531 0.514 0.486 1.06 0.545 0.493 1.11 3.065 4.8 3.188 1.0
KZP-17 31,932 0.40 5400 0.633 0.515 0.584 0.486 1.20 0.609 0.492 1.24 2.874 10.7 2.956 8.2
WP-6 32,523 0.40 5150 0.530 0.395 0.553 0.499 1.11 0.581 0.507 1.15 3.038 5.7 3.179 1.3
KZP-6 35,135 0.50 4500 0.748 0.655 0.593 0.563 1.05 0.617 0.565 1.09 3.031 5.9 3.086 4.2
KZP-10 27,333 0.50 4540 0.779 0.685 0.581 0.558 1.04 0.607 0.560 1.08 3.059 5.0 3.107 3.5
KZP-20 36,065 0.50 4680 0.839 0.764 0.617 0.553 1.12 0.640 0.553 1.16 2.804 12.9 2.730 15.2
KZP-12 34,263 0.60 3090 0.726 0.660 0.659 0.649 1.02 0.677 0.649 1.04 3.052 5.2 3.140 2.5
KZP-9 34,342 0.60 2800 0.630 0.602 0.707 0.654 1.08 0.721 0.649 1.11 2.909 9.7 2.776 13.8
KZP-21 35,069 0.60 3330 0.826 0.742 0.635 0.643 0.99 0.655 0.645 1.02 3.029 5.9 3.250 0.9

Avg. – – – – – – – 1.09 – – 1.13 – 6.6 – 4.5


Engineering Fracture Mechanics 202 (2018) 87–102
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

Table 6
Comparison of ft for the tests of Kim et al. [63]. (D × 2H × B = 0.2 m × 0.24 m × 0.12 m, a0/D = 0.5).
Specimen no. E/MPa fc/MPa c1 c2 ω0 Pmax/N Kini/ ac/D ft/MPa
MPa m1/2
Results in Calculated results (1)/(2) Calculated results Relative
Ref. [32] (1) with Ref. [32] (2) with Ref. [32] error/%

L60 29,940 31.27 1.18 7.95 188 3850 0.403 0.594 0.578 1.03 2.493 10.5
L40 33,720 47.41 1.18 7.51 166 4140 0.238 0.602 0.601 1.00 3.565 3.8
L20 40,570 82.80 1.18 6.70 140 7840 1.175 0.505 0.538 0.94 3.315 26.9
G60 28,790 32.57 1.18 7.92 185 3640 0.332 0.588 0.586 1.00 2.575 9.5
G40 33,400 45.85 1.18 7.55 167 3870 0.177 0.609 0.606 1.00 3.537 4.6
G20 39,570 85.68 1.18 6.64 138 7400 0.883 0.523 0.563 0.93 4.615 0.1
R60 26,510 34.87 1.18 7.85 181 3390 0.231 0.602 0.596 1.01 2.746 6.7
R40 31,580 55.28 1.18 7.32 159 3910 0.083 0.600 0.613 0.98 4.168 11.1
R30 34,350 66.91 1.18 7.04 150 5230 0.405 0.552 0.589 0.94 4.201 3.0
R20 38,140 88.80 1.18 6.57 137 8240 1.323 0.502 0.526 0.95 2.761 41.2

Avg. – – – – – – – – – 0.98 – 11.7

Table 7
Concrete properties used in the numerical simulations.
Poisson’s ratio 0.2
Young’s modulus 39.3 GPa
Fracture energy 112.6 N/mm
Crack width 2 mm
Cohesive elements width 0.2 mm

Fig. 12. Three-dimensional mesh of the TPB specimen.

8
a0/D = 0.352 (C32) Pmax = 7557 N
a0/D = 0.379 (C23) Pmax = 7034 N
a0/D = 0.381 (C22) Pmax = 6975 N
a0/D = 0.403 (C01) Pmax = 6634 N
a0/D = 0.426 (C24) Pmax = 6092 N
6
a0/D = 0.437 (C02) Pmax = 5885 N
a0/D = 0.440 (C03) Pmax = 5880 N
a0/D = 0.453 (C15) Pmax = 5660 N
a0/D = 0.465 (C21) Pmax = 5431 N
P/kN

a0/D = 0.478 (C04) Pmax = 5160 N


4 a0/D = 0.504 (C20) Pmax = 4655 N
a0/D = 0.507 (C16) Pmax = 4521 N
a0/D = 0.515 (C06) Pmax = 4420 N
a0/D = 0.529 (C26) Pmax = 4135 N
a0/D = 0.546 (C05) Pmax = 5958 N

2 a0/D = 0.549 (C19) Pmax = 3861 N


a0/D = 0.551 (C33) Pmax = 3820 N
a0/D = 0.556 (C17) Pmax = 3728 N
a0/D = 0.597 (C07) Pmax = 3142 N

0
0 200 400 600 800 1000
CMOD/ȝm
Fig. 13. Calculated P–CMOD curves by FEA.

98
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

Table 8
Comparison of peak loads by FEA and tests.
Number no. a0/D Calculated ft by proposed method/MPa Experimental peak load/N Calculated peak load by FEA/N Relative error/%

C01 0.403 3.66 6547 6634.3 1.32


C02 0.437 3.63 6057 5884.67 2.85
C03 0.44 3.72 5879 5880.11 0.02
C04 0.478 3.70 5612 5159.85 8.06
C05 0.546 3.77 4543 3958.23 12.87
C06 0.515 3.69 4543 4419.81 2.71
C07 0.597 3.76 3385 3142.17 7.17
C15 0.453 3.76 4899 5659.66 13.44
C16 0.507 3.57 5077 4521.11 10.95
C17 0.556 3.76 4276 3727.81 12.82
C19 0.549 3.76 4498 3860.62 14.17
C20 0.504 3.74 4676 4655.01 0.45
C21 0.465 3.73 5879 5431.23 7.62
C22 0.381 3.68 7660 6974.59 8.95
C23 0.379 3.71 6013 7034.52 14.52
C24 0.426 3.61 6124 6092.45 0.52
C26 0.529 3.63 4276 4134.67 3.31
C32 0.352 3.72 7571 7557.95 0.17
C33 0.551 3.77 3697 3820.43 3.23

10
Calculated peak loads by FEA
Experimental peak loads [57]
8

6
Pmax/kN

0
0.35 0.40 0.45 0.50 0.55 0.60
a0/D

Fig. 14. Comparison of peak loads by FEA and tests.

2H
a0

B
Fig. 15. The configuration of WS specimen.

99
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

parameters may be established in the future.

Acknowledgments

The work presented in the paper was supported by the National Natural Science Foundation of China (Nos. 51779069,
51578208), Natural Science Foundation of Hebei Province (No. E2017202030).

Appendix A

g′(a) and k′ (α) in Eqs. (10) and (11) can be expressed as follows:
1
k ′ (α ) = {(−2.15 + 12.16α−19.89α 2 + 10. 8α3)(1 + 2α )(1−α )3 2−(1.99−2.15α + 6.08α 2−6.63α3
D (1 + 2α 2)(1−α )3
3
+ 2.7α 4 ) × [2(1−α )3 2− (1 + 2α )(1−α )1 2]}
2 (a1)
1 2 2
g ′ (a) = (A1 + A′1 a)(2s1 2 + M1 s + M3 s 2) + A1 a (s−1 2 + M1 + M2 s1 2 + M3 s ) s′ + A2 a2 [(2s1 2 + M1 s + M2 s 3 2) s′
M2 s 3 2 +
2 3 3
3
M3 ⎛ a03 a0 a0 ⎤ 4 32 1 4 1 ⎛ a0 a0 ⎤
+ ⎜
2
−s′ + s 2 ⎞ + (2A2 a + A′2 a ) ⎡ s + M1 s + 2 5 2
M2 s + M3 1−⎛ ⎞ −3s ⎞
2 ⎝ a4 a a ⎠⎦ ⎣3 2 15 6 ⎝ ⎝a⎠ a ⎠⎦ (a2)
where
a0
s′ =
a2 (a3)

∂σs (CTOD) ∂σ (CTOD) ∂CTOD


A1′ = = s
∂a ∂CTOD ∂a (a4)

- A1′ (a−a0)−[ft −σs (CTOD)]


A2′ =
(a−a0 )2 (a5)
According to the Eq. (6)
2 3
∂σs (CTOD) −c CTOD ⎞ ⎡ c1 ⎛ c1 CTOD ⎞ c2 ⎛ c CTOD ⎞ ⎞ 1 ⎫
= ft ⎧exp ⎛ 2 ⎜ 3 ⎟ − ⎜1 + ⎛ 1

⎟− ⎟(1 + c13) exp(−c2)
⎜ ⎟

∂CTOD ⎨ ⎝ ω ⎢ ω
⎠⎣ 0 ⎝ ω ⎠ ω ⎝ ω0 ⎠ ⎠ ω0 ⎬
⎩ 0 0 0⎝
⎭ (a6)
∂CTOD
Substituting Eq. (8) to Eq. (7), the expression of can be obtained:
∂a
1 2
∂CTOD 6PS ⎡ 0.66 1.32α ⎤ ⎧ 2 a a 2
= 0.76−4.56α + 11.61α 2−8.16α3 + + s + (1.081−1.149α ) ⎡ 0 −⎛ 0 ⎞ ⎤ ⎫
∂a BD 2E ⎢
⎣ (1− α ) 2 (1− α ) 3 ⎥⎨
⎦⎩ ⎢
⎣ a ⎝ a⎠⎥ ⎦⎬⎭
−1 2
3PSa ⎡ 0.66 ⎤ ⎧ 2 a a 2
+ 0.76−2.28α + 3.87α 2−2.04α3 + s + (1.081−1.149α ) ⎡ 0 −⎛ 0 ⎞ ⎤ ⎫
BD 2E ⎢
⎣ (1− α ) 2 ⎥⎨
⎦⎩ ⎢
⎣ a ⎝ a⎠⎥ ⎦⎬⎭
2 2
1.149 ⎡ a0 a a a
× ⎧2ss′− −⎛ 0 ⎞ ⎤−(1.081−1.149α ) ⎡ 0 −2 03 ⎤ ⎫
⎨ D ⎢a ⎝a⎠⎦
⎣ ⎥ ⎢
⎣ a a ⎦⎥⎬ (a7)
⎩ ⎭

Appendix B

Formulas of the WS test which are different from those of TPB are listed as follows:
The external load P is expressed as [31]:
B D⎛ 2 ini ⎞
P= ⎜ g (a) + KIC ⎟

k (α ) ⎝ 2πa ⎠ (b1)
where
ini Pini
KIC = k (α )
B D (b2)
1
k (α ) = (2 + α )(0.886 + 4.64α−13.32α 2 + 14.72α3−5.6α 4 )
(1−α )3 2 (b3)
Eq. (8) is instead by:

100
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

P 1+α 2
CMOD = [2.163 + 12.219α−20.065α 2−0.9925α3 + 20.609α 4−9.9314α5] ⎛ ⎞
BE ⎝ 1−α ⎠ (b4)
Eqs. (10) and (11) are as follows:
2B D [g ′ (ac ) k (α c ) ac1 2−g (ac ) ac −1 2 (k ′ (α c ) ac + k (α c ) 2)]
ζ ′ (ac ) =
2π k 2 (α c ) ac (b5)

2B D k′ (α c ) ini
η′ (ac ) = KIc
k 2 (α c ) (b6)

References

[1] Darwin D, Scanlon A, Gergely P, et al. Cracking of concrete members in direct tension. J Am Concr Inst 1986;83(1):3–13.
[2] Sarfarazi V, Schubert W. Numerical simulation of tensile failure of concrete in direct, flexural, double punch tensile and ring tests. Periodica Polytech Civil Eng
2017;61(2):176.
[3] Swaddiwudhipong S, Lu HR, Wee TH. Direct tension test and tensile strain capacity of concrete at early age. Cem Concr Res 2003;33(12):2077–84.
[4] Zhang SL. Theoretical relationship and experimental comparison between splitting tensile strength and axial tensile strength of concrete. J Railway Sci Eng
1979(2):65–82+199 [in Chinese].
[5] ASTM. Committee C-9 on concrete and concrete aggregates. standard test method for splitting tensile strength of cylindrical concrete specimens. ASTM
International; 2011.
[6] ASTM. Standard Test Method for Flexural Strength of Concrete. Philadelphia, PA: American Society for Testing and Materials; 1999.
[7] Luong MP. Tensile and shear strengths of concrete and rock. Eng Fract Mech 1990;35(1):127–35.
[8] Jiang FT. Several problems in determination of tensile strength of concrete. J Hydroelect Eng 1986;9:27–32. [in Chinese].
[9] Raphael JM. Tensile strength of concrete. J Am Concr Inst 1984;81(2):158–65.
[10] Fédération Internationale du Béton(FIB). Model code for concrete structures 2010; 2010.
[11] American Concrete Institute. ACI 318-11 building code requirements for structural concrete and commentary; 2011.
[12] Popovics S. Strength and related properties of concrete: a quantitative approach. John Wiley and Sons; 1998.
[13] Zheng W, Kwan A, Lee P. Direct tension test of concrete. ACI Mater J 2001;98(1):63–71.
[14] Rashid MA, Mansur MA, Paramasivam P. Correlations between mechanical properties of high-strength concrete. J Mater Civil Eng-ASCE 2002;14(3):230–8.
[15] Shah SP. Determination of fracture parameters (KIcS and CTODc) of plain concrete using three-point bend tests. Mater Struct 1990;23(6):457–60.
[16] Golewski GL. Green concrete composite incorporating fly ash with high strength and fracture toughness. J Clean Prod 2017;172.
[17] Golewski GL. Improvement of fracture toughness of green concrete as a result of addition of coal fly ash. Characterization of fly ash microstructure. Mater
Charact 2017;134:335–46.
[18] Brühwiler E, Wittmann FH. The wedge Splitting test, a new method of performing stable fracture mechanics tests. Eng Fract Mech 1990;35(1–3):117–25.
[19] Wittmann FH, Rokugo K, Brühwiler E, et al. Fracture energy and strain softening of concrete as determined by means of compact tension specimens. Mater Struct
1988;21(1):21–32.
[20] Ince R. Determination of concrete fracture parameters based on two-parameter and size effect models using split-tension cubes. Eng Fract Mech
2010;77(12):2233–50.
[21] Shetty DK, Rosenfield AR, Duckworth WH. Mixed-mode fracture in biaxial stress state: application of the diametral-compression (Brazilian disk) test. Eng Fract
Mech 1987;26(6):825–40.
[22] Aliha MRM, Heidari-Rarani M, Shokrieh MM, et al. Experimental determination of tensile strength and KIc of polymer concretes using semi-circular bend (SCB)
specimens. Struct Eng Mech 2012;43(6):823–33.
[23] Aliha MRM. Indirect tensile test assessments for rock materials using 3-D disc-type specimens. Arab J Geosci 2014;7(11):4757–66.
[24] Bahmani A, Aliha MRM, Berto F. Investigation of fracture toughness for a polycrystalline graphite under combined tensile-tear deformation. Theor Appl Fract
Mech 2017.
[25] Mirsayar MM, Zollinger DG. Factors influencing stresses and movements in continuously reinforced concrete pavements – a review. Eng Solid Mech
2018;6(1):67–82.
[26] RILEM. Draft recommendation (50–FCM) Determination of the fracture energy of mortar and concrete by means of three-point bend test on notched beams.
Mater struct 1985;18(106):285–90.
[27] Draft RILEM. recommendations (TC 89–FMT) Determination of fracture parameters (KIcS and CTODc) of plain concrete using three-point bend tests, fracture
mechanics of concrete test methods. Mater Struct 1990;23:457–60.
[28] Zhang J, Leung CKY, Xu SL. Evaluation of fracture parameters of concrete from bending test using inverse analysis approach. Mater Struct 2010;43(6):857–74.
[29] Slowik V, Villmann B, Bretschneider N, Villmann T. Computational aspects of inverse analyses for determining softening curves of concrete. Comput Methods
Appl Mech Eng 2006;195(52):7223–36.
[30] Tin-Loi F, Que NS. Nonlinear programming approaches for an inverse problem in quasibrittle fracture. Int J Mech Sci 2002;44(5):843–58.
[31] Qing LB, Li QB. A theoretical method for determining initiation toughness based on experimental peak load. Eng Fract Mech 2013;99(1):295–305.
[32] Kumar S, Barai SV. Determining double-K fracture parameters of concrete for compact tension and wedge splitting tests using weight function. Eng Fract Mech
2009;76:935–48.
[33] Qing LB, Nie YT, Wang J, Hu Y. A simplified extreme method for determining double-K fracture parameters of concrete using experimental peak load. Fatigue
Fract Eng Mater Struct 2017;40:254–66.
[34] Qing LB, Dong MW, Guan JF. Determining initial fracture toughness of concrete for split-tension specimens based on the extreme theory. Eng Fract Mech
2018;189:427–38.
[35] Bažant ZP, Planas J. Fracture and size effect in concrete and other quasibrittle materials. Boca Raton and London: CRC Press; 1998.
[36] Wu ZM, Rong H, Zheng JJ, Xu F, Dong W. An experimental investigation on the FPZ properties in concrete using digital image correlation technique. Eng Fract
Mech 2011;78:2978–90.
[37] Trivedi N, Singh RK, Chattopadhyay J. Investigation on fracture parameters of concrete through optical crack profile and size effect studies. Eng Fract Mech
2015;147:119–39.
[38] Muralidhara S, Prasad BKR, Eskandari H, Karihaloo BL. Fracture process zone size and true fracture energy of concrete using acoustic emission. Constr Build
Mater 2009;24(4):479–86.
[39] Slate FO. X-ray technique for studying cracks in concrete with emphasis on methods developed and used at Cornell University. In: Wittmann FH, editor. Fracture
Mechanics of Concrete. Netherlands: Elsevier Science Publishers; 1983. p. 85–93.
[40] Hillerborg A, Modéer M, Petersson PE. Analysis of crack formation and crack growth in concrete by means of fracture mechanics and finite elements. Cem Concr
Res 1976;6(6):773–81.

101
L. Qing et al. Engineering Fracture Mechanics 202 (2018) 87–102

[41] Dong W, Zhou XM, Wu ZM. On fracture process zone and crack extension resistance of concrete based on initial fracture toughness. Constr Build Mater
2013;49(6):352–63.
[42] Dong W, Wu ZM, Zhou XM, Wang CJ. A comparative study on two stress intensity factor-based criteria for prediction of mode-I crack propagation in concrete.
Eng Fract Mech 2016;158:39–58.
[43] Chang KJ. On the maximum strain criterion—a new approach to the angled crack problem. Eng Fract Mech 1981;14(1):107–24.
[44] Mirsayar MM. Mixed mode fracture analysis using extended maximum tangential strain criterion. Mater Des 2015;86:941–7.
[45] Mirsayar MM, Park P. Mixed mode brittle fracture analysis of high strength cement mortar using strain-based criteria. Theor Appl Fract Mech 2016;86.
[46] Mirsayar MM, Berto F, Aliha MRM, et al. Strain-based criteria for mixed-mode fracture of polycrystalline graphite. Eng Fract Mech 2016;156:114–23.
[47] Mirsayar MM, Razmi A, Berto F. Tangential strain-based criteria for mixed-mode I/II fracture toughness of cement concrete. Fatigue Fract Eng Mater Struct
2017;41.
[48] Razmi A, Mirsayar M. On the mixed mode I/II fracture properties of jute fiber-reinforced concrete. Constr Build Mater 2017;148:512–20.
[49] Mirsayar MM, Razmi A, Aliha MRM, et al. EMTSN criterion for evaluating mixed mode I/II crack propagation in rock materials. Eng Fract Mech 2017;190.
[50] Mirsayar MM, Joneidi VA, Petrescu RVV, et al. Extended MTSN criterion for fracture analysis of soda lime glass. Eng Fract Mech 2017;178.
[51] Xu SL, Zhang XF. Determination of fracture parameters for crack propagation in concrete using an energy approach. Eng Fract Mech 2008;75(15):4292–308.
[52] Xu SL, Reinhardt HW. Determination of double-K, criterion for crack propagation in quasi-brittle fracture, Part I: experimental investigation of crack propa-
gation. Int J Fract 1999;98(2):111–49.
[53] Zhang XF, Xu SL. A comparative study on five approaches to evaluate double-K, fracture toughness parameters of concrete and size effect analysis. Eng Fract
Mech 2011;78(10):2115–38.
[54] Reinhardt HW, Cornelissen HAW, Hordjil DA. Tensile tests and failure analysis of concrete. J Struct Eng 1986;112(11):2462–77.
[55] Jenq Y, Shah SP. Two parameter fracture model for concrete. J Eng Mech 1985;111(10):1227–41.
[56] Tada H, Paris PC, Irwin GR. The stress analysis of crack handbook. St.Louis, Missouri, USA: Paris Productions Incorporated; 1973.
[57] Refai TME, Swartz SE. Fracture behavior of concrete beams in three-point bending considering the influence of size effects. 1987, Report No. 190, Engineering
Experiment Station, Kansas State University.
[58] Kumar S, Barai SV. Determining the double-K fracture parameters for three-point bending notched concrete beams using weight function. Fatigue Fract Eng
Mater Struct 2010;33(10):645–60.
[59] Wang YS, Hu XZ, Liang L, Zhu WC. Determination of tensile strength and fracture toughness of concrete using notched 3-p-b specimens. Eng Fract Mech
2016;160:67–77.
[60] Karihaloo BL, Nallathambi P. Notched beam test: mode I fracture toughness, fracture mechanics test methods for concrete. In: Shah SP, Carpinteri A, editors.
Report of RILEM Technical Committee 89-FMT. London: Chapman & Hall; 1991. p. 1–86.
[61] Zhao GF, Jiao H, Xu SL. Study on fracture behaviour with wedge splitting test method. In: van Mier JGM, Rots JG, Bakker A, editors. Fracture processes in
concrete, rock and ceramics. E. & F. N. Spon; 1991. p. 789–98.
[62] Zhu XF. Transport organischer Flüssigkeiten in Betonbauteilen mit Mikro-und Biegerissen: Eindring-und Durchströmungsvorgänge umweltgefährdender Stoffe
an feinen Trennrissen in Beton. Beuth; 1997.
[63] Kim JK, Lee CS, Park CK, Eo SH. The fracture characteristics of crushed limestone sand concrete. Cem Concr Res 1997;27(11):1719–29.
[64] Xu SL, Reinhardt HW. Determination of double-K, criterion for crack propagation in quasi-brittle fracture, Part III: compact tension specimens and wedge
splitting specimens. Int J Fract 1999;98(2):179–93.
[65] Xu SL. Determination of parameters in the bilinear, Reinhardt’s non-linear and exponentially non-linear softening curves and their physical meanings. Werkstoffe
Werkstoffprüfung im Bauwesen, Hamburg. Libri BOD 1999:410–24.
[66] CEB-Comite Euro-International du Beton-EB-FIP Model Code. 1990. Bulletin d’information no. 2123/214, Lausanne; 1993.
[67] Xu SL, Reinhardt HW. Determination of double-K, criterion for crack propagation in quasi-brittle fracture, Part II: analytical evaluating and practical measuring
methods for three-point bending notched beams. Int J Fract 1999;98(2):151–77.
[68] Hu SW, Xu AQ. Experimental validation and fracture properties analysis on wedge splitting concrete specimens with different initial seam-height ratios. Proc
Struct Integrity 2016;2:2818–32. [in Chinese].
[69] Wu ZM, Xu SL, Ding Y, et al. The double-K fracture parameter of concrete for non-standard three point bending beam specimens. Eng Sci 2001;3(4):76–81. [in
Chinese].
[70] Wu ZM, Xu SL, Liu JY. Study on crack propagation process of concrete and size effect of double-K fracture parameter by means of photoelastic coatings. J
Hydraul Eng 2001;32(4):34–40. [in Chinese].

102

You might also like