Construction and Building Materials 202 (2019) 58–72

Contents lists available at ScienceDirect

Construction and Building Materials

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

Mechanical properties tests and multiscale numerical simulations

for basalt fiber reinforced concrete
Xinjian Sun a,b, Zhen Gao a, Peng Cao a,c,⇑, Changjun Zhou d
a
School of Hydraulic and Electric Engineering, Qinghai University, Xining, People’s Republic of China
b
State Key Laboratory of Plateau Ecology and Agriculture, Xining, People’s Republic of China
c
College of Architecture and Civil Engineering, Beijing University of Technology, Beijing, 100124 China
d
School of Transportation & Logistics, Dalian University of Technology, Dalian, People’s Republic of China

h i g h l i g h t s

A damage constitutive model was developed for BFRC.

FE model with the damage model was utilized to simulate BFRC mechanical properties.

Effects of basalt fiber on the BFRC mechanical properties was investigated.

Parabolic equation proposed for describing the effect of the aggregate and fiber volume contents.

Regression formula obtained for size effect of mechanical properties of the optimum BFRC.

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

Article history: Basalt fiber reinforced concrete (BFRC) can be regarded as a composite of cement mortar, aggregate and
Received 18 July 2018 basalt fiber. In this paper, the influences of the basalt fiber’s length and content on the fundamental
Received in revised form 8 November 2018 mechanical properties of concrete were investigated by multi-scale simulation. At the mesoscopic scale,
Accepted 3 January 2019
a damage constitutive model was developed in accordance with the Mori-Tanaka homogenization theory
and progressive damage theory to predict the composited material properties of BFRC. At the macro-
scopic scale, the obtained material properties of BFRC from mesoscopic were input into the finite element
Keywords:
specimen model to simulate the mechanical performance of BFRC. By coupling the mesoscopic material
Basalt fiber reinforced concrete
Mechanical properties
model with the finite element macroscopic model, the effects of basalt fiber on the mechanical perfor-
Numerical simulation mance of BFRC specimen at macroscopic scale can be investigated. The compressive, splitting tensile,
Size effect and bending performances of BFRC were studied by both experiments and numerical simulation. The pre-
dicted results from the proposed multiscale model show a good agreement with the experimental results.
It is also found that with the increase of fiber content, the compressive and splitting tensile strengths of
concrete increase first followed by decrease while the bending strength keeps increase. For the different
lengths of basalt fiber, the BFRCs with 6 mm basalt fiber display a better compressive and splitting tensile
performance than BFRCs with 12 mm fiber, whereas the differences in bending strength is slight. Results
shows that BFRCs with 2‰ 6 mm basalt fiber can achieve the maximum strength. Moreover, the size
effect on BFRC’s basic properties of BFRC with 2‰ 6 mm basalt fiber was further explored by regression.
Those regression formula has very high R2 values.
Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction more complicated, the plain concrete with poor toughness and
brittleness cannot satisfy the requirement of mass concrete struc-
Due to the brittle fracture characteristics of cement concrete, ture for better durability and safety [4,5]. In recent years, due to
there is a potential fracture risk in engineering construction under harsh natural conditions such as arid and cold, large temperature
the effect of load and environment [1–3]. Especially, in the huge difference, high radiation and salt, fracturing phenomenon in con-
dam structure. Due to the influence of load and environment is crete have arisen on many large hydraulic structures on the Qing-
hai Tibet Plateau [6]. It should be notable that this kind of issue is
⇑ Corresponding author. being gradually solved with the development and application
E-mail address: [email protected] (P. Cao). of high strength concrete materials in the past thirty years. By

https://doi.org/10.1016/j.conbuildmat.2019.01.018
0950-0618/Ó 2019 Elsevier Ltd. All rights reserved.
 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 59

introducing toughening materials such as fiber, the compressive Although it has been proved that the mechanical properties of
strength, shear and fracture resistance of fiber reinforced concrete concrete can be greatly improved by adding fibers, the research
can be significantly improved, being widely introduced by many on the mechanical model of fiber reinforced concrete is relatively
papers [7–9]. Meanwhile, fiber reinforced concrete materials have backward. In the available model related studies, the fracture
been widely fabricated in hydraulic and civil buildings around the model of fiber reinforced concrete was mentioned firstly [39]. For
world in past decades [10,11]. In the research and application of instance, in the paper published by Zhang Jun [40], the increase
fiber reinforced concrete materials, carbon fiber is the most com- of fracture strength factor was used to present the increase of frac-
mon reinforced concrete material which has many advantages ture strength caused by fiber. On the other hand, based on the
such as lightweight, high strength, good stability, excellent combi- Hillerberg cohesive method [41], some studies were focused on
nation with cement and so on. However, due to its high cost, the the fracture strength of fiber reinforced concrete by simulation
application of carbon fiber reinforced concrete has been greatly and analyzing. According to the research published by Enfedaque,
limited since more and cheaper alternative fibers were used a fracture toughness cohesive force model of glass fiber reinforced
[12,13]. On the other hand, in the field of civil engineering, due concrete was predicted with the use of ABAQUS [42]. In addition,
to high strength and good cement combination performance, steel because the robustness and applicability of concrete can be verified
fiber is commonly used which can improve the fracture resistance by the elastoplastic damage model [43–45], in consequence, this
of concrete significantly [14–17]. By the mean of numerical calcu- model with various kinds of fiber content was applied by some
lation and experimental study, the toughening mechanism of steel scholars to study the fracture and damage behavior of fiber rein-
fiber for concrete under dynamic and static conditions has been forced concrete [46]. Based on the research conducted by Chira
studied by Su. It is suggested that a proper amount of steel fiber and Kumar, the damage behavior of concrete with different fiber
can increase the fracture resistance of concrete by several times contents was simulated by using the elastoplastic damage model
[18]. Nieuwoudt has found that the strength of the interface of concrete, it had been found that the numerical simulation had
between the steel fiber and cement matrix is at least equivalence good agreement with the experimental results [47]. With the
to that of the concrete, indicating that the steel fiber concrete development of computer technology, more concrete models from
material is more suitable for resistant fracture behavior [19]. How- mesoscopic have been initiated to study the fracture and stiffness
ever, as a hydraulic material, the concrete will be exposed to water degradation behavior of fiber reinforced concrete under the ulti-
perennial. For steel fiber reinforced concrete materials, water con- mate load [48–50]. For example, in the paper presented by Xu, a
vection and diffusion happens and eventually, the steel fiber is mesoscopic concrete model was established. In which, two-
eroded. In terms of this reason, the application of steel fiber rein- dimensional finite element method was applied to simulate the
forced concrete in hydraulic structures is strictly limited [20,21]. aggregate distribution and the long and narrow unit was recog-
Polyethylene fiber (short for PE fiber) reinforced concrete is also nized as fiber. Based on this simulation, the failure behavior of fiber
known as a common high toughness concrete material [22]. By reinforced concrete under indirect tension and impact compres-
considering double-K fracture criterion, the fracture energy of sion were simulated successfully [51,52]. By employing the beam
polythene reinforced concrete was studied by Cao. It has been element, the steel fiber in the steel fiber reinforced material was
found that the increase of PE fiber content has insignificant influ- simulated and a finite element model of three-dimensional steel
ence on fracture-initiation toughness, whereas it can improve the fiber reinforced concrete was established by Liang and Wu. In this
unstable fracture toughness obviously [23,24]. According to the model, the fiber was distributed randomly and a dense of network
research published by Bilodeau and Zhu [25,26], with a certain structure was formed. Meantime, the concrete was treated as an
amount of PE fiber, the fire resistance and freeze-thaw resistance equivalent isotropic medium and the embedded contact was sim-
of polyethylene fiber reinforced concrete material can be greatly ulated for the fiber and concrete. Through which, the secondary
improved and prompted. However, the industrial polyethylene hardening phenomenon of steel fiber reinforced concrete under
fiber materials appear in cluster form usually which will result in the ultimate bearing capacity was explained rationally [53]. On
difficulties in mixing procedure. Due to the wide use of cement the other hand, by applying CT scanning and three-dimensional
and fiber in hydraulic concrete, if the mixing effect of polyethylene finite element method, the anisotropic behavior of fiber reinforced
fiber and cement cannot be guaranteed, there will be a great poten- concrete was predicted and analyzed by Qsymah [54]. From the
tial risk on the safety and integrity of the structure [27]. As a sub- mesoscopic point of view, the aggregate and fiber distribution
stitute fiber, regarding on its excellent performance, basalt fiber should be considered in the process of finite element modeling.
has drawn great attention world-widely. With the industrialization Due to huge amount of calculation should be taken, this kind of
of basalt fiber, it has been widely applied in aviation, aerospace and model is not suggested for large-scale structure simulation and
automobile fields. In civil engineering and hydraulic engineering, analyzing. Therefore, it is urgent to develop a more competitive
basalt fiber is often cut into short pieces and mixed with concrete computing method in terms of accuracy and efficiency.
material enhance the fracture strength of concrete [28,29]. Com- At present, the research of basalt fiber reinforced concrete
pared with polythene fiber reinforced concrete, reinforced concrete (BFRC) in water conservancy is just in its start-up step. Especially,
with basalt fiber is not only conducive to cost saving [30], but also for the models of short-cut basalt fiber-reinforced concrete, there is
has better performance on binding strength with cement due to its no comprehensive experimental study and simulation yet. To fill
chemical property is more similar with cement [31,32]. Whereas this gap, based on experimental and finite element methods
the polyethylene fiber needs to be modified from both Chemical (FEM), this study aims to investigate the mechanical properties
and physical property to enhance the binding strength of the inter- of BFRC by multi-scale coupling analyses.
face between PE fiber and cement [33,34].
In the use of basalt fiber, it is usually weaved into specific form
[35] and the knitted basalt fabric has good performance on shock 2. Research objective and scope
resistant and fracture toughness [36,37]. However, it should be
noticed that the knitted basalt fiber cannot be widely used in the The purpose of this study is to explore the effect of basalt fiber
massive concrete structure. On the other hand, when using long on the fundamental mechanical properties of concrete based on
basalt fiber as enhanced phase, the concrete material has more experimental investigation and finite element modeling. The nov-
capability on fracture resistant and it has been used in the field elty of this study is that the multi-scale modeling method have
of civil construction and hydraulic construction [38]. been employed to simulate the compressive, indirect tensile and
60 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

bending strength of BFRC with different fiber contents. In meso- the particle gradation was shown in Table 2. On its saturated sur-
scale, the BFRC is regarded as a composite of plain concrete with face, the dry apparent density of gravel stone as well as middle
basalt fiber in which the basalt fiber is distributed to the matrix stone were 2680 kg/m3 and 2660 kg/m3, respectively. In addition,
of plain concrete. This composite is looked as a new equivalent water reducing agent (water reduction rate of 17% and gas content
medium and its material properties is calculated by Mori-Tanada of 1.3%) and Air entraining agent (water reduction rate of 6% and
method. Besides, to take the damage behavior and strength under gas content of 5.2%) were included as well.
different loading modes, Hashin method is adopted to define the The basalt fibers with 6 mm and 12 mm in length [55], manu-
damage strength of matrix, i.e. plain concrete, under tensile, com- factured by Shanghai Chen Qi Chemical Technology Co., Ltd, were
pressive and shear load. By this way, the damage strengths under selected in this research as shown in Fig. 1 and its performance
different load modes will not impact on each other, which is diffi- indexes were presented in Table 3.
cult to be realized in conventional elastoplastic models. In the
transformation between the meso-scale and macro-scale constitu- 3.2. Concrete mix proportion
tive models, firstly, the meso-scale properties of concrete are cal-
culated by MT method. Then, an applicable Jacobian matrix is The strength grade C25, the waterproofing grade W8 and the
generated by MT method from the calculated results. After that, frost resistance grade F300, named C25W8F300, were applied as
the generated Jacobian matrix is transferred to the UMAT subrou- the control group for the benchmark concrete. The mix proportion
tine of ABAQUS to analyze the mechanical performance of concrete of C25W8F300 is listed in Table 4. The water reducer and air entrain-
in macro-scale. Through this process, the multiscale modeling ing agent was added by the volume of cementitious materials, i.e., 1%
analysis can be realized by coupling the meso-scale properties and 0.04%, respectively. The 15–30 min slump was measured as 90–
with macro-scale performances of basalt fiber reinforced concrete. 110 mm while the density of fresh concrete is 2380 kg/m3.
Using this method, the compressive, splitting and bending
strengths of BFRC can be analyzed by coupling fiber contents with
3.3. Experiments
fiber lengths. Based on that, the change trends of compressive,
splitting and bending strengths of BFRC with the fiber content
During the preparation of basalt fiber reinforced concrete spec-
and the fiber length can be derived and the optimum combination
imens, the dry mixing of fiber method was proposed to improve the
of fiber content and fiber length in the engineering application can
dispersion. According to literatures [56,57], 0‰, 1‰, 2‰, 3‰, 4‰
be determined. In the end, the size effect of aggregate, as well as
the influence of aggregate content, on mechanical properties of
BFRC, were investigated by finite element method based on energy
releasing theory. Additionally, the regression formula were pro-
posed for the size effect of BFRC with determined basalt content
on mechanical properties.

3. Raw material and experimental method

3.1. Raw material

P.O 42.5 ordinary Portland cement was selected and its perfor-
mance indexes were shown in Table 1. The involved grade 2 fly
ashes were produced by Yongdeng Lian Electric Fly Ash Co., Ltd.
The fine and coarse aggregates were collected from the natural
aggregate and crushed gravel on site. Through tests, for fine aggre-
gate, its fineness modulus was 3, and the apparent density was
2680 kg/m3. On the other hand, for the second gradation coarse
aggregate, its particle size in the range from 5 mm to 40 mm and Fig. 1. Short chopped basalt fibers.

Table 1
Measured physical and mechanical properties of cement.

Cement kind Strength rank Bending Compressive Setting time Standard consistency Stability
strength strength
3d 28 d 3d 28 d
1 P.O 42.5 5.2 7.8 20.1 45.2 2.35 4.15 29 1.5
Code value P.O 42.5 3.5 6.5 17.0 42.5 45 min 10 h – 5.0

Table 2
The coarse aggregate grain composition.

Small aggregate Medium aggregate

Size (mm) Percentage Content (%) Cumulative percentage (%) Size (mm) Percentage content (%) Cumulative percentage (%)
<4.75 2.0 2.0 <19.0 2.8 2.8
4.75–9.5 32.8 34.8 19.0–26.0 29.3 32.1
9.5–16.0 55.5 90.2 26.0–31.5 44.4 76.5
16.0–19.0 6.6 96.8 31.5–37.5 23.5 100.0
>19.0 3.2 100.0 >37.5 0 100.0
 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 61

Table 3
The basalt fiber performance parameters.

Diameter (lm) Length (mm) Tensile strength (MPa) Elasticity modulus (GPa) Fracture elongation (%) Density (kg/m3)
17.4 6 (12) 2000 85 2.5 2699

Table 4
Mix proportion of the control group for the benchmark concrete (kg).

Water Cement Fly ash Sand Small aggregate Medium aggregate Water reducer Air entraining agent
130 217 54 694 708 708 2.71 0.108

and 5‰ of each type of basalt fiber in volume of fresh concrete were was no obvious change. However, once the maximum failure load
added. Therefore, there are totally 11 types of concrete were pre- was reached, a sudden rupture happened as shown in Fig. 2c which
pared and tested. For each test, 3 duplicated specimens were pre- shows obvious brittle.
pared and measured. 150 mm cubic specimens prepared for
compression and splitting tests while 100 mm  100 mm  4.2. Experiment results
400 mm rectangle specimens were prepared for three-point bend-
ing strength test. The total specimens were cured in standard curing 4.2.1. Experiment result calculation
room for 28 days before testing. According to elastic theory, the compressive strength of con-
crete cube was calculated regarding on formula (1):
4. Results and analyses on BFRC mechanical properties F
f cc ¼ ð1Þ
A
4.1. BFRC failure mode
In Eq. (1), f cc stands for compressive strength (MPa), Fis the
The failure modes of BFRC with different fiber volumes and maximum load when specimen failed (N) and A is the loading area
lengths were similar to that of ordinary concrete which were pre- (mm2). The calculated result presented in Table 5.
sented in Fig. 2. During the compression experiment, the surface The splitting tensile strength of concrete cube was calculated
and corner of specimen cube began to bulge and peel off. With regarding on formula (2):
the increase of stress load, cracks appeared gradually and extended
2F F
throughout the whole specimen finally as shown in Fig. 2a. In the f ts ¼ ¼ 0:637 ð2Þ
splitting test, a micro crack appeared in the middle of the speci- pA A
men. After releasing the load, with slightly knock, the test speci- In Eq. (2), f ts stands for splitting tensile strength (MPa), F is the
men was broken from the middle as shown in Fig. 2b. The failure maximum load when specimen failed (N) and A is the area of split-
section was mainly focused on the binding intersection between ting interface (mm2). The calculated result presented in Table 5.
coarse aggregate and cement base. Occasionally, the coarse aggre- The bending strength of concrete cube was calculated regarding
gate might be split as well. For bending test, during the load, there on formula (3):

Fig. 2. Concrete failure mode.

62 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

Table 5
28-day concrete compression, splitting tensile and bending strength test results.

Specimen Compression strength (MPa) Standard deviation Splitting tensile strength (MPa) Standard deviation Bending strength (MPa) Standard deviation
PC 29.9 0.258 2.29 0.036 3.83 0.036
B-6-0.1 31.4 0.852 2.36 0.121 4.21 0.057
B-6-0.2 33.4 1.875 2.38 0.092 4.02 0.110
B-6-0.3 37.2 1.462 2.50 0.108 3.89 0.093
B-6-0.4 35.9 1.042 2.51 0.051 4.32 0.042
B-6-0.5 33.9 0.726 2.40 0.045 4.91 0.094
B-12-0.1 30.5 0.497 2.28 0.096 4.00 0.083
B-12-0.2 32.4 2.139 2.27 0.057 4.10 0.059
B-12-0.3 33.5 1.289 2.26 0.036 3.95 0.139
B-12-0.4 26.8 0.589 2.28 0.057 4.78 0.064
B-12-0.5 25.5 0.497 2.20 0.057 4.94 0.054

3Fl 4.2.4. Bending strength

ff ¼ 2
ð3Þ
2bh To a certain extent, the bending strength can reflect the tough-
ness of concrete. Regarding on Table 5, it can be noted that no mat-
In Eq. (3), f f stands for bending strength (MPa), F is the maxi-
ter what the fiber lengths are, the bending strength of BFRC
mum load when specimen failed (N), b represents the section increase with the incorporation of fibers and reaches the maximum
width of specimen (mm), h is the section height of specimen value at a 5‰ content. The slight decrease found at 3‰ content
(mm) and l stands for the distance between supports which equals may be due to the dispersion problem, i.e., the fibers may be entan-
to 3 times of section height (mm). The calculated result presented gled together for some specimens.
in Table 5 as well.
The compressive, splitting and bending test results and their
5. Numerical simulation on BFRC mechanical properties
standard deviations were all listed in Table 5. The serial number
PC means the ordinary concrete specimen and the serial number
In order to theoretically analyze the mechanical properties of
of basalt fiber reinforced concrete specimen were named in the
BFRC under the mechanism of fracture, a model for predicting
form of B-x-y in which B stands for basalt fiber reinforced concrete,
the mechanical properties of basalt fiber reinforced concrete was
x represents for the length of fiber and y means fiber volume ratio.
established based on multiscale theory. From mesoscopic level,
For instance, B-6-0.1 means the specimen was made of basalt fiber
the fiber reinforced concrete material is equivalent to plain con-
with 6 mm in length and the total fiber volume was 1‰ of the total
crete and basalt fiber, of which the plain concrete is the matrix
specimen.
and basalt fiber is the inclusion. Moreover, from more mesoscopic
level, the plain concrete is equivalent to cement mortar, aggregate
4.2.2. Compressive strength and interface, of which cement mortar is the matrix phase and the
The compressive strength of the standard cube is an important aggregate wrapped by the interface is the inclusion phase. By
indicator for dividing the strength grade of concrete. From Table 5, applying the Mori-Tanaka homogenization method, the mechani-
no matter the length of fiber was 6 mm or 12 mm, with the fiber cal properties of each equivalent medium under different volume
volume increasing, the compressive strength increased first then composition can be predicted.
decreased. On the other hand, the compressive strength of mixed
concrete with 6 mm basalt fiber was higher than that with 5.1. Mean-field homogenization (MFH) for Two-Phase composites
12 mm. In addition, it can be seen that the maximum compressive

strength was reached in the test specimen containing 2.5‰–3‰ of In the multiscale method, for every macro point x, the macro
basalt fiber in volume. By analyzing the results listed in Table 5, the  
strain e is given and the macro stress r can be derived by solving
influence of fiber length and content on concrete was suggested as
boundary value problem (BVP), and vice versa. In mesoscale level,
that: due to the excellent mechanical properties and good compat-
the macro point is regarded as the center of the representative vol-
ibility with cement base, basalt fiber reinforced concrete with 
6 mm fiber had higher compressive strength over that of ordinary ume element (RVE), i.e. x 2 x, and the boundary condition is @ x.
concrete. Secondly, because of uneven mixing and fibrous mass This multiscale method is illustrated in Fig. 3. As the energy equiv-
phenomenon, when test specimen containing basalt fiber with alence in both scales, and therefore the strain-stress relationship in
12 mm in length or with relatively high volume, the compressive
strength was relatively low.

4.2.3. Splitting tensile strength

Splitting tensile strength was also measured which reflect the
tensile strength of concrete indirectly. According to Table 5, for
specimens with 6 mm basalt fiber, the tensile strength increased
first then decreased with the increasing of fiber volume and its
splitting tensile strength was greater than that of ordinary con-
crete. For specimen with 12 mm basalt fiber, the splitting tensile
strength decreased gradually with fiber volume increasing. By
means of analysis and calculation, it has been found that the split-
ting tensile strength was proportional to the compressive strength
and the tension-compression ratio was in the range of 0.062–
0.086. Fig. 3. Schematic diagram of the multiscale method.
 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 63

macroscale can be transferred to the relationship of average strain

 
hei and average stress hri, namely e ¼ hei and r ¼ hri.
The definition of stress mean field f on RVE is:
Z  
1
hf i ¼ f x; x dv ð4Þ
v x

In the equation, f is the internal stress/strain field, x is the macro
point, x is the meso point.
The definition of mean stress field on both matrix and inclusion
phases is:
Z  
1
hf ixr ¼ f x; x dv r ; r ¼ 0; 1 ð5Þ
vr xr

Where 0 and 1 represent for matrix and inclusion, respectively.

Then Eq. (6) stands for the relation of average strains of RVE, the
matrix, and the inclusion, while Eq. (7) stands for the relation of
average stress.
heix ¼ v 0 heix0 þ v 1 heix1 ð6Þ

hrix ¼ v 0 hrix0 þ v 1 hrix1 ð7Þ

where v 0 and v 1 are the volume of matrix and inclusion, respec- Fig. 4. Schematic diagram for double inclusion model.
tively, v0 þ v1 ¼ 1.
When a linear displacement occurred on the RVE’s boundary,
the average strain of each inclusion can be defined by the strain model, the fiber of stiffness C1 is surrounded with the plain con-
concentration tensor, Be : crete of stiffness C0 in its close surroundings while the outside
those areas are assumed as an unknown reference medium of stiff-
heix1 ¼ Be : heix0 ð8Þ ness Cr.
By applying the Mori-Tanaka homogenization method with
Then there exists a relation between heix1 , heix0 , and hei:
multi-level inclusions in DIGIMAT software, the constitutive model
heix1 ¼ Be : ½v 1 Be þ ð1  v 1 ÞI : hei
1 of BFRC was predicted. Firstly, the elastic mechanical parameters of
ð9Þ
the plain concrete were determined. The elastic mechanical prop-
where I represents the symmetric equivalent tensor, changing with erties of each phase were shown in Table 6. The elastic moduli
Be . and Poisson’s ratios of cement mortar, aggregate, and interface
Since MFH is based on Eshelby tensor, according to Eshelby were cited from Stock et al. [58] while the densities of aggregate
solution, the strain inside inclusion is evenly distributed and is and cement mortar were obtained in laboratory. It is noted that
related to the remote strain. the density of interface was assumed to be the same as cement
mortar. Therefore, the basic elastic parameters of the plain con-
eðxÞ ¼ He ðI; C 0 ; C 1 Þ : E; 8x 2 I ð10Þ
crete can be calculated by Mori-Tanaka homogenization method,
e which is shown in Table 7. Secondly, combining Tables 7 and 3,
where H represents the strain concentration tensor in a single
inclusion, which is defined 其中, as: the elastic parameters of plain concrete and fiber can be utilized
    in the Mori-Tanaka homogenization method to predict the
He ðI; C 0 ; C 1 Þ ¼ ½I þ nðI; C 0 Þ : C 1
0 : C 1  C 1
0  ð11Þ mechanical parameters of BFRC.

where nðI; C 0 Þ is Eshelby tensor, related to the shape and properties

of material. 5.2.2. Progressive damage model
Then the macro equivalent stiffness tensor of concrete can be For unidirectional fiber reinforced composites, the three-
obtained: dimensional Hashin failure criterion is widely applied [59]. In
 h   i  1 order to describe the strength of fiber concrete, the three-
C¼ v 1 C 1 : Be þ ð1  v 1 ÞC 0 : v 1 Be þ ð1  v 1 ÞI ð12Þ

And the relationship between stress and strain in macro scale Table 6
is: Elastic mechanical parameters of plain concrete in numerical simulation.

   Elasticity Poisson’s ratio Density (kg/m3)

r ¼ C:e ð13Þ modulus (GPa)

For a two-phase composite with moderate fiber volume, this Cement mortar 13.4 0.25 2000
method is quite accurate. All the phases in mesoscale are assumed Aggregate 74.5 0.15 2650
Interface 5.36 0.30 2000
ideally homogenous and isotropic for BFRC.

5.2. Mesoscale model for BFRC

Table 7
5.2.1. Double inclusion model for BFRC Elastic mechanical parameters of plain concrete calculated from Mori-Tanaka
In this study, the BFRC is regard as a composite that basalt fiber homogenization method.
embedded into the plain concrete. The double inclusion model is Elasticity modulus (GPa) Poisson’s ratio Density (kg/m3)
applied to simulate the plain concrete and basalt fiber composited
23.76 0.22 2260
matrix and the schematic diagram is presented in Fig. 4. In this
64 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

dimensional Hashin failure criterion, failure indicator and expo- Table 10

nential damage evolution methods were applied to simulate the Damage variable value of MLT model.

damage process of BFRC. Damage mode Damage variable value

Fiber damage D1 ¼ /ðmaxff 1 ; f 2 gÞ
(1) Three Dimensional Hashin Failure Criterion and Failure Matrix damage D2 ¼ /ðmaxff 3 ; f 4 gÞ
Indicator Transverse shear damage of Composite materials D3 ¼ 1  ð1  D1 Þð1  D2 Þ
Transverse shear damage of Composite materials D4 ¼ D3

Generally, the three-dimensional Hashin criterion can be

applied to simulate the failure behavior of fiber composite materi-
als. With the purpose to ensure the integrity of materials in differ- damage evolution function, each damage variables were
ent failure modes, four failure indicators were introduced: fiber determined.
tensile failure indicator F 1 , fiber compression failure indicator F 2 , In damage evolution Eqs. (14) and (15), a and b are material
matrix tensile failure indicator F 3 and matrix compression failure response parameters separately. In this paper, the value of a was
indicator F 4 . The value of each failure indicators can be determined set to 1 and the constitutive model can be obtained by changing
 
b which can result in stress-strain curve softening following Wei-
by the failure function F rf ¼ 1 and if any failure indicator value
bull distribution. The value of b can be estimated from the position
was equal or greater than 1, failure on composite material hap- with the maximum stress. When the value of b tended to be infin-
pened then. In Table 8, direction 1 was defined as the direction ity, the exponential damage evolution model was turned into
of fiber by all failure criteria. The direction perpendicular to the instantaneous damage mode. In this paper, from the trail calcula-
fiber in the plane was direction 2 treated as the matrix direction tion, we found when b was set to 0.5, the numerical results
and the normal direction of the plane was direction 3. X t and X c matched well with experimental results. Therefore, the value of b
are the tensile strength and compressive strength in fiber direction was set to 0.5 [60]. Fig. 5 shows the relationship between the fail-
of composite materials; Y t and Y c are the tensile strength and com- ure indicator f and the damage variable D during the damage evo-
pressive strength in matrix direction of composite materials, lution process when the response parameter of the material was
respectively; S and SI are the shear strength between composites. changed.
The detailed strength properties of BFRC were shown in Table 9.
D ¼ /ðf Þ; 0  D  1 ð14Þ
(2) Damage Evolution (
0 f  f min
In accordance with the Matzenmiller-Lubliner-Taylor (MLT) uðf Þ ¼ f ab f ab ð15Þ
Dmax  ð1  expð eb min ÞÞ f  f min
model, the damage evolution in this paper was finalized by assum-
ing each phase material of concrete subjecting to the Weibull dis- Finally, according to the definition of failure criterion, damage
tribution in meso-scopic level. In fact, the damage evolution is the variables and the law of damage evolution, the progressive damage
strain energy releasing process. Thus, the material is believed to be model of BFRC in the mesoscopic was obtained. Moreover, the
softened. In other words, it is the degradation of elastic modulus damage model was used as material property and imported into
and decreasing of bearing capacity from the macroscopic point. ABAQUS to simulate the basic mechanical properties of BFRC.
The mentioned soften form in this paper was expressed in expo-
nential form. In the period of damage evolution, the damage vari- 6. Simulation and verification
able was treated as the key issue due to it could reflect the damage
mechanism of materials directly. The damage variable D in this 6.1. Comparison on experimental and numerical results of mechanical
research was expressed by the damage evolution function /ðf Þ properties of BFRC
which can explain the relationship between failure indicator f
and damage variable D more accurately. As Table 10 shows, using 6.1.1. Compressive, splitting, and bending strengths of BFRC
The geometric size of the simulated model was the same as the
experimental size as shown in Fig. 4. After doing concrete model
Table 8 mesh, the hexahedral mesh was generated. After calculation, the
Three-dimensional Hashin failure criterion.

Failure mode Failure criterion

Fiber tensile failure r11  0 F 1 ðrÞ ¼
r211
þ
r212 þr313
X 2t S2
r11 < 0 r11
Fiber compression failure F 2 ðrÞ ¼ Xc
Matrix tensile failure r22 þ r33  0 F 3 ðrÞ ¼ ðr22 þr33 Þ
þ
2 r212 þr213
þ
r223 r22 r33
Y2 t S2 S2I
r2 þr2 r223 r22 r33
Matrix compression failure F 4 ðrÞ ¼ ðr22 þr2 33 Þ þ 12 2 13 þ
2

r22 þ r33 < 0 4S

 I  
S S2I
2
Yc r22 þr33
þ 2S I
 1 Yc

Table 9
Strength properties of basalt fiber reinforced concrete.

Strength parameter (direction) Symbol Strength (MPa)

Transverse tensile (1) Xt 1.49
Transverse compression (1) Xc 13.70
Longitudinal tensile (2,3) Yt 1.49
Longitudinal compression (2.3) Yc 13.70
Shear (1–2,1–3) S 3.39
Shear (2–3) SI 3.39 Fig. 5. The relationship between the failure indicator f and the damage variable D
during the damage evolution process.
 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 65

model presented in Fig. 4a had 4735 units and 5524 nodes, while
the model in Fig. 4b had 3855 units and 4600 nodes. For the model
presented in Fig. 4c, 5650 units and 6677 nodes were contained.
Within the finite boundary model, the bottom surface was defined
as one of the displacement boundaries with a fixed displacement
limiting the horizontal and vertical displacement while the upper
surface was defined as the other displacement boundary which is
a free boundary. By applying displacement load, the counterforce
at support was obtained. The strength parameters of the model
then were calculated by the input of displacement and counter-
force at support.
The compression, splitting and bending experiments and the
corresponding numerical models of BFRCs are shown in Fig. 6.
The numerical models analyzed by finite element method reflected
the strain distribution change during the stress loading. The result
of each test and numerical simulation of the compression, splitting
and bending are drawn in Figs. 7–9 respectively. It can be seen that
the numerical simulation result had a good agreement with the Fig. 7. Comparisons of compression experiment and numerical simulation of BFRC.
experimental data, indicating that the proposed multi-scale dam-
age model is useful for simulating the strength of BFRC. In addition,
through a comprehensive evaluation of numerical simulation and
experimental results, when the length of fiber was 6 mm and the
volume was 2‰, the finite element calculation results of compres-
sive strength, splitting tensile strength and flexural strength were
very close to the experimental results.

6.1.2. P-CMOD of BFRC in three-point bending test on notched beams

In this study, the load-crack mouth opening displacement, i.e.,
P-CMOD, in the three-point bending test on notched beams, as
shown in Fig. 10, was utilized to describe the deformation charac-
teristics of BFRC. The BFRC with 2‰ 6 mm basalt fiber were chosen
as the test concrete. Three groups of notched beams were tested as
shown in Table 11 and for each group three duplicated specimens
were prepared. The proposed finite element model was used to
simulate this kind of test. The numerical and experimental test
data were compared in Fig. 11. It can be seen that the numerical Fig. 8. Comparisons of splitting tensile experiment and numerical simulation of
results are very close to the experimental ones, indicating the BFRC.
effectiveness of the proposed method.

Fig. 6. Compression, splitting tensile and bending experiment and numerical model of BFRC.
66 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

Fig. 9. Comparisons of bending experiment and numerical simulation of BFRC. Fig. 11. Numerical results versus experimental results on P-CMOD curves of three-
point bending tests on notched BFRC beams.

content is 1.03 times 0.3 aggregate content’s. For the BFRCs incor-
porated with 6 mm fiber, all the compressive strengths of BFRCs
were higher than ordinary concrete. However, for the BFRCs intro-
duced with 12 mm fiber, the compressive strengths increase first
and then drop and reach the maximum value with a fiber content
of 3‰. Fig. 13(c) and (d) displayed the splitting strengths of 6 mm
and 12 mm length of fibers’ BFRCs with different aggregate con-
tent, respectively. The splitting strengths increase with aggregate
content increasing and the splitting strength of 0.7 aggregate con-
tent is 1.06 times larger than 0.3 aggregate content’s. In compar-
Fig. 10. The sketch of three-point bending test on notched beams. ison, the splitting strengths of BFRC with 6 mm fiber were
superior to BFRC with 12 mm fiber. The bending strengths of BFRCs
with different aggregate content were shown in Fig. 13(e) and (f).
Table 11 From the simulated results, the bending strengths increased with
Geometry information of specimens for three-point bending test on notched beams. an increasing aggregate content for both BFRCs with 6 mm and
Group Geometry (l  h  t) ðmm  mm  mmÞ a0 ðmmÞ S ðmmÞ 12 mm fibers. Moreover, the 0.7 aggregate content’s bending
strength is 1.07 times of the 0.3 aggregate content’s BFRCs.
Ⅰ 480  180  60 36 400
Ⅱ 640  240  80 48 540 The predicted results of the impacts of varied aggregate content
III 800  300  100 60 660 on the BFRCs’ mechanical performances from numerical simulation
are meaningful for the durability of concrete and the optimal com-
ponents of concrete’s raw materials. So as to quantitatively evalu-
ate the effects of two variables, i.e. aggregate content, fiber content
6.2. The influence of the aggregate volume in determining the strength
and the fiber’s size, on the mechanical performances of BFRCs,
of BFRC
regression analysis was conducted based on simulated mechanical
performances’ results. Since the regression analysis is related to
Furthermore, the volume contents of the aggregates in the
two variables, multiple regression analysis was adopted and con-
basalt fiber reinforced concrete material were investigated and dis-
ducted using STATA software. The regression formulas for com-
cussed by means of the multiscale simulation method.
pressive, splitting and bending strengths were shown below.
From the aforementioned analysis, it was identified that the
From Eqs. (16)–(21), the results showed that all the regression
multi-scale damage model proposed in this paper can accurately
equation can fit well with the simulation values.
predict the tensile, compression and bending strengths of BFRC
The compressive strength of BFRC with 6 mm fiber:
under ultimate strength. However, the effect of aggregate content
on the strength of BFRC also should be concerned. Therefore, the f ¼ 586; 571a2 þ 4027:18a þ 1:826b þ 27:81 ð16Þ
mechanical performances of BFRCs at different aggregate content
were also investigated based on the constituted model. Further- R2 ¼ 0:92
more, the impacts of aggregate content, fiber content and the
length of fiber on compressive, splitting and bending strengths The compressive strength of BFRC with 12 mm fiber:
were studied from mesoscopic. f ¼ 1; 733; 024a2 þ 9626:66a þ 39:45b ð17Þ
In this study, the mass ratios from 0.3 to 0.7 with a gap of 0.1
were selected and investigated. The compression, splitting and
R2 ¼ 0:96
bending strengths tests were simulated on the built multi-scale
damage model. The simulating results were illustrated in Fig. 12. Splitting strength of BFRC with 6 mm fiber:
Fig. 12(a) and (b) displayed the impacts of different aggregate con-
f ¼ 21; 121a2 þ 137:31a þ 0:329b þ 2:07 ð18Þ
tent on the compressive strength which indicated that the com-
pression strength of BFRC increases with the aggregate content
increasing. The compressive strength of BFRC with 0.7 aggregate
R2 ¼ 0:93
 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 67

Fig. 12. Impacts of aggregate content on BFRC’s mechanical properties.

68 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

Fig. 13. The strength size effect of BFRC.

 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 69

Splitting strength of BFRC with 12 mm fiber: plain concrete. In Fig. 8c and d, the size effect on splitting tensile
strength of basalt fiber reinforced concrete was described. It can
f ¼ 6485a þ 24:98a þ 0:324b þ 2:08
2
ð19Þ be seen that although the splitting tensile strength decreased with
the increase of specimen size, the tensile strength decreasing trend
R2 ¼ 0:97 became steady with the increase of specimen size decreasing.
Bending strength of BFRC with 6 mm fiber: Moreover, the splitting tensile strength size effect of specimens
with different basalt fiber content and fiber length was basically
f ¼ 30; 888a2 þ 2:93a þ 0:64b þ 3:55 ð20Þ in the same trend. Fig. 8e and f show the bending strength size
effect of BFRC, compared with the standard specimen size
R2 ¼ 0:91 (100 mm  100 mm  400 mm), the bending strength of non-
standard sized specimen was significantly lower. It can be seen
Bending strength of BFRC with 12 mm fiber: that the larger specimen size the smaller bending strength and
f ¼ 32; 520a2 þ 46:67a þ 0:637b þ 3:49 ð21Þ the size effect on bending strength became less significant. Firstly,
this phenomenon can be explained by the size effect of the con-
crete itself. On the other hand, because the length and the amount
R2 ¼ 0:97
of basalt fiber were relatively small, the improvement of bending
f stands for compressive, splitting and bending strengths. a is the strength on large scaled specimen was not significant.
content of fiber, whose values are 0‰, 1‰, 2‰, 3‰, 4‰ and In order to describe the strength size effect of BFRC quantita-
5‰; b is the aggregate content whose value are 0.3, 0.4, 0.5, tively, the basic mechanic property size effect of BFRC was ana-
0.6 and 0.7. lyzed in accordance with the size effect based on energy release
From Eqs. (16)–(21), it can be noted that the regression predic- theory proposed by Bazant [61]. For specimen with similar geo-
tion models agree well with the simulated results for the compres- metric, its size effect law of the structural nominal strength was
sive, splitting and bending strengths of BFRC, which indicates that written as follows:
these efficient regression models have a great significance for the
Bf
application of BFRC in large hydraulic structures. rN ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
t
ð22Þ
1 þ D=D0
6.3. Numerical analysis of the size effect on BFRC In equation,rN is the structural nominal strength, B is the non-
dimension constant, f t is the strength of the concrete structure,
Based on the achieved experimental data and calculated value, D0 is the geometric constant and D is the characteristic size of
the multi-scale damage model proposed in this paper can predict the structure.
the strength of fiber reinforced concrete accurately. Because of This equation reveals the relationship between structural nom-
the structural characteristics of the concrete itself, the failure pro- inal strength rN and characteristic size of structure D. Through
cess and mechanism are quite complicated, which causes the size experiment, the value of structural nominal strength, characteristic
effect of concrete strength. In order to verify the feasibility of size of structure and strength of concrete structure were obtained.
multi-scale damage model on predicting the strength of fiber rein- So the law of size effect can be derived once the value of non-
forced concrete with various sizes, the finite element compression, dimensional constant and geometric constant were determined.
splitting and bending models of basalt fiber reinforced concrete In order to facilitate the regression analysis of experimental data,
with different size were further established. Eq. (22) can be turned into a linear equation according to the prin-
The volumes of geometrical models were scaled to 2, 4, 6 and 8 ciple of the least square method.
times of the original model as shown Table 12. Through the finite
element analysis, the compressive strength, splitting tensile Y ¼ AX þ C ð23Þ
strength and bending strength of basal fiber concrete, considering
In which: X ¼ D; Y ¼ r2 ; C ¼ 1 1
; A ¼ C=D0
the size effect, were presented in Fig. 13. N ðBf t Þ2

According to Fig. 13, the obvious feature of size effect had been In Eq. (23), the value of X and Y are already known, by inputting
reflected by the proposed model. From Fig. 8.a and b, when the size the strength and size of each specimen, the value of A and C can be
of the specimen was smaller than 238 mm, the compressive derived then. In this paper, with the purpose to reduce the work-
strength decreased relatively fast with the size of specimen scaling load of numerical calculation, only the basic mechanical properties
up. For instance, the compressive strength of plain concrete cube of concrete with the optimum fiber content and length was applied
specimen with a length of 189 mm was decreased by 0.6 MPa over to achieve the size effect calculation formula. The corresponding
a specimen with 150 mm in length. Once the length of the speci- value of A and C were listed in Table 13.
men was larger than 238 mm, the compressive strength change In turn, the value geometric constant D0 and the multiplication
became less significant with the volume increasing. For example, of the non-dimensional constant B and the strength of concrete
the compressive strength of plain concrete cube specimen with a structure f t were obtained as shown in Table 14.
length of 300 mm was only decreased by 0.13 MPa over a specimen By substituting the value of D0 and Bf t in Eq. (22), the nominal
with 273 mm in length. In terms of this fact, it is believed that the strength size effects of BFRC specimens (containing 2‰ volume of
compressive strength of BFRC has a similar size effect change of the fiber with 6 mm in length) were derived as follows respectively.

Table 12
Geometric dimension of BFRC specimens.

Magnification times Compression (width  height  length)/mm Splitting tensile (width  height  length)/mm Bendingn (width  height  length)/mm
1 150  150  150 150  150  150 100  100  400
2 189  189  189 189  189  189 126  126  504
4 238  238  238 238  238  238 159  159  636
6 273  273  273 273  273  273 182  182  728
8 300  300  300 300  300  300 200  200  800
70 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

Table 13 The fitting curve of the size effect disclosed by Eqs. (24)–(26)
Calculation formula parameters of BFRC’ size effect with the optimum fiber content was described in Fig. 14. According to which, the theory of size
and length.
effect caused by energy release had a good applicability to analyze
A C the strength size effect of BFRC and the corresponding fitting for-
Compression 4:28  107 8:24  104 mula was consistent with the simulation result.
Splitting tensile 1:15  103
1:02  102 The obtained results above show that the influence of basic
Bending 6:08  104 5:00  103 mechanical properties of concrete with various basalt fiber content
and length. In next, the influence on size effect of concrete material
with different fiber content will be further studied and investigated.
Table 14
D0 and of Bf t BFRC’ size effect.
7. Conclusions
D0 Bf t
Compression 1924.46 34.84 In this study, mechanical properties of BFRC were investigated
Splitting tensile 8.92 9.89
Bending 8.22 14.14
through a series of experiments and multiscale numerical simula-
tion. The following conclusions can be drawn:

The specimen in compression test:
According to basic mechanical performance test of basalt fiber
34:84 reinforced concrete, with the increase of fiber content, the com-
rN ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð24Þ pressive and splitting tensile strength of concrete increased first
1 þ D=1924:46
followed by decreasing. Meanwhile, the bending strength
The specimen in splitting test: increased with the increasing of fiber volume. According to
the compressive and splitting tensile test, it is found that the
9:89
rN ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð25Þ mechanical performance of concrete with 6 mm fiber was bet-
1 þ D=8:92 ter than that with 12 mm fiber.
The specimen in bending test:
By applying Mori-Tanaka homogenization algorithm and con-
tinuum progressive damage theory, the damage model of con-
14:14
rN ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð26Þ crete was established to obtain material inputs for numerical
1 þ D=8:22

Fig. 14. Fitting curves of the strength size effect of BFRC.

 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72 71

simulation in finite element model. Numerical and experimen- [19] P.D. Nieuwoudt, A.J. Babafemi, W.P. Boshoff, et al., The response of cracked
steel fiber reinforced concrete under various sustained stress levels on both
tal results matched with each other very well.
the macro and single fiber level, Constr. Build. Mater. 156 (2017) 828–843.

Basalt fiber with a length of 6 mm and 2‰ of volume can mostly [20] V. Marcos-Meson, A. Michel, A. Solgaard, et al., Corrosion resistance of steel
improve the mechanical properties of concrete. fiber reinforced concrete – a literature review, Cem. Concr. Res. (2017).

A parabolic equation has been proposed for describing and pre- [21] C.G. Berrocal, K. Lundgren, I. Löfgren, Corrosion of steel bars embedded in fiber
reinforced concrete under chloride attack: state of the art, Cem. Concr. Res. 80
dicting the influence of the aggregate and fiber volume con- (2015) 69–85.
tents. The regression results can fit well with the simulated [22] S.M. Abtahi, M. Sheikhzadeh, S.M. Hejazi, Fiber-reinforced asphalt-concrete – a
results. review, Constr. Build. Mater. 24 (6) (2010) 871–877.
[23] P. Cao, D. Feng, C. Zhou, Study on fracture behavior of polypropylene fiber

The regression formula for the size effect of compressive reinforced concrete utilizing three-point bending beam test, Comput. Concr. 5
strength, splitting tensile strength and bending strength of the (14) (2017) 527–546.
optimum BFRC (6 mm in length and 2‰ volume of concrete) [24] A.C. Aydin, Self compactability of high volume hybrid fiber reinforced
concrete, Constr. Build. Mater. 21 (6) (2007) 1149–1154.
had good agreements with numerical simulation. [25] A. Bilodeau, V.K.R. Kodur, G.C. Hoff, Optimization of the type and amount of
polypropylene fibers for preventing the spalling of lightweight concrete
subjected to hydrocarbon fire, Cem. Concr. Compos. 26 (2) (2004) 163–174.
Conflict of interest [26] C. Zhu, J.G. Niu, J.J. Li, C.J. Wan, J.H. Peng, Topcu, Effect of aggregate saturation
degree on the freeze–thaw resistance of high performance polypropylene fiber
lightweight aggregate concrete, Constr. Build. Mater. 145 (2017) 367–375.
None. [27] M. Yamaguchi, K. Murakami, K. Takeda, et al., An experimental study on mix
proportion for high toughness of fiber-reinforced concrete with high
molecular weight polyethylene fiber and its basic characteristics, J. Struct.
Acknowledgements Constr. Eng. 73 (634) (2008) 2091–2100.
[28] J. Lian, Experimental research on the mechanical behavior of chopped basalt
fiber reinforced concrete, Indust. Constr. 37 (6) (2007) 8–10.
This work is financially supported by the National Natural [29] A.K. Mohanty, L.T. Drzal, M. Misra, Engineered natural fiber reinforced
Science Foundation of China (Grant No. 51769028, 51508137, polypropylene composites: influence of surface modifications and novel
powder impregnation processing, J. Adhes. Sci. Technol. 16 (2000) 999–1015.
51509139), Natural Science Foundation of Qinghai Province in
[30] J. Li, J. Niu, C. Wan, et al., Comparison of flexural property between high
China (Grant No. 2018-ZJ-750, 2017-ZJ-933Q), and Fundamental performance polypropylene fiber reinforced lightweight aggregate concrete
Research Funds for the Central Universities (DUT17RC(3)006). and steel fiber reinforced lightweight aggregate concrete, Constr. Build. Mater.
Thanks for the associate professor Leng Zhen in the Hong Kong 157 (2017) 729–736.
[31] V.A. Rybin, A.V. Utkin, N.I. Baklanova, Alkali resistance, microstructural and
Polytechnic University to provide many valuable discussions. mechanical performance of zirconia-coated basalt fibers, Cem. Concr. Res. 53
(2) (2013) 1–8.
[32] B. Wei, H. Cao, S. Song, Tensile behavior contrast of basalt and glass fibers after
References chemical treatment, Mater. Des. 31 (9) (2010) 4244–4250.
[33] C. Zhang, V.S. Gopalaratnam, H.K. Yasuda, Plasma treatment of polymeric
fibers for improved performance in cement matrices, J. Appl. Polym. Sci. 76
[1] Y. Ding, S. Liu, The investigation on the workability of fiber cocktail reinforced
(2000) 1985–1996.
self-compacting high performance concrete, Constr. Build. Mater. 22 (2008)
[34] A.M. López-Buendía, M.D. Romero-Sánchez, V. Climent, et al., Surface treated
1462–1470.
polypropylene (PP) fibers for reinforced concrete, Cem. Concr. Res. 54 (2013)
[2] H. Higashiyama, N. Banthia, Correlating flexural and shear toughness of
29–35.
lightweight fiber-reinforced concrete, ACI Mater. J. 105 (3) (2008) 251–257.
[35] W. Mingchao, Z. Zuoguang, L. Yubin, et al., Chemical durability and mechanical
[3] Z. Zhang, C.T. Hsu, Shear Strengthening of Reinforced Concrete Beams Using
properties of alkali-proof basalt fiber and its reinforced epoxy composites, J.
Carbon-Fiber-Reinforced Polymer Laminates, J. Compos. Constr. 9 (2) (2005)
Reinf. Plast Compos (2008).
158–169.
[36] A.K. Bandaru, S. Patel, Y. Sachan, et al., Low velocity impact response of 3D
[4] P.K. Mehta, P.J.M. Monteiro, Concrete: Microstructure, properties and
angle-interlock Kevlar/basalt reinforced polypropylene composites, Mater.
materials, 3 ed., McGraw-Hill, New York, 2006.
Des. 105 (2016) 323–332.
[5] Y. Zhang, M. Zhang, Transport properties in unsaturated cement-based
[37] N. Morova, Investigation of usability of basalt fibers in hot mix asphalt
materials – a review, Constr. Build. Mater. 72 (2014) 367–379.
concrete, Constr. Build. Mater. 47 (2013) 175–180.
[6] J.F. Huo, D.P. Liu, Experimental study on frost resistance durability of
[38] D. Wang, L. Wang, X. Gu, et al., Effect of basalt fiber on the asphalt binder and
polypropylene fiber reinforcement lightweight aggregate concrete, Low
mastic at low temperature, J. Mater. Civ. Eng. 25 (3) (2013) 5–364.
Temperature Arch. Technol. (2008).
[39] V.C. Li, E. Liang, Fracture processes in concrete and fiber reinforced
[7] S. Yin, R. Tuladhar, F. Shi, M. Combe, T. Collister, Nagaratnam Sivakugan, Use of
cementitious composites, J. Eng. Mech. 112 (6) (1986) 566–586.
macro plastic fibers in concrete: a review, Constr. Build. Mater. 93 (2015) 180–
[40] J. Zhang, V.C. Li, Simulation of crack propagation in fiber-reinforced concrete
188.
by fracture mechanics, Cem. Concr. Res. 34 (2) (2004) 333–339.
[8] D.J. Kim, A.E. Naaman, S. El-Tawil, Comparative flexural behavior of four fiber
[41] A. Hillerberg, M. Modeer, P.E. Peterson, Analysis of crack formation and crack
reinforced cementitious composites, Cem. Concr. Compos. 30 (2008) 917–928.
growth by means of fracture mechanics and finite elements, Cem. Concr. Res. 6
[9] W. Pansuk, T.N. Nguyen, Y. Sato, et al., Shear capacity of high performance fiber
(6) (1976) 773–782.
reinforced concrete I-beams, Constr. Build. Mater. 157 (2017) 182–193.
[42] A. Enfedaque, M.G. Alberti, J.C. Gálvez, et al., Numerical simulation of the
[10] P.N. Balaguru, S.P. Shah, Fiber Reinforced Cement Composites, McGraw-Hill
fracture behaviour of glass fiber reinforced cement, Constr. Build. Mater. 136
Inc, New York, 1992.
(2017) 108–117.
[11] A.M. Brandt, Fiber reinforced cement-based (FRC) composites after over 40
[43] Z.P. Bazant, B.H. Oh, Crack band theory for fracture of concrete, Mater. Struct.
years of development in building and civil engineering, Compos. Struct. 86
16 (3) (1983) 155–166.
(2008) 3–9.
[44] J. Lubliner, J. Oliver, S. Oller, E. Onate, A plastic-damage model for concrete, J.
[12] P.J. Heffernan, M.A. Erki, Fatigue behavior of reinforced concrete beams
Solids Struct. 25 (1989) 299–326.
strengthened with carbon fiber reinforced plastic laminates, J. Compos. Constr.
[45] P. Cao, D. Feng, C. Zhou, A modified damage-plasticity coupled area-weighted
8 (2) (2004) 132–140.
nonlocal model for simulating ductile fracture and softening behaviors of
[13] M.B.S. Alferjanil et al., Use of Carbon Fiber Reinforced Polymer Laminate for
materials, Math. Prob. Eng. 2014 (2014) 10.
strengthening reinforced concrete beams in shear: a review, Int. Refereed J.
[46] G. Barluenga, Fiber–matrix interaction at early ages of concrete with short
Eng. Sci. 2 (2) (2013) 45–53.
fibers, Cem. Concr. Res. 40 (5) (2010) 802–809.
[14] C.X. Qian, P. Stroeven, Development of hybrid polypropylene-steel fiber
[47] A. Chira, A. Kumar, T. Vlach, et al., Property improvements of alkali resistant
reinforced concrete, Cem. Concr. Res. 30 (1) (2000) 63–69.
glass fibers/epoxy composite with nanosilica for textile reinforced concrete
[15] F. Altun, T. Haktanir, K. Ari, Effects of steel fiber addition on mechanical
applications, Mater. Des. 89 (2015) 146–155.
properties of concrete and RC beams, Constr. Build. Mater. 21 (3) (2007) 654–
[48] C. Peng, J. Feng, C. Zhou, et al., Investigation on statistical characteristics of
661.
asphalt concrete dynamic moduli with random aggregate distribution model,
[16] P. Song, S. Hwang, Mechanical properties of high-strength steel fiber
Constr. Build. Mater. 148 (2017) 723–733.
reinforced concrete, Constr. Build. Mater. 18 (9) (2004) 669–673.
[49] P. Cao, F. Jin, C. Zhou, et al., Steady-state dynamic method: an efficient and
[17] B.E. Barragan, R. Gettu, M.A. Martin, R.L. Zerbino, Uniaxial tension test for steel
effective way to predict dynamic modulus of asphalt concrete, Constr. Build.
fiber reinforced concrete – a parametric study, Cem. Concr. Compos. 25 (7)
Mater. 111 (2016) 54–62.
(2003) 767–777.
[50] P. Cao, F. Jin, D. Feng, et al., Prediction on dynamic modulus of asphalt concrete
[18] Y. Su, J. Li, C. Wu, et al., Mesoscale study of steel fiber-reinforced ultra-high
with random aggregate modeling methods and virtual physics engine, Constr.
performance concrete under static and dynamic loads, Mater. Des. 116 (2017)
Build. Mater. 125 (2016) 987–997.
340–351.
72 X. Sun et al. / Construction and Building Materials 202 (2019) 58–72

[51] Z. Xu, H. Hao, H.N. Li, Mesoscale modelling of dynamic tensile behaviour of fiber [57] CECS13:2009, Standard for test methods of fibrous concrete. [S].Beijing: China
reinforced concrete with spiral fibers, Cem. Concr. Res. 42 (2012) 1475–1493. Planning Press, 2010:35-38.
[52] Z. Xu, H. Hao, H.N. Li, Mesoscale modelling of fiber reinforced concrete material [58] A.F. Stock, D.J. Hannant, R.I.T. Williams, The effect of aggregate concentration
under compressive impact loading, Constr. Build. Mater. 26 (1) (2012) 274–288. upon the strength and modulus of elasticity of concrete, Mag. Concr. Res.
[53] Xiangwei Liang, Wu Chengqing, Meso-scale modelling of steel fiber reinforced (1979).
concrete with high strength, Constr. Build. Mater. 165 (2018) 187–198. [59] K.C. Warren, R.A. Lopez-Anido, S.S. Vel, et al., Progressive failure analysis of
[54] A. Qsymah, R. Sharma, Z. Yang, et al., Micro X-ray computed tomography three-dimensional woven carbon composites in single-bolt, double-shear
image-based two-scale homogenisation of ultra high performance fiber bearing, Compos. B Eng. 84 (2016) 66–276.
reinforced concrete, Constr. Build. Mater. 130 (2017) 30–240. [60] Zhen Gao, Peng Cao, Xinjian Sun, Yawei Zhao, Compressive strength analysis
[55] A.E.K. Mike, Property analysis of Basalt Fiber Fast Setting and Hardening and microscopic characterization of basalt fiber reinforced concrete, J.
Cement Reinforced Concrete, Tianjin University, 2014. Hydroelectr. Eng. 37 (08) (2018) 111–120.
[56] The Ministry of Water Resources of the People’s Republic of China. SL352- [61] ZdenekP Bazant, Fracture and Size Effect in Concrete and Other Quasibrittle
2006, Test code for hydraulic concrete[S]. Materials, CRC Press, 1998.