Zhang 2021
Zhang 2021
Zhang 2021
h i g h l i g h t s
X-ray CT tests and IDBC method were combined to analyze the concrete damage process.
The failure modes of concrete under tensile and compression load were compared.
The differences of concrete under tensile and compression loads were studied.
The relationship between fractal dimension, CT number and stress were studied.
a r t i c l e i n f o a b s t r a c t
Article history: Concrete is prone to be damaged due to external forces. Traditional concrete failure analysis and research
Received 20 August 2020 mostly focus on the characterization of macro – mechanical properties, which cannot describe the dam-
Received in revised form 22 January 2021 age of internal structure and the process of concrete failure after being subjected to a load on a meso-
Accepted 11 February 2021
scopic level. Thus, a series of CT tests were carried out on the damage process of concrete under
Available online 11 March 2021
tensile and compression loads, and CT images of the whole process of concrete failure were obtained.
On this basis, the fractal dimension calculated by the Improved Differential Box Counting (IDBC) method
Keywords:
was used to quantitatively study the change of the internal structure of concrete during the loading pro-
Concrete
CT test
cess. Then the mean values of the CT number of concrete CT images were statistically analyzed. The
Tensile load results revealed that the fractal dimension of different stress stages in each section increased with the
Compression load increase of scan times under tensile load, while the fractal dimension increasing first, then decreasing
Fractal dimension and finally increasing with the increase of scan times under compression load. Correspondingly, with
CT number the increase of scan times, the mean values of CT number under tensile load showed a trend of decreas-
Damage process ing, while a trend of decreasing first, then increasing and finally decreasing of the mean values of CT num-
ber was exhibited when the concrete specimen was subjected to compression load, which indicated that
the concrete specimen undergone the process of expansion, crack propagation, and failure under tensile
load, whereas the concrete experienced the process of compaction, crack propagation, and failure under a
compression load. Furthermore, research results have shown that the fractal dimension and mean value
of CT number can be used to quantitatively describe the damage process of concrete, whether it is sub-
jected to tensile or compression load.
Ó 2021 Published by Elsevier Ltd.
https://doi.org/10.1016/j.conbuildmat.2021.122693
0950-0618/Ó 2021 Published by Elsevier Ltd.
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
crete failure. Liu et al. [9] and Guo et al. [10] utilized Hopkinson the concrete damage process and calculated the fractal dimension
tensile bar (SHTB) to conducted dynamic tensile tests on concrete, of the CT scan section, which found that the fractal dimension
and Cao et al. [11] investigated the dynamic response of cement- develops regularly with crack propagation and can better reflect
tailings matrix composites under SHPB compression load. Yan the evolution law of damage cracks of concrete materials.
et al. [12] did a dynamic tensile test of concrete and concluded that In this study, CT technique and fractal theory are coupled to
the results obtained from the traditional experimental analysis study the concrete damage process when the concrete is subjected
could not fully reveal the inherent variation of dynamic character- to uniaxial tensile and compression loads. The self-developed por-
istics, and had unavoidable limitations. From the above explana- table loading equipment is used to perform uniaxial tensile and
tion, in the traditional research of concrete materials, the damage compression tests on concrete specimens. Through CT real-time
characteristics of concrete are mostly studied from the macro- scanning, the CT images of the whole process of concrete damage
scopic point of view, which fails to reflect the macroscopic under two loading types are obtained, and the three-dimensional
mechanical properties of concrete from the internal destruction reconstruction of concrete specimen is carried out by using the
mechanism of concrete, and also cannot quantitatively analyze obtained CT images. Then, the self-editing MATLAB program is
the evolution process of cracks in concrete. used to calculate the fractal dimension of the concrete CT images
Due to the micro-crack damage is a prime cause for deteriora- with different cross-sections and different stress stages. Finally,
tion of concrete structures [13], in recent years, some non- the comparative study on fractal characteristics of concrete failure
destructive testing methods have been used to detect the damage under uniaxial tensile and compression loads is carried out
of concrete materials. And there are two main non-destructive through CT test results and fractal theory.
testing methods: acoustic emission method and CT testing method.
Zeng et al. [14] obtained both structural response and acoustic 2. Materials and methods
emission monitoring results during a four-point flexural test of
an I-section prestressed concrete beam. The change in AE signal 2.1. Materials and specimen preparation
activity during incremental loading cycles was evaluated to deter-
mine the beam’s response in terms of damage generation. Besides, The specification of the concrete cylinder specimen is
the AE technique for detecting different types of damage in the cor- U60 mm 120 mm. The age of the concrete specimen is 300 days,
rosion process was described through the characterization of cor- and the aggregate gradation is the first grade. The mix proportions
responding AE sources in pre-stressed concrete cylinder pipes of concrete are shown in Table 1.
[15]. Ren et al. [16] used AE data to study the fracture process of When making concrete specimens, blocks with a pouring size of
mortar and concrete beams under three-point bending tests. 150 mm 150 mm 450 mm are generally used. After curing at
Reboul et al. [17] employed the AE technique to investigate the room temperature, the concrete specimens are directly drilled on
damage levels and mechanisms of two reinforced concrete wall- these blocks firstly. Then cut the ends to get the required length
slab connections under a cyclic reverse loading. However, the AE of the test specimen, which is 120 mm. During the preparation of
technique has some disadvantages, such as the difficulty in estab- the test specimen, the problems of specimen alignment, adhesion,
lishing a quantitative relationship between the parameters of con- and installation also need to be solved.
crete cracks and frequency and seismic amplitude, as well as the
uncertainty of the research on the evolution process of concrete
2.1.1. Specimen alignment
cracks [18], therefore, CT technique was used to detect the con-
To ensure that the test specimen and the upper and lower con-
crete damage. Tian et al. [19] used X-ray computed tomography
necting rods maintain the coaxial in the process of bonding the test
(CT) to study the concrete damage characteristics under uniaxial
specimen, a V-shaped groove is designed. The process of specimen
compression. The X-CT technique and gray-level co-occurrence
alignment is shown schematically in Fig. 1.
matrices (GLCM) were coupled to analyze the internal damage of
concrete [20,21]. Whereas, neither the acoustic emission technique
nor the CT scanning technique can quantitatively describe the con- 2.1.2. Specimen adhesion
crete damage process, thus scholars have proposed various meth- A type of building structural adhesive developed and produced
ods to quantitatively describe the concrete damage. by the Dalian Institute of Physics and Chemical Engineering is
For quantitatively describing the process of concrete damage, a selected to bond the specimen, commonly known as sticky steel
concrete cube finite element model with an embedded SA sensor adhesive. The performance indexes of the building structural adhe-
was established using a concrete damaged plasticity constitutive sive are listed in Table 2. In the process of bonding the specimen,
model to simulate the damage process under uniaxial compression the V-shaped groove is placed vertically to bond the specimen
[22]. Song et al. [23] proposed a three-dimensional multi-level for reducing the error of the test curve caused by the deformation
stress corrosion model (MSC) to characterize the damage induced of the adhesive.
by the variable cyclic load levels during consecutive loading stages
in numerical simulations. However, as a commonly used engineer- 2.1.3. Specimen installation
ing material, concrete exhibits a series of fractal features during its During specimen installation, the test machine is placed verti-
formation [24]. Hence, more and more scholars have used fractal cally. Meanwhile, the operator holds the connecting rod with his
theory to study the concrete damage process. Rezaie et al. [25] hand and puts it into the pressure chamber while releasing the test
used fractal dimensions to quantitate the cracking pattern on walls specimen, which effectively reduces the probability of the test
after earthquakes. Besides, using the fractal dimension as a math- specimen fracture during installation.
ematical tool, Zamen et al. [26] conducted damage assessment on
concrete materials under different applied loads after being sub- 2.2. Test devices
jected to extreme compression loads. Moreover, the correlations
among the fractal dimension, compressive strength, and perme- To measure the damage process of concrete specimens under
ability of concrete incorporating silica fume were studied, and uniaxial tensile and compression loads, the portable CT real-time
the results showed that increasing silica fume content improves dynamic loading device independently developed by the Xi’an
compressive strength but decreases permeability and fractal University of Technology is used with an X-ray CT scanning device
dimension [27]. Dang et al. [28,29] carried out the CT tests on [30]. In this study, the combination of loading device and the X-ray
2
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Table 1
The proportion of concrete (kg/m3).
V-shaped groove
Table 2
Performance index of building structural adhesive.
2.3. Methodology
End cover
Table 3
The load and corresponding displacement under tensile load.
Fig. 5. The fractured image of CONC-042. Fig. 6. The fractured image of CONC-061.
Table 4
The load and corresponding displacement under compression load.
5
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Fig. 7. Scanning CT images in each stress stage of 4 cross-sections in uniaxial tensile test: (a) 1st cross-section; (b) 2nd cross-section; (c) 3rd cross-section; (d) 4th cross-
section.
and 16(d) are the three-dimensional images of the specimen after 4.4. Segmentation results of CT images using IDBC method
failure.
It can be seen from Figs. 15 and 16 that the concrete specimens Before using MATLAB software to calculate the fractal dimen-
show heterogeneity. When the concrete specimen is damaged by a sion of CT images, it is processed by binarization first to provide
tensile load, the crack is roughly perpendicular to the loading more accurate crack morphology. As shown in Figs. 17 and 18,
direction, and after being damaged by a compression load, the the areas of aggregate and mortar are black, and the areas of pores
crack appears parallel to the loading direction. and cracks are white.
Fig. 8. Concrete scanning longitudinal section under tensile load: (a) 1st scan; (b) 3rd scan; (c) 6th scan.
6
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Fig. 10. Scanning CT images in each stress stage of 4 cross-sections in uniaxial compression test: (a) 1st cross-section; (b) 2nd cross-section; (c) 3rd cross-section; (d) 4th
cross-section.
7
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Fig. 11. Concrete scanning longitudinal section under uniaxial compression load: (a) 1st scan; (b) 3rd scan; (c) 5th scan.
8
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Fig. 14. Threshold segmentation by using Line-Probe method: (a) the result of threshold segmentation; (b) change of CTN along the straight line.
Fig. 15. Three-dimensional reconstruction of the concrete specimen under tensile load: (a) three-dimensional reconstruction of the complete specimen before loading;
(b) corresponding aggregate; (c) corresponding mortar; (d) three-dimensional reconstruction of the complete specimen after fracture.
Fig. 16. Three-dimensional reconstruction of the concrete specimen under compression load: (a) three-dimensional reconstruction of the complete specimen before loading;
(b) corresponding aggregate; (c) corresponding mortar; (d) three-dimensional reconstruction of the complete specimen after fracture.
The variation of fractal dimension with the stress of four cross- tion of fractal dimension fully reflects the brittle fracture charac-
sections is shown in Fig. 19. teristics of concrete [29].
It can be seen from Fig. 19 that when the concrete specimen is Fig. 20 shows the mean values of the CT number which belong
subjected to uniaxial tensile load, the fractal dimension D shows a to the four CT scanning sections, and all mean values of the CT
continuously increasing trend. Taking cross-section 1 as an exam- number are calculated by ENVI image processing software. As
ple, the fractal dimension D = 2.2758 at the first scan, however, can be seen from Fig. 20, the variation trend of the mean values
there is no significant change on the fractal dimension D before of CT number in the four cross-sections is consistent, but the mean
the sixth scan with the increase of stress. The fractal dimension values of CT numbers do not change significantly, which is because
D of the fifth scan is 2.2763, which is 0.022% higher than that of the tensile damage process of the concrete specimen is very short,
the first scan, which is because the stress level of the previous and it is difficult to capture the CT image when the concrete is frac-
scans is relatively low and the damage development of concrete tured. There are almost no cracks in the cross-sectional view of the
specimen is slow. At the sixth scan, the fractal dimension CT scan, therefore the mean values of the CT number do not change
D = 2.2865, which is an increase of 0.0120 compared to the fifth obviously. However, from the perspective of the mean values of the
scan. This change is caused by the concrete specimen is completely CT number in the longitudinal section, it can be seen that with the
fractured after the peak stress. The roughness of concrete reaches a scan times (stress) increase, the mean values of the CT number
maximum, thus the fractal dimension increases greatly. The varia- basically decrease linearly before the sixth scan. At the sixth scan,
9
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Fig. 17. Segmentation images in each stress stage of 4 cross – sections in tensile test: (a) 1st cross-section; (b) 2nd cross-section; (c) 3rd cross-section; (d) 4th cross-section.
the mean values of the CT number in this scan stage are signifi- increases gradually. After the peak stress, the specimen has been
cantly reduced. Tian et al. [35] and Lei et al. [36] both studied fractured and the roughness of the specimen is relatively large,
the mesoscopic damage process of concrete under tensile load which leads to the fractal dimension reaches a maximum.
and the same conclusions were reached. Fig. 22 shows the mean values of the CT number in the cross
and longitudinal sections of the CT scan after the concrete speci-
5.2. Calculation of fractal dimension and CT numbers of concrete CT men is subjected to a compression load. As can be seen from this
images under uniaxial compression load figure, the change law of the mean values of the CT number in
the cross and longitudinal sections is the same. When the stress
The curves of the fractal dimension with the stress at each is in the range of 0–10.6 MPa, the mean values of the CT number
stress stage in the four cross-sections of the concrete specimen in the cross and longitudinal sections are decreased, but the
under the compression load are acquired as shown in Fig. 21. decrease range is not large. During the subsequent stress growth
From Fig. 21, it can be seen that the change of fractal dimension of 10.6–25.8 MPa, the mean values of CT number in the cross
with the stress under compression load is different from that under and longitudinal sections increased, indicating that the specimen
tensile load. In the first two scanning stages, the fractal dimensions has experienced a short compaction stage, the voids and micro-
of the four cross-sections all increase slightly with the increase of cracks in the specimen are closed, which is reflected in the CT
stress. With the increase of stress, the fractal dimension D image that the mean values of CT number increase significantly.
decreases a little bit at the third scan, which is because the con- In the next stress in the range of 25.8–29.9 MPa, the mean values
crete specimen has experienced a compaction stage, that is, the ini- of CT number in the cross and longitudinal sections are reduced,
tial voids and micro-cracks in the specimen are closed due to the and the reduction is large, which illustrates that the cracks start
external load. After the third scan, the fractal dimension D to penetrate and expand at this stage. When the stress exceeds
10
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
Fig. 18. Segmentation images in each stress stage of 4 cross – sections in compression test: (a) 1st cross-section; (b) 2nd cross-section; (c) 3rd cross-section; (d) 4th cross-
section.
Fig. 19. The relationship between fractal dimension and stress of four cross- Fig. 20. Variation of the mean CT number in the cross and longitudinal sections
sections at different stress stages under tension load. under tensile load.
11
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
6. Conclusions
12
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
to the lowest. Nevertheless, the mean values of the CT num- aggregate and aluminate cement, Constr. Build. Mater. 147 (2017) 865–875,
https://doi.org/10.1016/j.conbuildmat.2017.04.138.
ber show a trend of first decreasing, then increasing, and
[9] P.F. Liu, X.P. Zhou, Q.H. Qian, et al., Dynamic splitting tensile properties of
finally decreasing as the scan times increase under uniaxial concrete and cement mortar, Fatigue Fract. Eng. Mater. Struct. 43 (4) (2020)
compression load. 757–770, https://doi.org/10.1111/ffe.13162.
(4) Through the quantitative analysis of fractal dimension and [10] R.Q. Guo, H.Q. Ren, L. Zhang, et al., Direct dynamic tensile study of concrete
materials based on mesoscale model, Int. J. Impact Eng. 143 (2020) 103598,
mean values of CT number, it is concluded that the specimen https://doi.org/10.1016/j.ijimpeng.2020.103598.
undergoes the process of compaction, expansion, crack [11] S. Cao, E. Yilmaz, W.D. Song, Dynamic response of cement-tailings matrix
propagation, and failure under compression load, and only composites under SHPB compression load, Constr. Build. Mater. 186 (2018)
892–903, https://doi.org/10.1016/j.conbuildmat.2018.08.009.
expansion, crack propagation, and failure processes are [12] D.M. Yan, G. Lin, Dynamic tensile experiments of concrete and its data analysis
observed under tensile load. The changes of fractal dimen- method, J. Harbin Inst. Technol. 38 (4) (2006) 622–643. https://doi.org/0367-
sion and mean value of CT number can better reflect the 6234(2006)04-0622-04.
[13] A. Sidiq, R.J. Gravina, S. Setunge, F. Giustozzi, Microstructural analysis of
damage process of concrete materials. These two variables healing efficiency in highly durable concrete, Constr. Build. Mater. 215 (2019)
can be used as quantitative damage state parameters to 969–983, https://doi.org/10.1016/j.conbuildmat.2019.04.233.
describe the evolution process of concrete under uniaxial [14] H. Zeng, J.A. Hartell, M. Soliman, Damage evaluation of prestressed beams
under cyclic loading based on acoustic emission monitoring, Constr. Build.
tensile and compression loads. Mater. 255 (2020) 119235, https://doi.org/10.1016/
j.conbuildmat.2020.119235.
CRediT authorship contribution statement [15] R. Goldaran, A. Turer, Application of acoustic emission for damage
classification and assessment of corrosion in pre-stressed concrete pipes,
Measurement 160 (2020) 107855, https://doi.org/10.1016/j.
Le Zhang: Investigation, Conceptualization, Methodology, Soft- measurement.2020.107855.
ware, Writing-original draft. Faning Dang: Conceptualization. [16] D. Ren, B.G. Liu, S.J. Chen, D.W. Yin, M.Y. Yu, H. Liu, L. Wu, Visualization of
acoustic emission monitoring of fracture process zone evolution of mortar and
Weihua Ding: Data curation, Resources. Lin Zhu: Editing, concrete beams under three-point bending, Constr. Build. Mater. 249 (2020)
Investigation. 118712, https://doi.org/10.1016/j.conbuildmat.2020.118712.
[17] N. Reboul, C. Grazide, N. Roy, et al., Acoustic emission monitoring of reinforced
concrete wall-slab connections, Constr. Build. Mater. 259 (2020) 119661,
Declaration of Competing Interest https://doi.org/10.1016/j.conbuildmat.2020.119661.
[18] J. Zhang, Study on Concrete Dynamic Characteristic of DAGANGSHAN Arch
The authors declare that they have no known competing finan- Dam by CT Test, Xi’an University of Technology, 2010.
[19] W. Tian, N. Han, Analysis on meso-damage processes in concrete by X-ray
cial interests or personal relationships that could have appeared computed tomographic scanning techniques based on divisional zones,
to influence the work reported in this paper. Measurement 140 (2019) 382–387, https://doi.org/10.1016/j.
measurement.2019.04.026.
[20] P.H. Wang, H.X. Qiao, Y.S. Zhang, Y.K. Li, Q. Feng, K.F. Chen, Meso-damage
Acknowledgments evolution analysis of magnesium oxychloride cement concrete based on X-CT
and grey-level co-occurrence matrix, Constr. Build. Mater. 255 (2020) 119373,
This study is financially sponsored by the National Natural https://doi.org/10.1016/j.conbuildmat.2020.119373.
[21] L. Zhu, F.N. Dang, Y. Xue, W.H. Ding, L. Zhang, Analysis of micro-structural
Science Foundation of China (Nos. 51679199, 51979225), the Spe- damage evolution of concrete through coupled X-ray computed tomography
cial Funds for Public Industry Research Projects of the Ministry of and gray-level co-occurrence matrices method, Constr. Build. Mater. 224
Water Resources (No. 201501034-04) and the Key Laboratory for (2019) 534–550, https://doi.org/10.1016/j.conbuildmat.2019.07.007.
[22] S.Y. Zhao, S.L. Fan, J. Yang, S. Kitipornchai, Numerical and experimental
Science and Technology Coordination & Innovation Projects of investigation of electro-mechanical impedance based concrete quantitative
Shaanxi Province (No. 2014SZS15-Z01). The authors would like to damage assessment, Smart Mater. Struct. 29 (5) (2020) 055025, https://doi.
thank the National Natural Science Foundation of China, Special org/10.1088/1361-665X/ab58e9.
[23] Z.Y. Song, H. Konietzky, M. Herbst, Three-dimensional particle model based
Funds for Public Industry Research Projects of the Ministry of
numerical simulation on multi-level compressive cyclic loading of concrete,
Water Resources and Key Laboratory for Science and Technology Constr. Build. Mater. 225 (2019) 661–677, https://doi.org/10.1016/
Coordination & Innovation Projects of Shaanxi Province for the j.conbuildmat.2019.07.260.
[24] V.E. Saouma, C.C. Barton, Fractals, fracture and size effect in concrete, J. Eng.
financial support for this project.
Mech. 120 (4) (1994) 835–854, https://doi.org/10.1061/(ASCE)0733-9399
(1994)120:4(835).
References [25] A. Rezaie, A.J.P. Mauron, K. Beyer, Sensitivity analysis of fractal dimensions of
crack maps on concrete and masonry walls, Autom. Constr. 117 (2020)
103258, https://doi.org/10.1016/j.autcon.2020.103258.
[1] G.L. Xue, E. Yilmaz, W.D. Song, S. Cao, Analysis of internal structure behavior of
[26] S. Zamen, E.D. Niri, Fractal analysis of nonlinear ultrasonic waves in phase-
fiber reinforced cement-tailings matrix composites through X-ray computed
space domain as a quantitative method for damage assessment of concrete
tomography, Compos. Part B: Eng. 175 (2019) 107091, https://doi.org/10.1016/
structures, NDT&E Int. 111 (2020) 102235, https://doi.org/10.1016/j.
j.compositesb:2019.107091.
ndteint.2020.102235.
[2] D. Li, D.T. Niu, Q. Fu, D.M. Luo, Fractal characteristics of pore structure of hybrid
[27] Q. Lü, Q.L. Qiu, J. Zheng, J.Y. Wang, Q. Zeng, Fractal dimension of concrete
basalt–polypropylene fibre-reinforced concrete, Cem. Concr. Compos. 109
incorporating silica fume and its correlations to pore structure, strength and
(2020), https://doi.org/10.1016/j.cemconcomp.2020.103555. UNSP 103555.
permeability, Constr. Build. Mater. 228 (2019) 116986, https://doi.org/
[3] P. Lin, Brittle failure behaviour of medium containing flaws and pores, Chin. J.
10.1016/j.conbuildmat.2019.116986.
Rock Mech. Eng. 21 (8) (2002), https://doi.org/10.3321/j.issn:1000-
[28] F.N. Dang, G.Y. Lei, W.H. Ding, H.F. Ma, H.Q. Chen, Study on the CT meso-test
6915.2002.08.037. 1283–1283.
experiment of static and dynamic failure processes of concrete, J. Hydroelectr.
[4] O. Galao, F.J. Baeza, E. Zornoza, P. Garcés, Carbon nanofiber cement sensors to
Eng. 34 (1) (2015) 189–196. https://doi.org/1003-1243(2015)34:1<189:
detect strain and damage of concrete specimens under compression,
HNTJDL>2.0.TX;2-W.
Nanomaterials 7 (12) (2017) 413, https://doi.org/10.3390/nano7120413.
[29] F.N. Dang, J.Y. Fang, W.H. Ding, Fractal comparison research of fracture of
[5] A.M. Rajabi, F.O. Moaf, H.S. Abdelgader, Evaluation of mechanical properties of
concrete samples under static and dynamic uniaxial tensile using CT, Chin. J.
two-stage concrete and conventional concrete using nondestructive tests, J.
Rock Mech. Eng. 34 (S1) (2015) 2922–2928, https://doi.org/10.13722/j.cnki.
Mater. Civ. Eng. 2 (7) (2020) 04020185, https://doi.org/10.1061/(ASCE)
jrme.2014.0561.
MT.1943-5533.0003247.
[30] W. Tian, F.N. Dang, W.H. Ding, X.Y. Liang, H.Q. Chen, Development of a
[6] T. Lee, J. Lee, Setting time and compressive strength prediction model of
dynamic loading apparatus for CT experiment and its application, Rock
concrete by nondestructive ultrasonic pulse velocity testing at early age,
Soil Mech. 31 (1) (2010) 309–313, https://doi.org/10.16285/j.
Constr. Build. Mater. 252 (2020) 119027, https://doi.org/10.1016/
rsm.2010.01.048.
j.conbuildmat.2020.119027.
[31] S.S. Jin, J.X. Zhang, B.S. Huang, Fractal analysis of effect of air void on freeze–
[7] K. Horníková, L. Scheinherrová, R. Štefan, M. Foglar, Experimental investigation
thaw resistance of concrete, Constr. Build. Mater. 47 (2013) 126–130, https://
of physical, thermal, hygral and mechanical properties of cementitious
doi.org/10.1016/j.conbuildmat.2013.04.040.
composites at high temperatures, Constr. Build. Mater. 255 (2020) 119244,
[32] S.S. Jin, G.P. Zheng, J. Yu, A micro freeze-thaw damage model of concrete with
https://doi.org/10.1016/j.conbuildmat.2020.119244.
fractal dimension, Constr. Build. Mater. 257 (2020) 119434-1–119434-8,
[8] A. Baradaran-Nasiri, M. Nematzadeh, The effect of elevated temperatures on
https://doi.org/10.1016/j.conbuildmat.2020.119434.
the mechanical properties of concrete with fine recycled refractory brick
13
L. Zhang, F. Dang, W. Ding et al. Construction and Building Materials 285 (2021) 122693
[33] J. Li, Q. Du, C.X. Sun, An improved box-counting method for image fractal [35] W. Tian, F.N. Dang, Y.L. Xie, CT test analysis of meso damage and fracture
dimension estimation, Pattern Recogn. 42 (11) (2009) 2460–2469, https://doi. process of concrete under tensile loading, J. Civ. Archit. Environ. Eng. 37 (2)
org/10.1016/j.patcog.2009.03.001. (2015) 73–78. https://doi.org/:1674-4764(2015)02-0073-06.
[34] L. Zhang, F.N. Dang, W.H. Ding, L. Zhu, Quantitative study of meso-damage [36] G.Y. Lei, H.Y. Wang, J.H. Sun, F.N. Dang, Study on the failure mechanism of
process on concrete by CT technology and improved differential box counting concrete under compressive and tensile loads based on CT testing, Build.
method, Measurement 160 (2020) 107832, https://doi.org/10.1016/j. Struct. 48 (Supp. 2) (2018) 644–650, https://doi.org/10.19701/j.jzjg.2018.
measurement.2020.107832. S2.130.
14