Case Studies in Construction Materials 17 (2022) e01544

Contents lists available at ScienceDirect

Case Studies in Construction Materials

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

Influences of short-term heavy rainfall on interlayer properties

and microstructure of concrete dam
Wei-Jia Liu a, Yao-Sheng Tan b, c, Qing-Bin Li a, Ning Yang b, c, Chun-Feng Liu b, c,
Xiao-Feng Gao a, d, Yu Hu a, *, Xu-Jing Niu a, e, **, Bo Zhang f
a
State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, China
b
China Three Gorges Group Corporation, Beijing 100038
c
China Three Gorges Construction Engineering Corporation, Chengdu 610041, China
d
College of Civil Engineering, Zhejiang University of Technology, Hangzhou 310023, China
e
School of Mechanics and Civil Engineering, China University of Mining and Technology (Beijing), Beijing 100083, China
f
Emergency Science Research Institute, Chinese Institute of Coal Science, Beijing 100013, China

A R T I C L E I N F O A B S T R A C T

Keywords: Rainwater adversely affects the interlayer bonding quality of layered concrete. This study tested
Concrete dam the influences of different treatment measures (i.e., discharging the rainwater accumulated on the
Short-term heavy rainfall layer and covering the insulation quilt) on the interlayer mechanical properties of concrete under
Interlayer properties
the condition of short-term heavy rainfall, including compressive strength (CS), splitting tensile
Ultrasonic non-destructive test
Damage degree
strength (STS), and shear strength (SS). Meanwhile, the influence of short-term heavy rainfall on
the interlayer microstructure of concrete was analyzed using a scanning electron microscope
(SEM) and mercury intrusion porosimetry (MIP). In addition, the ultrasonic non-destructive
testing method was employed to quantify the deterioration degree of the concrete layer under
the condition of short-term heavy rainfall. The results show that short-term heavy rainfall can
reduce the CS, STS, and SS of concrete by 50%, 70%, and 37%, respectively. Meanwhile, the
porosity and damage degree near the concrete interlayer were also increased by short-term heavy
rainfall. This is mainly because rainfall increases the water-binder ratio (w/b) of concrete,
resulting in the deterioration of concrete microstructure. The increase of w/b is not conducive to
the development of concrete mechanical properties. Concrete with a large w/b will retain more
water after hydration. The water not involved in hydration leaves more pores after evaporation,
increasing the total porosity of concrete. Moreover, there is a linear relationship between the
concrete damage degree and the interlayer STS and SS loss rate, as well as the correlation co­
efficients are 0.959 and 0.983, respectively. The porosity of concrete is linearly correlated with
the interlayer STS and SS, and the correlation coefficients are 0.973 and 0.909, respectively.
Draining the rainwater accumulated on the layer and covering the insulation quilt can improve
the interlayer CS, STS, and SS of concrete. This is mainly because the above measures reduce the
influence of rainwater on the interlayer.

* Corresponding author.
** Corresponding author at: State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, China.
E-mail addresses: [email protected] (Y. Hu), [email protected] (X.-J. Niu).

https://doi.org/10.1016/j.cscm.2022.e01544
Received 15 July 2022; Received in revised form 18 September 2022; Accepted 6 October 2022
Available online 12 October 2022
2214-5095/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

1. Introduction

China builds cascade hydropower facilities in the Jinsha River Basin (Fig. 1) to cope with climate change and achieve the goal of
carbon neutrality by 2060. These power generation facilities mainly include Xiluodu extra-high arch dam (285.5 m), Baihetan extra-
high arch dam (289 m), and Wudongde extra-high arch dam (270 m), etc. [1]. Most of these arch dams are located in river valleys with
harsh climatic conditions. The region is prone to severe weather such as strong winds, high temperatures and heavy rainfall. Previous
studies have shown that concrete pouring in a harsh environment will degrade the interlayer properties. Furthermore, the degradation
of interlayer properties will adversely affect the safety and anti-sliding stability of the dam [2–4].
At present, the researches on the effect of environmental factors on the interlayer properties of concrete mainly focuses on tem­
perature, humidity and wind speed. Niu et al. [5] analyzed the influences of environmental factors (i.e., temperature, humidity and
wind speed) on the interlayer STS and chloride ion penetration resistance of concrete. The results showed that the interlayer properties
of concrete decreased significantly with the increase of temperature and wind speed. This is mainly because the higher temperature
and wind speed accelerate the hardening process of the lower concrete and make the pore structure rough. In addition, high tem­
perature led to uneven distribution of hydration products and induced microcracks in the structure. The research in literature [6]
showed that increasing relative humidity was beneficial to cement hydration. Ribeiro et al. [7] found that increasing the relative
humidity of the air can significantly increase the tensile strength of RCC. Xu et al. [8] analyzed the influences of different wind speeds
(i.e., 0, 2.5 and 5 m/s) on water content and interlayer STS of concrete. Their results showed that the dry and windy environment
would adversely affect the interlayer mechanical properties of concrete. This is mainly due to the rapid evaporation of concrete
moisture in the windy environment, which reduces the hydration degree of the lower concrete.
In addition, studies have shown that solar radiation can also adversely affect concrete properties. Feng et al. [9] found that the
temperature of concrete will rise due to long-term solar radiation. They suggested that the impact of solar radiation should be
minimized during construction. Moelich et al. [10] analyzed the influence of solar radiation on the plastic shrinkage cracking of
concrete. Their results showed that the plastic shrinkage and cracking degree of concrete in the sunlight were much greater than in the
shade. This is attributed to solar radiation accelerating the loss of pore water in concrete.
However, it is noteworthy that rainfall during construction is also a crucial factor affecting the properties of concrete. El-amary
et al. [11] studied the influences of rainfall duration (10, 30, 45 and 60 min) and rainfall start time (15, 35, 45 min from casting)

Fig. 1. Distribution of Jinsha River cascade hydropower stations.

2
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

on the CS of concrete. Their results showed that the CS of concrete decreased with the increase of rainfall duration. In addition, the
earlier the rainfall started, the more obvious the decrease in concrete CS. Lei et al. [12] studied the influences of rainfall intensity (0,
2.6, 5 and 8 mm / h) on the compactability, density and interlayer bonding strength of RCC. Their results showed that when the rainfall
intensity was less than 2.6 mm/h, the compaction, density and interlayer bonding quality of concrete were not affected. When the
rainfall intensity was greater than 8 mm / h, the shear strength and compaction quality of concrete significantly decreased. At this
time, the construction should be stopped.
The above literature review shows that the ambient temperature, relative humidity, wind speed, solar radiation and rainfall in the
concrete pouring process all affect the interlayer properties of concrete. However, the current research on the properties of concrete by
rainfall mainly focuses on integrally poured specimens.There are few reports on the research of layered pouring concrete. In addition,
there is no in-depth discussion of concrete microstructure changes under rainfall conditions in the relevant literature. The mechanism
of interlayer properties deterioration of dam under strong rainfall is still unclear.
Therefore, this study evaluated the influence of strong rainfall on the interlayer bonding properties of dam from two aspects:
macro-mechanical properties and micro-structure. To reveal the mechanism of the deterioration of the interlayer bonding properties of
concrete under rainfall conditions. In addition, this study also evaluated the influences of two treatment measures, namely, discharging
the rainwater accumulated on the layer (Fig. 2a) and covering the insulation quilt (Fig. 2b), on the interlayer properties of concrete
under the conditions of short-term heavy rainfall. This study is based on the environmental conditions of the construction site of the
Wudongde ultra-high arch dam. Wudongde dam is located in the low-latitude plateau monsoon region. The combined action of various
weather systems and complex terrain on the plateau leads to frequent disastrous weather such as rainstorms and short-term heavy
rainfall. The average annual precipitation in Wudongde is 758 mm. The frequency of short-term heavy rainfall is 5.5 times / year [13].
Given the large scale and long construction period of extra-high arch dams, construction in the rain is inevitable (Fig. 2). Moreover,
with global warming, extreme rainfall weather will become more frequent [14,15]. The above research has reference significance for
constructing mass concrete in rainy areas.

2. Experimental procedures

2.1. Materials

Low-heat Portland (LHP) cement with a specific surface area of 319 m2/kg and a density of 3200 kg/m3 was used in this research.
The density and fineness of the fly ash used in this study were 1800 kg/m3 and 8.2%, respectively. The chemical and mineral com­
positions of cementitious material are listed in Table 1.
Table 2 shows the mixed proportion of concrete. Among them, the water to binder ratio (w/b) of concrete mixture is 0.48. Artificial
sand (basalt) with a fineness modulus of 2.76 and an apparent density of 2780 kg/m3 was used. The particle size of coarse aggregate
(crushed basalt) is 5–20 mm. The morphologies of cement and aggregates are illustrated in Fig. 3. In addition, polycarboxylate
superplasticizers and air-entraining agents were used in this study.

2.2. Specimen preparation

The test was divided into four groups, with 6 specimens in each group. The first group of specimens was poured with concrete as a
whole (Fig. 4a), and the other groups of specimens were poured in layers. For the second group of specimens, the lower layer concrete
was poured first. After that, a short-term heavy rainfall simulation was performed on the concrete layer (Fig. 4b, rainfall 10 mm/h,
rainfall duration 3 h), and finally, the upper layer concrete (height 75 mm) was poured. The third group was the experiment of
draining rainwater from the concrete surface. Based on the second group of test rainfall, the accumulated water on the concrete surface
was discharged (Fig. 5b). The mass of discharged materials (rainwater and cementitious materials) accounts for 20% of the total
rainfall, making the concrete surface free from visible rainwater. The fourth group was tested for covering the concrete surface with an
insulation quilt. Cover the insulation quilt while raining (Fig. 4c). The insulation quilt was made of thermal insulation cotton with a
thickness of about 20 mm. Then, they were placed in a standard curing room (20 ± 0.5 ℃, RH=98%) for 28d.
In this study, the rainfall process was simulated using a sprinkler rainfall device. The rainfall intensity was controlled by adjusting

Fig. 2. Pictures of dam construction under rainfall conditions: (a) discharge the rainwater accumulated on the concrete surface (b) cover the
concrete surface with thermal insulation quilt.

3
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Table 1
Properties of cement and fly ash.
Properties Cement Fly ash

Chemical composition（% by mass） SiO2 23.24 50.6

Al2O3 4.07 27.1
Fe2O3 3.02 7.1
CaO 61.64 4.5
MgO 4.74 1.2
SO3 2.05 0.3
Alkali content 0.45 1.3
Loss on ignition 1.08 4.4
Mineral composition (% by mass) C3S 32.4 Vitreous body (54%)
C2S 47.5 Quartz (26%)
C3A 0.2 Mullite (17%)
C4AF 14.6 Magnetite (1.8%)
Gypsum 3.5 Hematite (0.2%)
Physical properties Specific gravity (g/cm3) 3.20 1.8
Specific surface (m2/kg) 319 1390
Compressive strength (MPa) 7 days 21.3 /
28 days 48.4 /

Table 2
Mix proportion of concrete (kg/m3).
W/B Cement Fly ash Artificial sand Coarse aggregate Superplasticizer Air-entraining agent

0.48 100 123 777 1323 1.388 0.069

Fig. 3. Morphologies of (a) cement, (b) artificial sand and (c) coarse aggregates.

Fig. 4. Concrete layer process method: (a) integral pouring, (b) no treatment measures, (c) covering insulation quilt.

the nozzle size and water supply pressure. Meanwhile, a graduated beaker was used to detect the rainfall of each sprinkler.
The specimens were numbered C0, A1, A2 and A3, respectively. C0 represents integrally poured concrete. A1 indicates that
rainwater accumulated on concrete surfaces under heavy rainfall conditions is not discharged. A2 indicates that rainwater accumu­
lated on the concrete surface will be discharged after a short time of heavy rainfall. A3 indicates that the insulation quilt is covered on
the layer under heavy rainfall conditions (Table 3).

4
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Fig. 5. Concrete layer state after short-term heavy rainfall: (a) no treatment measures, (b) discharge rainwater that accumulated on the layer, (c)
covering insulation quilt.

Table 3
Grouping and layer treatment of concrete specimens.
Groups Layer processing method Specimen size

C0 Integral pouring 150 mm × 150 mm× 150 mm

A1 Layered pouring (no treatment measures)
A2 Layered pouring (discharge rainwater that accumulated on the layer)
A3 Layered pouring (covering insulation quilt)

2.3. Test procedures

2.3.1. CS and STS test

The CS and STS test of the cube specimens was carried out in the light of standard DL/T 5150–2017 [16]. For CS test, load
continuously at the speed of 0.3 ~ 0.5 MPa / s until the specimen was damaged (Fig. 6a). For STS test, the specimens were split along
the interlayer at loading rates of 0.05 MPa/s (Fig. 6b). The average value of the six samples was taken as the test result.

2.3.2. Interlayer SS test

YZW100 rock fracture direct shear instrument under multi-field coupling condition was used in the concrete SS test (Fig. 7). Put the
specimen into the shear box, and place the force transfer plate above the specimen. According to the designed normal maximum load,
the normal stress is applied in 5 grades (0.6, 1.2, 1.8, 2.4 and 3 MPa). The SS of each normal stress level was taken as the average value
of 5 specimens. The tangential load was controlled by displacement (0.01 mm/s) [16]. Calculate the normal stress and shear stress
under normal loads at all levels according to formulas (1) and (2):
Fσ
σi = (1)
A

Fτ
τi = (2)
A

Where, σi is the normal stress; τi is the shear stress; Fσ is the total normal load; Fτ is the shear load;A is the effective shear area of the
shear plane.

Fig. 6. Concrete mechanical properties test: (a) CS, (b) STS.

5
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Fig. 7. Concrete shear strength test.

Calculate the friction coefficient and cohesion according to formula (3):

(3)
′ ′
τ = σf + c

Where, τ is the ultimate shear strength; σ is the normal stress; f is the friction factor; c is the cohesion.
′ ′

2.3.3. Ultrasonic non-destructive test

Ultrasound can detect damage inside concrete [17]. This study measured the longitudinal wave velocity of concrete samples by
33522 A Ultrasonic test instrument [18]. When measuring the longitudinal wave velocity through the concrete layer, the ultrasonic
probe should be placed in the middle of the two sides of the concrete. The surface of ultrasonic probe and test piece shall be coated with
a coupling agent to make them closely fit (Fig. 8).

2.3.4. MIP
The MIP method was used to study the effect of short-term heavy rainfall on the pore structure of concrete. The test equipment was
a MicroActive AutoPore V 9600 mercury porosimeter. After curing for 28 days, drill cores in the center of concrete (Fig. 9). The test
samples were prepared into particles with a diameter of approximately 5 mm. The samples were soaked in isopropanol for 7 days to
stop the hydration reaction [19]. Prior to MIP test, the samples were put in a vacuum dish to remove moisture from the pore structure.

2.3.5. SEM
SEM analysis (Hitachi S-4800, Hitachi Limited, Japan) was used to characterize the microstructure of concrete. SEM analyzed the
samples of each group of specimens after curing for 28 days. Select the sample from the concrete center (Fig. 9). The sample was then
crushed to a size of about 5 mm. Before the experiment, the specimens were put in a vacuum drying dish for 3 days. Then the samples
were placed in an ion sputtering apparatus for gold spraying. The micro-morphological differences between the samples were observed
using an accelerating voltage of 20 kV and a magnification of 1000 × .

Fig. 8. An ultrasonic velocity test of concrete.

6
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Fig. 9. Drilling and coring of concrete specimens.

3. Results and discussion

3.1. Mechanical properties

3.1.1. CS and STS

Fig. 10 shows the effects of different treatment measures on the CS (Fig. 10a) and interlayer STS (Fig. 10b) of concrete under strong
rainfall condition. The CS and STS of layered concrete under short-term heavy rainfall conditions are lower than the strength of bulk
concrete (C0). The CS of C0 was 31.72 MPa. The CS of A1 (no treatment measures), A2 (discharge rainwater that accumulated on the
layer) and A3 (covering insulation quilt) were 15.94, 26.31 and 27.92 MPa, which were 50%, 17% and 12% lower than that of C0,
respectively. The STS of C0 was 1.40 MPa. The interlayer STS of A1, A2 and A3 were 0.42, 0.59 and 0.80 MPa, which were decreased
by 70%, 58% and 43% compared with C0, respectively. By comparison, it can be found that the STS of concrete decreases more than
the CS under short-term heavy rainfall. This is mainly related to the different sensitivities of different strengths to the internal defects of
concrete. It has been proved that the tensile strength of concrete is more sensitive to defects such as cracks and pores than the
compressive strength [20,21]. It is also proved that short-term heavy rainfall has a significant influence on the interlayer properties of
dam.
In general, short-term heavy rainfall significantly reduced the mechanical properties of concrete. This is mainly because rainfall
increases the w/b of concrete. The increase of w/b leads to the increase of residual water after concrete hardening. The water not
involved in hydration will gradually evaporate with the setting and hardening of concrete. Finally, many pores are left, which increases
the total porosity of the concrete (this inference is confirmed in subsection 3.2.2). In addition, a large amount of free water migrates
and bleeds under the action of pore water pressure, which also increases the porosity of concrete [22]. A previous study [23] indicated
that the total porosity of concrete increases with the increase of water content. Metha et al. believed that the concrete surface produced
a layer of laitance with the increase of surface water content [24]. Due to the large water-cement ratio of this part of the laitance, a
sparse and porous weak zone was formed between the layers, which affected the development of the interlayer mechanical properties.

45 2.0
40 Bulk concrete Layered concrete 1.8 Bulk concrete Layered concrete
Splitting tensile strength (MPa)
Compressive strength (MPa)

35 1.6
-50% -17% -12% -70% -58% -43%
30 +75%
1.4
+65%
1.2
25 +40% +90%
1.0
20
0.8
15
0.6
10 0.4
5 0.2
0 0.0
C0 A1 A2 A3 C0 A1 A2 A3
Groups of concrete Groups of concrete

(a) (b)
Fig. 10. Variation of concrete strength under different working conditions: (a) CS, (b) STS.

7
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

The influence of rainwater on the interlayer mechanical properties of dam can be reduced by discharging rainwater accumulated on
the layer or covering the insulation quilt. For example, the CS and STS of A2 (discharge rainwater that accumulated on the layer) were
increased by 65% and 40%, respectively, compared to A1 (no treatment measures). Compared with A1 (no treatment measures), the CS
and STS of A3 (covered with insulation quilt) was increased by 75% and 90%, respectively.

3.1.2. SS
Fig. 11a shows the influence of different treatment measures on the SS of concrete under rainfall conditions. The SS of bulk concrete
(C0) was 8.37 MPa. The interlayer SS of A1 (no treatment measures), A2 (draining rainwater) and A3 (covered with insulation quilt)
were 5.28, 6.99 and 7.56 MPa, which were 37%, 16% and 10% lower than bulk concrete, respectively. Compared with A1, draining the
rainwater accumulated on the layer and covering the insulation quilt can increase the SS of concrete by 32% and 43%, respectively.
Fig. 11b shows the relationship between peak SS and normal stress of concrete under short-term heavy rainfall. Where cohesion c is
the intercept of the straight line and friction coefficient f is the straight-line slope. Table 4 shows the values of c and f. The cohesion of
bulk concrete (C0) was 4.7 MPa. The cohesion of A1, A2 and A3 were 2.3, 3.8 and 3.9 MPa, decreased by 51%, 19% and 17%,
respectively, compared with bulk concrete. Cohesion depends on Coulomb force, van der Waals force and cementation between
particles. For concrete poured in layers, the greater the particle density at the layer, the greater the original cohesion. Furthermore, the
more cementitious material between the particles, the greater the cohesion. The cohesion of concrete decreased significantly under
rainfall conditions. The main reason is that the rainwater increases the w/b of concrete, bring about a decrease in the particle density of
concrete near the interlayer. In addition, some cementitious materials will be lost due to drainage measures. This is also the reason for
the decrease in cohesion of group A2 concrete. Changes in cohesion also suggested laterally that short-term heavy rainfall increased
porosity near the concrete interlayer (this inference was confirmed in 3.2.1).
The size of friction coefficient depends on sliding friction and occlusal friction between particles [25]. The friction coefficients for
C0, A1, A2, and A3 were 1.2, 1.0, 1.0, and 1.2, respectively. The friction coefficients of A1 and A2 were reduced by 17% than bulk
concrete. This may be due to the loose microstructure of concrete layers caused by short-term heavy rainfall, which reduces the
occlusal friction between cement particles.

3.1.3. Ultrasonic test results

The propagation velocity of ultrasonic waves in concrete can reflect the compactness of concrete [26]. The propagation velocity of
the ultrasonic wave in the specimen can be calculated by the interval time between the initiation point of the transmitting wave and
receiving wave [18]. The variation of ultrasonic velocity of concrete under rainfall conditions is shown in Fig. 12a. The ultrasonic
velocity through the bulk concrete (C0) was 5052 m/s. The wave velocities passing through A1 (no treatment measures), A2 (draining
rainwater) and A3 (covered with insulation quilt) were 4628, 4838 and 4961 m/s, which decreased by 8%, 4% and 2%, respectively,
compared with those through the C0. This may be due to short-term heavy rainfall causing defects between concrete layers (the SEM
test results summarized in 3.2.2 proved this conjecture). Studies have shown [17,27,28] that when the concrete age, aggregate type
and internal humidity are the same, the propagation rate of ultrasonic waves in solid media will slow down with the increase of defects.
Moreover, such defects can be reduced by draining rainwater (A2) and covering insulation quilt (A3).
Zhu et al. [29] proposed a damage variable defined by ultrasonic velocity (Eq. 4). This damage variable can reflect the deterioration
degree of concrete strength.

D = 1 − (Vt /V0 )2 (4)

Fig. 11. Results of concrete shear test: (a) SS, (b) ultimate SS vs. normal stress.

8
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Table 4
Interlayer shear resistance parameters of concrete under short-term heavy rainfall.
Groups Working conditions Cohesion c (MPa) friction coefficient f Correlation coefficient R2

C0 Bulk concrete 4.7 1.2 0.911

A1 No treatment measures 2.3 1.0 0.997
A2 Discharge rainwater accumulated on layers 3.8 1.0 0.942
A3 Covered with insulation quilt 3.9 1.2 0.982

Fig. 12. Ultrasonic nondestructive testing results: (a) ultrasonic velocity, (b) strength loss rate vs. damage degree.

D is the concrete damage degree; Vt is the ultrasonic speed through the layered concrete, and V0 is the ultrasonic speed through C0.
( )
y1 = 2.096x + 0.375 R2 = 0.959 (5a)
( )
y2 = 2.199x + 0.009 R2 = 0.983 (5b)

The damage degree was linearly correlated with the loss rate of interlayer STS and interlayer SS (Eqs. 5a and 5b), and the cor­
relation coefficients (R2) were 0.959 and 0.983, respectively (Fig. 12b). The loss rate of interlayer STS and interlayer SS increases with
the increase of concrete damage degree. This is consistent with the result of previous study [5]. In practical engineering, the interlayer
STS and SS of concrete can be estimated by the damage degree obtained from the ultrasonic nondestructive testing results. The
calculation results can provide a basis for the interlayer bonding quality control of the dam.

3.2. Microstructure

3.2.1. Pore structure

Fig. 13a shows the cumulative intrusion curve of mercury. The cumulative pore volumes of A1 (no treatment measures), A2
(draining rainwater), A3 (covered with insulation quilt) and C0 (bulk concrete) were 0.0727, 0.0641, 0.0597 and 0.0503 mL/g,
respectively. Due to short-term heavy rainfall, the total pore volume near the concrete interlayer increased significantly. That is, short-
term heavy rainfall adversely affected the internal pore structure of concrete. This is mainly because the short-term heavy rainfall
increased the w/b of concrete, increasing porosity.
Fig. 13b shows the log-differential pore size distribution curves (dV/dlog D) of different grouped specimens. The pore size showed
an obvious multi-peak distribution, with one major peak and two secondary peaks, respectively. Fig. 13b shows that there is no third
peak in bulk concrete. Under the action of short-term heavy rainfall, an obvious third peak appeared in concrete. After draining
rainwater and covering the insulation quilt, the third peak gradually shifts to the left (small aperture direction). The pore size range
was changed from 500–3000 nm to 200–500 nm. This shows that the above two treatments can effectively reduce the macropore
content and make the concrete microstructure denser.
To further analyse the influence of different treatment measures on the change of concrete pore structure under rainfall conditions,
the pores were divided into five levels: gel pores (＜10 nm)，fine capillary pores (10–50 nm)，middle capillary pores (50–100 nm)，
large capillary pores (100 nm ~10 µm) and macropores (≥10 µm) [30]. Among them, macropores have a detrimental influence on the
properties of concrete [31]. The content of macropores (≥10 µm) and large capillary pores in concrete increased significantly under
rainfall conditions (Fig. 13c). The volume of fine capillary pores (10–50 nm) was significantly reduced. This indicates that the pores
near the interlayers were roughened by short-term heavy rainfall. In addition, compared with no treatment measures, draining

9
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Fig. 13. Concrete pore structure: (a) cumulative pore volume, (b) pore size distribution, (c) pore volume fraction and (d) interlayer bonding
strength vs. porosity.

rainwater and covering the insulation quilt can reduce the content of macropores (≥10 µm). Based on the comprehensive analysis of
the changes of concrete mechanical properties (Figs. 10 and 11) and pore structure (Fig. 13b), the increase of macropores and large
capillary pores content is the main reason for the deterioration of concrete mechanical properties.
In general, the increase of concrete porosity will lead to the deterioration of mechanical properties [1,32,33]. Fig. 13d shows the
porosity vs. interlayer SS and STS of concrete. The porosity was linearly correlated with interlayer SS and interlayer STS. The interlayer
SS and interlayer STS of concrete decreased with increased porosity. This result is consistent with previous studies [34]. Fang et al.
believed that the STS decreases significantly with increasing of the porosity on the splitting surface [34].
Concrete pore structure (i.e., pore shape, area and volume, etc.) has obvious fractal characteristics. The complexity of concrete pore
structure can be quantified as a fractal dimension by fractal theory. For porous media such as concrete, the surface fractal dimension
(SFD) is generally between 2 and 3. If the fractal dimension is > 3 or < 2, the object is considered non-fractal [35,36]. The research
shows that the SFD has a certain correlation with the porosity and strength of concrete [1,37]. The porosity increases with the decrease
of SFD. Furthermore, the concrete strength is positively correlated with SFD.
The fractal model based on thermodynamics can quantitatively reflect the complexity of concrete pore structure. In the mea­
surement of mercury porosimetry porosity, the work done by external pressure on mercury is equal to the increment of surface energy
of mercury liquid entering the pores [38]. Therefore, the following equation can be obtained:
dW = − PdV = − γ cos θdS (6)

Where P is the pressure applied during mercury injection (Pa), V is the volume of mercury injected at pressure P (m3), S is the pore
surface area (m2) and θ is the contact angle.
Based on Eq. (6), Zhang [39,40] et al. deduced the expression of pore fractal dimension based on thermodynamics:
logW n = log Qn + C (7)

10
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Where Wn is the accumulated surface energy during the n-th stage of mercury intrusion, Qn is a function of pore radius and pore volume
during the n-th stage of mercury intrusion and C is the product of the surface tension and the cosine of the mercury contact angle.
Where Wn and Qn can be calculated according to Eqs. (8) and (9):
∑
n
Wn = Pi ΔV i (8)
i=1

Pi is the pressure of the i-th mercury injection (MPa); ΔV i is the i-th mercury injection amount (m3).

Qn = r2−n Ds
V Dn s /3 (9)

Where rn is the pore radius (m) corresponding to the n-th mercury injection, V n is the total mercury volume (m3) injected into the pores
before the n-th mercury injection and Ds is the surface fractal dimension of pores.
By substituting Eqs. (8) and (9) into (7), the following expressions can be obtained:
( ) ( 1/3 )
Wn V
log 2 = Ds log n +C (10)
rn rn

Perform linear regression analysis on mercury intrusion data according to Eq. 10, and the slope of the obtained straight line is SFD
of concrete.
Fig. 14 shows the surface fractal dimension of samples after pore size subdivision. In the range of gel pores (<10 nm), the SFD value
of A1 is the smallest (about 2.611), which is close to the Menger sponge model [40]. The SFD of C0, A1 and A2 is a transitional
non-fractal region (SFD>3) in the range of fine capillary to middle capillary pores (10–100 nm). In the macropore (≥10 µm) range, the
SFD value of A1 is the largest (about 2.637). This suggests that the macroporous structure of A1 is more complex.
The overall fractal dimensions (Df) of A1 (no treatment measures), A2 (draining rainwater), A3 (covered with insulation quilt) and
C0 (bulk concrete) were 2.909, 2.929, 2.973 and 2.978, respectively. We found that the fractal dimension of layered concrete (i.e., A1,
A2 and A3) is lower than C0. In addition, there is also a high correlation between SFD and porosity (Table 5). The SFD of concrete
decreases with increased porosity, indicating that the concrete pore structure may have undergone serious degradation. Concrete
splitting tensile strength and shear strength tests prove this conjecture. The results of literature [37] also support this conjecture.

3.2.2. Scanning electron microscope

Fig. 15 shows the microstructure of concrete after curing for 28d. Many irregularly shaped macropores (Fig. 15a) were formed in
the interlayer of concrete under strong rainfall conditions. This is mainly because rainfall increases the w/b of concrete. Only a small
part of the water infiltrated into the concrete is involved in the hydration reaction. Water not involved in hydration occupies much
space. Many air voids are left as the concrete sets and hardens after the residual water evaporates. In addition, draining the rainwater
accumulated on the layer can reduce the number of air voids (Fig. 15b). However, from the strength change of concrete (Fig. 10), the
improvement effect of drainage measures on mechanical properties is limited. When the concrete surface is covered with an insulation
quilt, the microstructure of the concrete near the interlayer is denser (Fig. 15c). The SEM results verified the characterization results of
MIP on the concrete pore structure.

3.3. Effects of short-term heavy rainfall

3.3.1. Mechanical properties and microstructure

Short-term heavy rainfall significantly affected the interlayer mechanical properties, damage degree and microstructure of con­
crete. The CS (Fig. 10a), STS (Fig. 10b) and SS (Fig. 11a) of concrete decreased by 50%, 70% and 37%, respectively, under the action of
short-term heavy rainfall, which was compared with bulk concrete (C0). Meanwhile, the results of ultrasonic nondestructive testing
showed that the short-term heavy rainfall increased the damage degree of concrete near the interlayer (Fig. 12).
Furthermore, the short-term heavy rainfall increased the porosity of concrete near the interlayer. The porosity of concrete near the
interlayer increased from 11.53% (C0) to 16.22% under the action of short-term heavy rainfall. The content of macropores (≥10 µm)
and large capillary pores (100 nm ~ 10 µm) in the concrete significantly increased under the action of short-term heavy rainfall, while
the volume of fine capillary pores decreased significantly (Fig. 13c).
In summary, short-term heavy rainfall causes defects (such as pores and microcracks) near the concrete interlayer. Furthermore,
the defects such as pores and micro-cracks increased the damage degree inside the concrete, which led to the deterioration of the
interlayer mechanical properties.

3.3.2. Mechanism for the effects of short-term heavy rainfall

Table 5 shows that the CS, STS and SS of dam concrete decreased significantly under short-term heavy rainfall conditions. The
influence of strong rainfall on the interlayer mechanical properties of dam concrete can be effectively alleviated by discharging
rainwater accumulated on layers and covering insulation quilt. In addition, with the decrease of concrete porosity, the CS, STS, SS,
ultrasonic wave velocity and fractal dimension increased, while the damage degree decreased. This is mainly related to the change of
concrete microstructure:
Freshly poured concrete is less hydrated. At this time, rainwater easily penetrates the interior of the concrete, causing the increase

11
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Fig. 14. Logarithm plots of W n /r2n vs. V 1/3

n /rn in various pore ranges: (a) C0, (b) A1, (c) A2, (d) A3.

Table 5
Properties parameters of concrete under short-term heavy rainfall.
Groups CS (MPa) STS (MPa) SS (MPa) Ultrasonic velocity (m/s) Damage degree Porosity (%) overall SFD (Df)

A1 15.9 0.42 5.28 4628 0.161 16.22 2.909

A2 26.3 0.59 6.99 4838 0.083 14.54 2.929
A3 27.9 0.80 7.56 4961 0.038 13.36 2.973
C0 31.70 1.40 8.37 5052 / 11.53 2.978

of actual water content. The increase of w/b is not conducive to the development of concrete mechanical properties [41]. This is
mainly because concrete with a large w/b will retain more water after hydration. The water not involved in hydration leaves more
pores after evaporation (Fig. 15a), increasing the total porosity of concrete. In addition, a large amount of free water migrates and
bleeds under the action of pore water pressure, which also increases the porosity of concrete [22].
Drainage measures can only reduce the rainwater accumulated on the concrete surface but cannot drain the water that seeps into
the concrete. Moisture infiltrating into the concrete increases the w/b, increasing the macropore content (Fig. 13c). In addition, the
drainage of rainwater will also lead to the loss of some cementitious materials. This also leads to the increase of the w/b and total
porosity of concrete (Table 5). This is also why the concrete interlayer bonding strength still cannot reach the strength of bulk concrete
after taking drainage measures.
Compared with the measures of draining rainwater, covering with an insulation quilt can more reduce the influence of strong
rainfall on the mechanical properties of dam concrete. This is mainly because the insulation quilt prevents rainwater from infiltrating
the concrete interior. In addition, the insulation quilt can also weaken the influence of environmental factors on the lower layer of

12
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Fig. 15. Micro morphology of concrete under short-term heavy rainfall: (a) no treatment measures, (b) discharge rainwater accumulated on layers
and (c) bulk concrete and covering with insulation quilt.

concrete [42].

4. Conclusion

This study analyzed the influences of different treatment measures (no treatment measures, discharge rainwater accumulated on
layers and covering with insulation quilt) on the CS, SS and STS of concrete under strong rainfall conditions. Then, the changes of pore
structure and micromorphology of concrete under short-term heavy rainfall are analyzed by MIP and SEM. The conclusions are as
follows:

(1) Short-term heavy rainfall has an adverse effect on the CS, STS and SS of concrete. Draining rainwater accumulated on the layer
and covering the insulation quilt can improve the microstructure near the concrete layer and improve the interlayer mechanical
properties of concrete. Compared with no treatment measures, the interlayer STS of concrete increases by 40% and 90% after
draining rainwater and covering with insulation quilt; The SS increased by 70% and 74%. MIP and SEM test results show that
short-term heavy rainfall increases the macropore content and significantly increases the porosity of concrete. This is mainly
because the short-term heavy rainfall increases the w/b of concrete. In addition, draining rainwater or covering insulation quilt
can reduce the porosity of concrete. Furthermore, the increase of macropores and large capillary pores near the interlayer leads
to the deterioration of interlayer mechanical properties of concrete.
(2) The overall SFD correlates with concrete strength and porosity. As the overall SFD decreases, the porosity increases, while the
mechanical properties of concrete decrease. The interlayer bonding strength of concrete can be predicted by ultrasonic velocity,
damage degree, porosity and overall SFD. The damage degree is linearly correlated with the loss rate of interlayer STS and
interlayer SS (Eqs. 5a and 5b), and the correlation coefficients (R2) are 0.959 and 0.983, respectively. The interlayer properties
of concrete can be predicted by ultrasonic velocity in a construction site. In addition, porosity is linearly correlated with
interlayer STS and interlayer SS, with correlation coefficients of 0.909 and 0.973, respectively.

This paper mainly studied the influence of traditional treatment measures (discharging rainwater accumulated on the layer and
covering insulation quilt) on the interlayer properties of concrete under short-term heavy rainfall. These two treatment measures have
a limited effect on improving concrete interlayer properties. To better improve the interlayer properties of concrete, mortar containing
nanomaterials (such as nano-silica and nano-clay, etc.) or fibers (such as polyvinyl alcohol fibers, carbon fibers, and steel fibers) can be
laid on the layer in the future.

CRediT authorship contribution statement

Wei-Jia Liu: Project administration, Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review &
editing. Yao-Sheng Tan: Data curation, Investigation. Qing-Bin Li: Funding acquisition, Supervision. Ning Yang: Data curation.

13
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

Chun-Feng Liu: Investigation. Xiao-Feng Gao: Validation. Bo Zhang: Validation. Yu Hu: Funding acquisition, Methodology, Su­
pervision. Xu-Jing Niu: Methodology, Formal analysis, Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data Availability

Data will be made available on request.

Acknowledgments

The research is funded by the National Natural Science Foundation of China (Nos. 51979145, 51839007, 52008230 and
52009123), the Research Projects of China Three Gorges Group Corporation (Contract number: BHT/0802 and WDD/0458), and
Zhejiang Provincial Natural Science Foundation of China (No. LY22E090004).

References

[1] L. Wang, M. Jin, Y.H. Wu, et al., Hydration, shrinkage, pore structure and fractal dimension of silica fume modified low heat Portland cement-based materials,
Constr. Build. Mater. 272 (2021), 121952, https://doi.org/10.1016/j.conbuildmat.2020.121952.
[2] W.J. Liu, Y.B. Wang, Q.B. Li, et al., Research on interlayer bonding quality control method of dam concrete based on equivalent age, Materials 14 (18) (2021)
5192, https://doi.org/10.3390/ma14185192.
[3] X. Feng, Y. Zhang, Z. Fan, et al., Experimental study on impact of horizontal weak layer on dynamic failure mechanism of Roller Compacted Concrete Arch Dam,
J. Hydraul. Eng. 47 (12) (2016) 1493–1501, https://doi.org/10.13243/j.cnki.slxb.20160116.
[4] Y. Ye, G. Liu, P. Li, F. Chen, Dealing with leakage of Xibing roller compacted concrete thin arch dam, Adv. Sci. Technol. Water Resour. 25 (2005) 27–31, https://
doi.org/10.3880/j.issn.1006-7647.2005.03.009.
[5] X.J. Niu, Q.B. Li, W.J. Liu, et al., Effects of ambient temperature, relative humidity and wind speed on interlayer properties of dam concrete, Constr. Build.
Mater. 260 (2020), 119791, https://doi.org/10.1016/j.conbuildmat.2020.119791.
[6] A. Joshaghani, M. Balapour, A. Ramezanianpour, Effect of controlled environmental conditions on mechanical, microstructural and durability properties of
cement mortar, Constr. Build. Mater. 164 (2018) 134–149, https://doi.org/10.1016/j.conbuildmat.2017.12.206.
[7] A.C.B. Ribeiro, J. Díez-Cascón, A.F. Gonçalves, Roller compacted concrete - tensile strength of horizontal joints, Mater. Struct. 34 (2001) 413–417, https://doi.
org/10.1617/13665.
[8] W.B. Xu, Y. Hu, Q.B. Li, Interlayer strength of hydraulic concrete and layer surface treatment improvement in Wudongde Dam, J. Tsinghua Univ. Sci. Technol.
57 (2017) 845–850.
[9] L.S. Feng, Y.H. Zhou, Study on impact to RCC construction from solar radiant heat, Hongshui River 20 (01) (2001) 14–19.
[10] G.M. Moelich, J.E. van Zyl, N. Rabie, et al., The influence of solar radiation on plastic shrinkage cracking in concrete, Cem. Concr. Compos. 123 (2021), 104182,
https://doi.org/10.1016/j.cemconcomp.2021.104182.
[11] A.S. El-amary, B.T. Alsulami, N.M.W. AL-Bustami, Experimental study on the effect of rainfall on fresh concrete, ARPN J. Eng. Appl. Sci. 10 (15) (2015)
6773–6780.
[12] L. Yang, J.J. Shi, Experimental study on the impact of rainfall on RCC construction, J. Constr. Eng. Manag. 136 (5) (2010) 477–483, https://doi.org/10.1061/
(ASCE)CO.1943-7862.0000156.
[13] J. Wang, W. Duan, Z.M. Ma, Statistical study on heavy rain since the construction of wudongde hydropower station in Jinsha river, , Plateau Mt. Meteorol. Res.
39 (02) (2019) 62–69, https://doi.org/10.3969/j.issn.1674-2184.2019. 02.010.
[14] Climate Change 2021: The Physical Science Basis. Working Group I contribution to the Sixth Assessment Report of the In-tergovernmental Panel on Climate
Change.
[15] R.J.H. Dunn, L.V. Alexander, M.G. Donat, et al., Development of an updated global land in-situ-based dataset of temperature and precipitation extremes:
HadEX3, J. Geophys Res Atmos. 125 (16) (2020), https://doi.org/10.1029/2019JD032263.
[16] China Electric Power Press, Test Code for Hydraulic Concrete; DL/T 5150-2017, China Electric Power Press, Beijing, China, 2017.
[17] P.K. Mehta, P.J.M. Monteiro, Concrete: Microstructure, Properties, and Materials, McGraw-Hill, New York, NY, USA, 2013.
[18] D.H. Yin, Q.J. Xu, Comparison of sandstone damage measurements based on non-destructive testing, Materials 13 (22) (2020) 5154, https://doi.org/10.3390/
ma13225154.
[19] Z. Zhang, G.W. Scherer, Physical and chemical effects of isopropanol exchange in cement-based materials, Cem. Concr. Res. 145 (2021), 106461, https://doi.
org/10.1016/j.cemconres.2021.106461.
[20] A.M. NEVILLE, Properties of Concrete, Pearson Education Limited, Harlow, 2011, p. 328.
[21] Y.N. CHAN, G.F. PENG, M. ANSON, Residual strength and pore structure of high-strength concrete and normal strength concrete, Cem. Concr. Compos 21 (1)
(1999) 23–27, https://doi.org/10.1016/S0958-9465(98)00034-1.
[22] P. Zhou, S.L. Xie, Q. Li, Effect of water-binder ratio on properties and pore structure of concrete, Bull. Chin. Ceram. Soc. 37 (03) (2018) 974–978, https://doi.
org/10.16552/j.cnki.issn1001-1625.2018.03.036.
[23] D. Zhu, Z. Li, Effect of aggregates and water contents on the properties of magnesium phospho-silicate cement, Cem. Concr. Compos. 27 (1) (2005) 11–18,
https://doi.org/10.1016/j.cemconcomp.2004.03.003.
[24] P.K. Mehta, P.J.M. Monteiro, Concrete: Microstructure, Properties, and Materials, McGraw-Hill, New York, NY, USA, 2013.
[25] Y.D. Lou. Study on the effects of layer processing on the layer adhesion properties of RCC, Zhejiang University. 2015. Master Thesis, Hangzhou, China, 2015.
[26] V. Malhotra, N.J.Carino Mohan, Handbook on Nondestructive Testing of Concrete, CRC Press, 2003.
[27] S. Maalej, Z. Lafhaj, M. Bouassida, et al., Micromechanical modelling of dry and saturated cement paste: porosity assessment using ultrasonic waves, Mech. Res.
Commun. 51 (3) (2013) 8–14, https://doi.org/10.1016/j.mechrescom.2013.03.002.
[28] M.G. Hernández, J.J. Anaya, L.G. Ullate, et al., Application of a micromechanical model of three phases to estimating the porosity of mortar by ultrasound, Cem.
Concr. Res. 36 (4) (2006) 617–624, https://doi.org/10.1016/j.cemconres.2004.07.018.
[29] J.S. Zhu, Y.P. Song, Research on fatigue damage of concrete under biaxial compressive loading using ultrasonic velocity method, Chin. J. Rock. Mech. Eng. 23
(13) (2004) 2230–2234.
[30] L. Zhang, J. Zhou, Fractal characteristics of pore structure of hardened cement paste prepared by pressurized compact molding, Constr. Build. Mater. 259
(2020), 119856, https://doi.org/10.1016/j.conbuildmat.2020.119856.

14
W.-J. Liu et al. Case Studies in Construction Materials 17 (2022) e01544

[31] Z.Q. Zhang, B. Zhang, P.Y. Yan, Hydration and microstructures of concrete containing raw or densified silica fume at different curing temperatures, Constr.
Build. Mater. 121 (2016) 483–490, https://doi.org/10.1016/j.conbuildmat.2016.06.014.
[32] F. Vodak, K. Trtik, O. Kapickova, The effect of temperature on strength – porosity relationship for concrete, Constr. Build. Mater. 18 (7) (2004) 529–534,
https://doi.org/10.1016/j.conbuildmat.2004.04.009.
[33] C. Lian, Y. Zhuge, S. Beecham, The relationship between porosity and strength for porous concrete, Constr. Build. Mater. 25 (11) (2011) 4294–4298, https://doi.
org/10.1016/j.conbuildmat.2011.05.005.
[34] X. Fang, C. Wang, H. Li, et al., Influence of mesoscopic pore characteristics on the splitting-tensile strength of cellular concrete through deep-learning based
image segmentation, Constr. Build. Mater. 315 (2022), 125335, https://doi.org/10.1016/j.conbuildmat.2021.125335.
[35] Q. Zeng, M.Y. Luo, X.Y. Pang, et al., Surface fractal dimension: an indicator to characterize the microstructure of cement-based porous materials, Appl. Surf. Sci.
282 (2013) 302–307, https://doi.org/10.1016/j.apsusc.2013.05.123.
[36] Q. Zeng, K.F. Li, F.C. Teddy, et al., Surface fractal analysis of pore structure of high-volume fly-ash cement pastes, Appl. Surf. Sci. 257 (3) (2010) 762–768,
https://doi.org/10.1016/j.apsusc.2010.07.061.
[37] S. Jin, G. Zheng, J. Yu, et al., A micro freeze-thaw damage model of concrete with fractal dimension, Constr. Build. Mater. 257 (2020), 119434, https://doi.org/
10.1016/j.conbuildmat.2020.119434.
[38] H.M. Rootare, C.F. Prenzlow, Surface areas from mercury porosimeter measurements, J. Phys. Chem. 71 (8) (1967) 2733–2736, https://doi.org/10.1021/
j100867a057.
[39] B. Zhang, S. Li, Determination of the surface fractal dimension for porous media by mercury porosimetry, Ind. Eng. Chem. Res. 34 (4) (1995) 1383–1386,
https://doi.org/10.1021/ie00043a044.
[40] B. Zhang, W. Liu, X. Liu, Scale-dependent nature of the surface fractal dimension for bi- and multi-disperse porous solids by mercury porosimetry, Appl. Surf.
Sci. 253 (3) (2006) 1349–1355, https://doi.org/10.1016/j.apsusc.2006.02.009.
[41] S. Zheng, T. Liu, G. Jiang, et al., Effects of water-to-cement ratio on pore structure evolution and strength development of cement slurry based on
HYMOSTRUC3D and micro-CT, Appl. Sci. 11 (7) (2021) 3063, https://doi.org/10.3390/app11073063.
[42] Y. Hu, X.J. Niu, N. Yang, et al., Study on the risk of early plastic cracking of low heat cement concrete, Concrete 03 (2021) 19–22, https://doi.org/10.3969/j.
issn.1002-3550.2021.03.005.

15