Case Studies in Construction Materials 18 (2023) e01980

Contents lists available at ScienceDirect

Case Studies in Construction Materials

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

Case study

Sound absorption performance of a triple-hole structure in green

ceramsite concrete for high-speed-railway sound barriers:
Experiments and neural network modeling
Chunshen Wang a, Xinyu Ren a, Quanwei Diao a, Xiaoyan Pan b, Weiqing Su b,
Liang Chang b, Bin Lin a, Shuai Yan a, *
a
Key Laboratory of Advanced Ceramics and Machining Technology of Ministry of Education, Tianjin University, Tianjin 300350, China
b
China Railway Design Corporation, Tianjin 300308, China

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

Keywords: Green ceramsite concrete is an environmentally friendly sound-absorbing material for high-speed
Sound absorption railway sound barriers. However, given its low sound absorption, this material is not widely used.
Green ceramsite Therefore, this study proposes a triple-hole structure to improve the sound absorption of green
High-speed railway
ceramsite concrete. The triple-hole structure comprised micro-sized holes in the ceramsite,
Triple-hole structure
Noise control
medium-sized holes between the ceramsite, and surface macro-sized holes. The surface macro-
sized holes were obtained by changing the mold shape, and the other holes were formed by
formula control. The influence of the surface macro-sized holes on the acoustic performance of
the triple-hole structure was analyzed using an impedance tube test. Results showed that the
sound absorption coefficient increased when the hole diameter decreased or the surface porosity
increased. The optimal surface macro-sized holes increased the average sound absorption coef­
ficient by 83 %. The surface macro-sized holes improved the acoustic performance by absorbing
the reflection of incident sound and not by resonance effect. This effect was attenuated when the
relative surface area was increased by more than 100 %. Compared with the ANN method, the
finite element simulation method cannot accurately reflect the acoustic effects of the triple hole
structure. Back propagation neural network was used to predict the effect and can realize an
accuracy of 95.5 %. Thus, this study proposes a reference for the design of sound-absorbing
concrete based on frequency.

1. Introduction

The construction of a global high-speed railway has become the status quo and trend. The ensuing noise pollution has greatly
affected the health of urban residents and the environment [1], so it is also being paid increasing attention [2,3]. Ballast and
sound-absorbing building materials are frequently utilized to solve this problem [4,5]. Sound-absorbing materials include porous,
resonance, and composite materials [6]. Porous materials include fibrous materials, sponge materials, and concrete materials [7].
Among the porous materials, fibrous materials are poor in terms of sound absorption band [8], and sponge materials are easily ignited
[9]. While concrete materials have the advantage of the availability of their raw materials, simple preparation, environmentally

* Corresponding author.
E-mail address: [email protected] (S. Yan).

https://doi.org/10.1016/j.cscm.2023.e01980
Received 18 December 2022; Received in revised form 3 February 2023; Accepted 4 March 2023
Available online 9 March 2023
2214-5095/© 2023 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/).
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

friendly, and flame-retardant properties [10–13]. Thus, concrete materials have been widely used.
Among the concrete materials, foamed concrete and concrete with aggregate have received extensive research and application
[14]. Foamed materials with good sound absorption coefficient need a complex fabrication method to accurately control the foam
structure [8]. Concrete with aggregate is widely used owing to the easy availability of its raw materials and its simple preparation [12].
These materials achieve sound absorption through the porous structure naturally formed between particles [15]. Traditional concrete
with aggregate is made from raw materials such as perlite and is highly dependent on mineral resources [16]. As a kind of lightweight
aggregate, ceramsite is easily obtained, can be fired from river silt and solid waste, and has the advantage of environmental protection.
Thus, it has a good application prospect [17]. Therefore, this study chose green ceramsite fired from river silt as raw materials for
research.
In previous studies, researchers have adjusted formulations and created the needed aggregates to improve the acoustic performance
of concrete. Yoon et al. [18] mixed bottom ash aggregate into porous concrete, Naenudon et al. [19]used aggregates with waste fibers,
and Luo et al. [20] adjusted the particle size of the ceramsite aggregates and the thickness of the material to improve the SAC. In
addition, recycled rubber particles from waste tires [21,22], gravel [23], aluminum powder [24], cork [25], high furnace slag [26],
bagasse [27], and other raw materials [5,28] were also used as aggregates.
Researchers also adopted the method of structural design to improve the sound absorption coefficient (SAC). Rodrigues et al. [29]
investigated the effect of pore properties, Atalla et al. [30] designed non-uniform sponge materials and Pereira et al. [31–33] used a
resonance structure. Pyo et al. [34,35] obtained the resonance cavity by surface perforating on the cement mortar and revealed the
cooperative effect of macro- and micro-sized holes. The method used by Pyo provided us with a new idea of material design. Therefore
introducing surface macro-sized holes into concrete with lightweight aggregate with a multi-stage pore structure may have a better
sound absorption performance.
The artificial neural network (ANN) prediction method has been widely used in concrete performance optimization. The
compressive strength and load curves of concrete can be predicted by neural networks [36,37]. Researchers have used artificial neural
networks (ANN) to predict the acoustic environment in cars with high accuracy [38]. Therefore, the ANN is used to accurately predict
the acoustic effect. Moreover, finite element simulation methods are also used to predict this effect, and the results were compared with
those of ANN.
This study characterized the SAC of triple-hole concrete with lightweight aggregate, and the influence of the various parameters,
such as the diameter of holes, surface porosity, and air cavity, on the acoustic performance is studied. Meanwhile, the ANN is used to
predict the acoustic effects of the triple-hole structure. Therefore, a method to design the acoustic performance of green ceramsite
concrete is introduced based on this.

2. Materials and methods

2.1. Materials

Fig. 1 shows the application scenario of finished products and the preparation process of raw materials, and Table 1 illustrates the

Fig. 1. Materials: (a) application scenario of finished products and (b) preparation process of raw materials.

2
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

density of the materials used. The finished sound barrier is placed in the middle of two porous metal plates and does not carry
structural capacity when used. The materials used for the experiment are as follows.

• CEM I 42.5N cement (Gongyi Ruixue Color Cement Co. LTD, China) conforming to standard EN197-1 [39] was used as the
cementing agent.
• Polypropylene staple fiber (Tengxin New Material Co., LTD, Shanghai, China) was used as the main additive fiber in the experi­
ment, and the fiber diameter was less than 100 µm. Fig. 2(a) shows the length distribution of the fibers used.
• The aggregate was green ceramsite made from river silt and water treatment sludge (Huainan Jinrui New Building Materials Co.,
LTD., Anhui, China). Fig. 2(b) shows the particle size distribution of the ceramsite aggregate used. The water absorption was 0.16
and the porosity (according to EN 12390-5:19 [40]) was 0.58. The properties were measured according to EN 1097-6:2013 [41].

• A polycarboxylic acid water reducer (Hongxiang New Material Co., LTD, Shandong, China) was used as the main additive. The
water reduction rate was 25 %.

2.2. Specimen preparation

In the experiment, the specimen was made following the steps shown in Fig. 3. The dried ceramsite was wetted and mixed with
cement with additives, and then mixed with water to form a semi-dry material. The mixed material then was poured into the mold and
pressed to form. After the specimens were formed, the mold is released for use. Table 2 shows the details of concrete [42–44]. The
specimens were cylinders with a diameter of 100 mm and a height of 50 mm, some of which have a triple-hole structure. Then the
specimens were placed in the steam curing box for curing after demolding. The curing time and temperature were set following the
standard ACI 517-2R [45]. The curing temperature was not higher than 60 ℃, the air humidity was not less than 90 %. After four hours
of standing, four hours of warming, four hours of holding temperature, and four hours of cooling, the preliminary preparation of the
specimen was completed. The prepared specimens were dried at room temperature for 14 days before being used in the
sound-absorption test to ensure the accuracy of the test. Three specimens for each set of parameters were fabricated to reduce the
chance of error. The specimen is a cylindrical cement with a height of 50 mm and a diameter of 100 mm. The porosity of the specimen
was 0.34 by water inlet test.
The triple-hole structure is a structure with holes of various sizes: micro, medium, and macro. The micro-sized holes inside the
material are smaller than the ceramsite, the medium-sized holes between the ceramsite are about the size of the ceramsite, and the
surface macro-sized holes are larger than the ceramsite. Fig. 4 shows the triple-hole structure. Table 3 shows the details of the surface
macro-sized holes.

2.3. Sound absorption testing methods

The sound absorption of concrete was characterized by an impedance tube (Type 4206, Hottinger Brüel & Kjaer Ibérica, Nærum,
Denmark). When the impedance tube tests the absorption coefficient, a sound source at one end of the tube emits a flat wave that
travels through the tube and bounces off the specimen, forming a standing wave in front of it. The SAC can be obtained by measuring
the standing wave ratio between the wave belly and the node nearest to the specimen. Fig. 5 shows a picture of the specimens for the
test. Fig. 6 presents the diagram of the test. The PTFE raw tape was wrapped around the sides of specimens by ISO 10534-2:1998 [46]
to ensure the consistency of the test. The specimen was placed in two different situations during the test to investigate the synergistic
effect of the backing cavity and the triple-hole structure.

2.4. Finite element simulation methods

Finite element simulation methods were used to predict the effects of this triple-hole structure. COMSOL Multiphysics software was
used for finite element simulation analysis. The finite element simulation model of porous sound-absorbing concrete was developed
based on the theoretical model of JCA. With the accurate parameters of the radius, tortuosity factor, porosity and flow resistance of the
concrete, the simulation calculation of the material’s SAC can be completed. The model is composed of a perfect matching layer, an air
layer, a material layer, and an air cavity layer. A cylindrical material with a thickness of 50 mm, an air cavity layer with a height of
50 mm, and an air layer with a height of 600 mm were established to simulate the internal acoustic environment of the impedance
tube. The surface of the material layer was perforated to simulate the macro-sized holes. Ideal material parameters were used as the FE
simulation parameters, and the difference between medium- and micro-sized holes is ignored to investigate the synergistic effect of the
triple-hole structure. The porosity of the ideal material is 0.34, and the flow resistance is 21,300 Pa s/m2, just like the actual material.
The tortuosity is 20, the thermal characteristic length is 0.0019667 mm, the viscous characteristic length is 0.0016159 mm, and the

Table 1
Details of the ceramsite, cement, water reducer, and fiber.
Property Ceramsite Cement Water reducer Fiber

density (kg/m3) 648.7 3123 1073 913

3
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 2. (a) Length distribution of fibers and (b) particle size distribution of aggregate.

Fig. 3. Process of making specimens.

Table 2
Details of concrete (kg/m3).
Aggregate Cement Fiber Water reducer Water

636.62 127.32 0.38 1.02 89.13

aerodynamic viscosity is 1.8 * 10-5 Pa s [47]. Fig. 7 shows the FEM simulation model.

2.5. Artificial neural network methods

The artificial neural network (ANN) was used to predict the acoustic effects of the macro-sized holes. This network includes input,
hidden, and output layers. Fig. 8 shows the structure and the training flow chart of the artificial neural network. In the figure, D and P
are the parameters of the macro-sized holes in Table 3, h is the depth of holes, f is the frequency, and α is the SAC. Ek is the total
instantaneous energy errors of neural networks, and e is the error of one node. The kth training session is represented by pk . The
forward and reverse calculation processes make the prediction accurate.
The BP neural network used had two hidden layers containing 30 neurons. The ReLu function was used as the activation function of
neurons. The obtained process parameters and their corresponding SAC are divided into training sets and test sets according to
different samples. Among them, specimens D5-P9 and D15-P4.5 were used as test set data, and the rest data were used as training set.
80 % of the original data in the training set was utilized to train the model, while the remaining 20 % was utilized for validation. The
details of the surface macro-sized holes and SAC were used as input values and output values for training. We used correlation

4
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 4. Photos of the triple-hole structure: (a) micro-sized holes, (b)medium-sized holes, and (c) macro-sized holes.

Table 3
Details of surface macro-sized holes.a
Specimen label Diameter (mm) Surface porosity (%)

D5-P14 5 14
D5-P9 5 9
D5-P4.5 5 4.5
D10-P14 10 14
D10-P9 10 9
D10-P4.5 10 4.5
D15-P14 15 14
D15-P9 15 9
D15-P4.5 15 4.5
a
D is the diameter of the surface macro-sized holes, P is the surface porosity, and the depth of the holes
is 25 mm.

coefficients R2 and error values to evaluate the model quality. The maximum iteration is 1000, and the learning rate is 0.001.

3. Results and discussion

3.1. Results of the sound absorption test by using an impedance tube

3.1.1. Acoustic performance of the triple-hole structure without an air cavity

Fig. 9 shows the SAC of the specimens with the same surface porosity without an air cavity. The test specimen was 50 mm thick and
close to the rear wall of the impedance tube to avoid an air cavity. The concrete without surface macro-sized holes has an absorption
peak at 200–400 Hz and an absorption valley at 600–800 Hz. The overall SAC is lower than 0.4. The triple-hole structure significantly
improves the overall sound absorption coefficient of the material. When the porosity is 14 %, the specimen with a diameter of 5 mm
achieves the highest peak value of about 0.64 at 600–800 Hz. The SAC is the highest at 1600 Hz, with a value of 0.81. When the
porosity is 9 %, the specimen with a 5 mm aperture has a high peak value at 600–800 Hz, and the SAC decreases at 0–200 Hz. When
the porosity is 4.5 %, the specimen with a diameter of 5 mm obtains a high peak value of SAC (about 0.61) at 400–600 Hz. The result
shows that when the surface porosity is the same, the specimen with 5 mm holes reaches the highest peak value in every group.
However, the absorption coefficient is lower than that of the original specimen at 0–200 Hz. This result shows that the smaller the hole,
the greater the SAC increase.
When the hole diameter is the same, the specimens with 14 % surface porosity reach the highest peaks of SAC in every group. After
600 Hz, the SAC of the specimens in the porosity of 14 % even reaches above 0.6. Meanwhile, the specimens with 5 % surface porosity
reach the lowest peaks of SAC in every group. This result proves that for the specimens with the same hole diameter, the larger the
surface porosity is, the greater the increase of the SAC will be. The diameter of the macro-sized holes significantly affects the location of
the absorption peak. The smaller the diameter, the more the peak shifts to the right.

3.1.2. Acoustic performance of the triple-hole structure backed by an air cavity

In practical application, sound-absorbing materials frequently cooperate with an air cavity [48]. In this study, 50 mm between the
specimen and the rear wall of the impedance tube was used to simulate this condition. The SAC increases backed by an air cavity,
which is the same as the study reported by Kim et al. [49,50]. Fig. 10 shows the SAC backed by an air cavity. In every group, the
specimen with a 5 mm hole diameter obtains the highest SAC. This result shows that when the surface porosity is the same, in the case

5
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 5. Specimens for sound absorption test: (a) D5-P14, (b) D5-P9, (c) D5-P4.5, (d) D10-P14, (e) D10-P9, (f) D10-4.5, (g) D15-P14, (h) D15-P9, and
(i) D15-P4.5.

Fig. 6. Diagram of sound absorption test.

backed by an air cavity, the sound absorption performance gets better with the decrease in the diameter of the holes. In the specimen
with a porosity of 14 %, the specimen with a hole diameter of 5 mm obtains a higher SAC. The first absorption peak appears at
200–300 Hz for all three specimens, and the absorption valley appears at 400–800 Hz. Compared with the original specimen, the
absorption peak shifts to the right obviously, which is the same as those without an air cavity. In the specimen with a porosity of 9 %,
the specimen with a diameter of 5 mm has the best sound absorption coefficient. The SAC of the specimens is almost higher than 0.42
after 200 Hz when using a triple-hole structure backed by an air cavity, indicating an improvement. This result shows that when the
hole diameter is the same in the case backed by an air cavity, the SAC is better when the surface porosity is larger, similar to the
condition without an air cavity.

3.2. Analysis of the mechanism of the surface macro-sized holes

The friction between the hole wall and the air in the holes makes the sound absorbed [51]. Cox et al. [7] found that large holes

6
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 7. FEM simulation model: (a) material layer model and (b) integral model with a cavity.

Fig. 8. Flow chart of the training process.

increase sound absorption by lowering the incoming sound’s reflection like diffusers. To better investigate this effect, the average SAC
is utilized to evaluate the sound absorption. Eq. (1) shows the calculation of the average SAC [52].
∫ x2 /
a= ax dx (x2 − x1 ) (1)
x1

Where ax is the SAC at x Hz. Given that the impedance tube can measure SAC from 50 Hz to 1600 Hz, the frequency range chosen for
the analysis in this study is from 50 (x1) Hz to 1600 (x2) Hz [53]. The increase of the surface area resulting from surface macro-sized
holes is calculated as relative surface area increment (ΔS). The relative surface area increment and average SAC increment are
calculated by using Eqs. (2)–(5).
ΔS = (SA − SN )/SN ⋅ 100% (2)
/
SN = πd2 4 (3)
/ /
SA = πd2 4 + πPd2 h D (4)

7
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 9. SAC of specimens with the same surface porosity: (a) No surface macro-sized hole specimen, (b) P = 14 %, (c) P = 9 %, and (d) P = 4.5 %.

Δa = aA − aN (5)
In the equations, S denotes the effective surface area of the specimen with incident sound, and A and N represent different
specimens. A represents the test specimen, N is the standard specimen, d is the diameter of the specimen, D is the diameter of the
surface macro-sized holes, h is the depth, and P is the surface porosity. Fig. 11 shows the change in Δa as a function ofΔS with and
without a cavity. The average SAC increases with the increase in the effective surface area. When the relative surface area increases by
100 %, the increase of Δa becomes inconspicuous. This is because the surface area needs to be increased by decreasing the hole
diameter when the surface porosity is constant or by increasing the surface porosity when the hole diameter is constant. Both ways
make the walls for air friction decrease. When the hole diameter is considerably small, the holes do not reduce reflection as Cox et al.
[7] suggested. Consequently, the increase in sound absorption is not evident. From the point of practical use, the relative surface area
increased by the surface macro-sized holes ought to be less than 100 %. Moreover, the diameter of the surface macro-sized holes should
be larger than the ceramsite to form a triple-hole structure. When this triple-hole concrete is used with a backing cavity, the SAC can be
greatly improved.
Fig. 11(a) and (b) show the same trends. Therefore, we can conclude that the air cavity does not invalidate the triple-hole structure.
The macro-sized hole increases the surface area of the material and simultaneously reduces the average thickness of the material. As
the result, a thinner material will have a lower SAC. The addition of a backing cavity to a porous sound-absorbing material can be
approximately seen as increasing the thickness of the material. Therefore, the backing cavity can approximately raise the thickness of
the material to solve this problem, so that the surface macroscopic size hole can play a better role.
To distinguish the effects of holes of different sizes, we decomposed the SAC of the specimen into the original material SAC and the
SAC changed by surface macro-sized holes (Δα). Fig. 12 shows the results. The peak gradually moves in the low-frequency direction
with a decrease in surface porosity. The peak value increases with the decrease in diameter. The peak of Δα moves toward lower
frequencies and becomes smaller when the cavity is adopted. The location of this peak is less affected by hole diameter and more

8
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 10. SAC of specimens backed by an air cavity with the same surface porosity: (a) No surface macro-sized hole specimen, (b) P = 14 %, (c)
P = 9 %, and (d) P = 4.5 %.

Fig. 11. Diagram of the relation between surface area increments and incremental SAC (a) without an air cavity and (b) backed by an air cavity.

9
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 12. Δα of the specimens: (a) D5-P14, (b) D5-P9, (c) D5-P4.5, (d) D10-P14, (e) D10-P9, (f) D10-4.5, (g) D15-P14, (h) D15-P9, and (i) D15-P4.5.

affected by surface porosity.

Park et al. [34] found that the average SAC is not improved when the surface perforated structures influence sound absorption like a
resonant cavity, resulting in changes in the position of the absorption peaks. Accordingly, the SAC is significantly reduced at a specific
frequency. However, the results of this study do not conform to this phenomenon. This suggests that the triple pore structure absorbs
sound not through resonance but through the friction between air and the pore wall and the viscous resistance of air.

3.3. Results of the FEM simulation

Fig. 13 shows the △α obtained by simulation calculation. The trend of simulation results is basically consistent with experimental
results. When there is no air cavity, the double-hole structure reduces the sound absorption coefficient from 0 Hz to 750 Hz, the most at
450 Hz, and forms a peak of increasing sound absorption coefficient at 1200 Hz. The positions of the peaks are less affected by hole
diameter but more affected by hole surface porosity. The larger the hole diameter is, the smaller the peak value of sound absorption is.
The smaller the surface porosity is, the more the peak moves to the low frequency direction. When the air cavity is adopted, the
positions of the peaks move to the low frequency range, and the peak of negative peak becomes smaller, while the positive peak value
increases. This indicates that the coupling effect of the air cavity and the double-hole structure. The reason for the difference in
simulation is that the ideal materials are simulated instead of the actual materials. The results show that for ideal materials with a
single-hole structure, the use of surface macro-sized holes can also improve the SAC of the material. However, this improving effect is
less evident at 1200–1600 Hz than in materials with a triple-hole structure. This shows the limitations of the finite element simulation
method. In the finite element simulation method, the difference between the medium- and the micro-sized holes is not easy to reflect in

10
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 13. Δα of the FEM simulation: (a) D5-P14, (b) D5-P9, (c) D5-P4.5, (d) D10-P14, (e) D10-P9, (f) D10-4.5, (g) D15-P14, (h) D15-P9, and (i)
D15-P4.5.

Fig. 14. Variation of MSE (mean squared error).

the model. This will lead to a large deviation in the simulation of concrete with a triple-hole structure. Finite element simulation must
use accurate flow resistance, porosity, and other parameters of the material to complete the prediction and analysis of the performance
of the material, while ANN method does not rely on accurate material physical parameters. ANN method can complete the

11
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

Fig. 15. (a) Comparison of predicted and measured values for the validation set data and (b) cross plot of predicted and measured values for
new data.

Fig. 16. Comparison with other studies: (a) average SAC at 50–1600 Hz [54–58] and (b) the relative increase in average SAC [12,55,59,60].

establishment of a prediction model through several groups of material manufacturing parameters and the sound absorption coeffi­
cient of the finished product and can achieve high accuracy. This reduces the number of tests for the prediction and optimization of the
performance of new specimens, so an ANN method was introduced.

3.4. Results of the ANN prediction

When predicting the SAC of the triple-hole structure, the traditional prediction methods did not have high accuracy due to the
randomness of the micro- and medium-sized holes. The BP neural network was used to predict the influence of the macro-sized holes
and the backing cavity to better design the holes. The design parameters of the macro-sized holes were used as input parameters, and
the SAC is used as output parameters for prediction. Fig. 14 shows the variation of error in the iteration process.
Fig. 15(a) shows the results of the predicted and the measured value of the validation set in the data. The comparison shows that the
correlation coefficients can reach 0.9705. This finding illustrates the feasibility of using ANN to predict the effect of the triple-hole
structure on acoustic performance. Fig. 15(b) shows the results of the predicted and the measured value of the new specimens
using the trained model. The result shows that the empirical model established can predict the effect of new preparation parameters to
accurately optimize the specimens.
Fig. 16 shows the advantage of the triple-hole structure optimized by neural networks. The average SAC of the material was
predicted to be the best when the surface porosity was 15 %, the diameter of the large hole was 5 mm, and the concrete was used with
the backing cavity. The average SAC of the material at 50–1600 Hz reached 0.43 with the triple-hole structure and a cavity, which can
meet the requirements of high-speed railway. Moreover, the triple-hole structure can effectively improve the sound absorption

12
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

coefficient of the material compared with other optimization methods.

4. Conclusion

This study presented a triple-hole structure that was applied to green ceramsite concrete for high-speed railway sound barriers. This
triple-hole structure contains holes of three varied sizes, namely, micro-sized holes inside the ceramsite, medium-sized holes between
the ceramsite, and surface macro-sized holes. The SAC of the concrete with lightweight aggregate was studied by using an impedance
tube. The influence of the diameter of the surface macro-sized holes and surface porosity on sound absorption was also studied. The
mechanism of the triple-hole structure was explored. The results were as follows:

1. The triple-hole structure can improve the sound absorption of the green ceramsite concrete whether with or without a cavity. The
average SAC of the triple-hole specimens is increased by up to 89.5 % compared with the specimens without macro-sized holes. The
triple-hole structure can make the concrete with thin thickness have strong acoustic performance.
2. The sound absorption mechanism of the surface macro-sized holes is to expose the area to reduce the reflection of the incident
sound to the sound source. When the relative surface area caused by the surface macro-sized holes increased by more than 100 %,
the increase of the SAC became insignificant because the material removal ratio was considerably large, or the effect of reducing the
incident sound reflection was not obvious.
3. The surface perforated structure can be used simultaneously with the backing cavity. The combination of the two can significantly
improve the low frequency acoustic absorption coefficient of the material.
4. With reasonable design, the BP neural network can predict this effect with 95.5 % accuracy. This network can be used to design the
structure of concrete with lightweight aggregate.
5. ANN method can predict the properties of triple-hole structure materials more accurately. It does not rely on the parameter test of
the finished material, which is a more promising prediction and optimization method compared with the FEM method.

CRediT authorship contribution statement

Chunshen Wang: Visualization, Writing – original draft, Writing – review & editing, Methodology, Formal analysis, Validation,
Investigation. Xinyu Ren: Formal analysis, Investigation. Quanwei Diao: Formal analysis, Investigation. Xiaoyan Pan: Resources,
Supervision. Weiqing Su: Resources, Supervision. Liang Chang: Resources, Supervision. Bin Lin: Resources, Supervision. Shuai Yan:
Conceptualization, Formal analysis, Writing – review & editing, Resources, Supervision, Investigation.

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

The data that has been used is confidential.

Acknowledgements

This work was financially supported by Peiyang Scholar Young Backbone Teacher Fund (Grant no. 2020XRG-0042).

References

[1] S.C. Li, S.H. Shi, L. Bu, Rail network must protect giant pandas, Nature 545 (7654) (2017), https://doi.org/10.1038/545289a (289-289).
[2] F. Li, T. Zhou, Effects of objective and subjective environmental pollution on well-being in urban China: a structural equation model approach, Soc. Sci. Med.
249 (2020), 112859, https://doi.org/10.1016/j.socscimed.2020.112859.
[3] P. Galvín, A. Romero, J. Dominguez, Vibrations induced by HST passage on ballast and non-ballast tracks, Soil Dyn. Earthq. Eng. 30 (9) (2010) 862–873,
https://doi.org/10.1016/j.soildyn.2010.02.004.
[4] X.Y. Zhang, D. Thompson, H. Jeong, G. Squicciarini, The effects of ballast on the sound radiation from railway track, J. Sound Vib. 399 (2017) 137–150, https://
doi.org/10.1016/j.jsv.2017.02.009.
[5] N.I.R.R. Hannan, S. Shahidan, N. Ali, N.M. Bunnori, S.S.M. Zuki, M.H.W. Ibrahim, Acoustic and non-acoustic performance of coal bottom ash concrete as sound
absorber for wall concrete, Case Stud. Constr. Mater. 13 (2020), https://doi.org/10.1016/j.cscm.2020.e00399.
[6] X. Sagartzazu, L. Hervella-Nieto, J. Pagalday, Review in sound absorbing materials, Arch. Comput. Methods Eng. 15 (3) (2008) 311–342, https://doi.org/
10.1007/s11831-008-9022-1.
[7] T. Cox, P. d’Antonio, Acoustic Absorbers and Diffusers: Theory, Design and Application, Crc Press, 2016.
[8] T.S. Tie, K.H. Mo, A. Putra, S.C. Loo, U.J. Alengaram, T.C. Ling, Sound absorption performance of modified concrete: a review, J. Build. Eng. 30 (2020), https://
doi.org/10.1016/j.jobe.2020.101219.
[9] C. Liu, T. Zhang, Y. Luo, Y. Wang, H. Wang, Multifunctional polyurethane sponge coatings with excellent flame retardant, antibacterial, compressible, and
recyclable properties, Compos. Part B Eng. 215 (2021), 108785, https://doi.org/10.1016/j.compositesb.2021.108785.
[10] R. Fediuk, M. Amran, N. Vatin, Y. Vasilev, V. Lesovik, T. Ozbakkaloglu, Acoustic properties of innovative concretes: a review, Materials 14 (2) (2021) 398,
https://doi.org/10.3390/ma14020398.

13
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

[11] T.S. Tie, K.H. Mo, A. Putra, S.C. Loo, U.J. Alengaram, T.C. Ling, Sound absorption performance of modified concrete: a review, J. Build. Eng. 30 (2020), 101219,
https://doi.org/10.1016/j.jobe.2020.101219.
[12] P. Wang, C. Zhao, Study on reducing railway noise by porous concrete sound-absorbing panel, Mater. Res. Innov. 19 (2015) 1156–1160, https://doi.org/
10.1179/1432891714z.0000000001269.
[13] C.J. Churchill, D.K. Panesar, Life-cycle cost analysis of highway noise barriers designed with photocatalytic cement, Struct. Infrastruct. Eng. 9 (10) (2013)
983–998, https://doi.org/10.1080/15732479.2011.653574.
[14] K. Dhasindrakrishna, K. Pasupathy, S. Ramakrishnan, J. Sanjayan, Progress, current thinking and challenges in geopolymer foam concrete technology, Cem.
Concr. Compos. 116 (2021), https://doi.org/10.1016/j.cemconcomp.2020.103886.
[15] M. Vasina, D.C. Hughes, K.V. Horoshenkov, L. Lapcik, The acoustical properties of consolidated expanded clay granulates, Appl. Acoust. 67 (8) (2006) 787–796,
https://doi.org/10.1016/j.apacoust.2005.08.003.
[16] I. Palomar, G. Barluenga, J. Puentes, Lime-cement mortars for coating with improved thermal and acoustic performance, Constr. Build. Mater. 75 (2015)
306–314, https://doi.org/10.1016/j.conbuildmat.2014.11.012.
[17] M. Vasina, D. Placha, M. Mikeska, L. Hruzik, G.S. Martynkova, Sound absorption study of raw and expanded particulate vermiculites, Appl. Phys. a-Mater. Sci.
Process. 122 (12) (2016), https://doi.org/10.1007/s00339-016-0527-4.
[18] H.N. Yoon, J. Seo, S. Kim, T. Kil, D. Jang, J.H. Bae, H.K. Lee, Sound-absorption and NOx-removal performances of TiO2-incorporated porous concrete made with
bottom ash aggregates, Adv. Concr. Constr. 12 (1) (2021) 1–11, https://doi.org/10.12989/acc.2021.12.1.001.
[19] S. Naenudon, A. Vilaivong, Y. Zaetang, W. Tangchirapat, A. Wongsa, V. Sata, P. Chindaprasirt, High flexural strength lightweight fly ash geopolymer mortar
containing waste fiber cement, Case Stud. Constr. Mater. 16 (2022), https://doi.org/10.1016/j.cscm.2022.e01121.
[20] X.Y. Luo, W.F. Li, X.L. Jin, Effect of aggregate size and gradation on the ceramsite sound-absorbing materials, in: Proceedings of the International Conference on
Optical, Electronic Materials and Applications, Chongqing, PEOPLES R CHINA, 2011, pp. 450–+. 〈https://doi.org/10.4028/www.scientific.net/AMR.216.450〉.
[21] A.C. Corredor-Bedoya, R.A. Zoppi, A.L. Serpa, Composites of scrap tire rubber particles and adhesive mortar – noise insulation potential, Cem. Concr. Compos.
82 (2017) 45–66, https://doi.org/10.1016/j.cemconcomp.2017.05.007.
[22] N.F. Medina, D. Flores-Medina, F. Hernandez-Olivares, Influence of fibers partially coated with rubber from tire recycling as aggregate on the acoustical
properties of rubberized concrete, Constr. Build. Mater. 129 (2016) 25–36, https://doi.org/10.1016/j.conbuildmat.2016.11.007.
[23] S. Lin, G.B. Zhang, J.B. Sun, S.L. He, C.F. Chen, A.M. Morsy, X.Y. Wang, Acoustic performance optimization of a cementitious composite with a porous medium,
J. Build. Eng. 44 (2021), 103362, https://doi.org/10.1016/j.jobe.2021.103362.
[24] J. Yoon, H. Kim, T. Koh, S. Pyo, Microstructural characteristics of sound absorbable porous cement-based materials by incorporating natural fibers and
aluminum powder, Constr. Build. Mater. 243 (2020), 118167, https://doi.org/10.1016/j.conbuildmat.2020.118167.
[25] R.M. Novais, L. Senff, J. Carvalheiras, A.M. Lacasta, I.R. Cantalapiedra, J.A. Labrincha, Simple and effective route to tailor the thermal, acoustic and
hygrothermal properties of cork-containing waste derived inorganic polymer composites, J. Build. Eng. 42 (2021), https://doi.org/10.1016/j.
jobe.2021.102501.
[26] J. Li, X.G. Zhuang, C. Leiva, C. Arenas, A. Cornejo, X. Querol, N. Moreno, O. Font, C.F. Pereira, Utilization of boiler slag from pulverized-coal-combustion power
plants in China for manufacturing acoustic materials, Energies 13 (21) (2020) 5705, https://doi.org/10.3390/en13215705.
[27] P. Kathirvel, M. Gunasekaran, S. Sreekumaran, A. Krishna, Effect of partial replacement of ground granulated blast furnace slag with sugarcane bagasse ash as
source material in the production of geopolymer concrete, Mater. Sci. - Medzg. 26 (4) (2020) 477–481, https://doi.org/10.5755/j01.ms.26.4.23602.
[28] A.O. Dawood, H. Al-Khazraji, R.S. Falih, Physical and mechanical properties of concrete containing PET wastes as a partial replacement for fine aggregates, Case
Stud. Constr. Mater. 14 (2021), https://doi.org/10.1016/j.cscm.2020.e00482.
[29] P.C. Rodrigues, N.T.d.S. Braga, E.S. Arruda Junior, L.d.N. Pinheiro Cordeiro, G.d.S. Vieira de Melo, Effect of pore characteristics on the sound absorption of
pervious concretes, Case Stud. Constr. Mater. 17 (2022), https://doi.org/10.1016/j.cscm.2022.e01302.
[30] N. Atalla, R. Panneton, F.C. Sgard, X. Olny, Acoustic absorption of macro-perforated porous materials, J. Sound Vib. 243 (4) (2001) 659–678, https://doi.org/
10.1006/jsvi.2000.3435.
[31] C. Lagarrigue, J.P. Groby, V. Tournat, O. Dazel, O. Umnova, Absorption of sound by porous layers with embedded periodic arrays of resonant inclusions,
J. Acoust. Soc. Am. 134 (6) (2013) 4670–4680, https://doi.org/10.1121/1.4824843.
[32] O. Doutres, N. Atalla, H. Osman, Transfer matrix modeling and experimental validation of cellular porous material with resonant inclusions, J. Acoust. Soc. Am.
137 (6) (2015) 3502–3513, https://doi.org/10.1121/1.4921027.
[33] A. Pereira, L. Godinho, L. Morais, The acoustic behavior of concrete resonators incorporating absorbing materials, Noise Control Eng. J. 58 (1) (2010) 27–34,
https://doi.org/10.3397/1.3264649.
[34] S. Park, K.A. Moges, S. Pyo, Experimental study on the sound absorption performance of surface-perforated mortar, Constr. Build. Mater. 307 (2021), 124824,
https://doi.org/10.1016/j.conbuildmat.2021.124824.
[35] K.A. Moges, S. Park, S. Pyo, Sound absorption characteristics of surface perforated mortar with micro-sized pores, Appl. Acoust. 195 (2022), 108811, https://
doi.org/10.1016/j.apacoust.2022.108811.
[36] M. Zarringol, H.-T. Thai, Prediction of the load-shortening curve of CFST columns using ANN-based models, J. Build. Eng. 51 (2022), https://doi.org/10.1016/j.
jobe.2022.104279.
[37] H. Naderpour, A.H. Rafiean, P. Fakharian, Compressive strength prediction of environmentally friendly concrete using artificial neural networks, J. Build. Eng.
16 (2018) 213–219, https://doi.org/10.1016/j.jobe.2018.01.007.
[38] G.P. Tan, D.F. Wang, Q. Li, Vehicle interior sound quality prediction based on back propagation neural network, in: Proceedings of the 2nd International
Conference on Challenges in Environmental Science and Computer Engineering (CESCE), Haikou, PEOPLES R CHINA, 2011, pp. 471–7. 〈https://doi.org/10.
1016/j.proenv.2011.12.075〉.
[39] BSI, EN 197-1:2011: Cement, Composition Specifications and Conformity Criteria for Common Cement, BSI, London, UK, 2011.
[40] BSI, EN 12390-5, Testing Hardened Concrete–Part 5: Flexural Strength of Test Specimens, British Standards Institution-BSI and CEN European Committee for
Standardization, BSI, London, UK, 2009.
[41] BSI, EN 1097-6: 2013-Tests for mechanical and physical properties of aggregates, Determination of Particle Density and Water Absorption, BSI, London, UK,
2013.
[42] P. Gle, E. Gourdon, L. Arnaud, K.V. Horoshenkov, A. Khan, The effect of particle shape and size distribution on the acoustical properties of mixtures of hemp
particles, J. Acoust. Soc. Am. 134 (6) (2013) 4698–4709, https://doi.org/10.1121/1.4824931.
[43] S.B. Park, D.S. Seo, J. Lee, Studies on the sound absorption characteristics of porous concrete based on the content of recycled aggregate and target void ratio,
Cem. Concr. Res. 35 (9) (2005) 1846–1854, https://doi.org/10.1016/j.cemconres.2004.12.009.
[44] C. Arenas, C. Leiva, L.F. Vilches, H. Cifuentes, M. Rodriguez-Galan, Technical specifications for highway noise barriers made of coal bottom ash-based sound
absorbing concrete, Constr. Build. Mater. 95 (2015) 585–591, https://doi.org/10.1016/j.conbuildmat.2015.07.107.
[45] A.C. 517, Accelerated Curing of Concrete at Atmospheric Pressure—State of the Art, ACI Manual of Concrete Practice, ACI. Committee 517, Farmington Hills,
Mich, 1992.
[46] ISO, ISO 10534-2: Acoustics—Determination of Sound Absorption Coefficient and Impedance in Impedance Tubes—Part 2: Transfer-function Method, ISO
Geneva, Switzerland, 1998.
[47] J. Allard, N. Atalla, Propagation of Sound in Porous Media: Modelling Sound Absorbing Materials 2e, John Wiley & Sons, 2009.
[48] Y. Luo, J.-j. Lou, Y.-b. Zhang, J.-r. Li, Sound-absorption mechanism of structures with periodic cavities, Acoust. Aust. 49 (2) (2021) 371–383, https://doi.org/
10.1007/s40857-021-00233-6.
[49] H. Kim, J. Hong, S. Pyo, Acoustic characteristics of sound absorbable high performance concrete, Appl. Acoust. 138 (2018) 171–178, https://doi.org/10.1016/j.
apacoust.2018.04.002.

14
C. Wang et al. Case Studies in Construction Materials 18 (2023) e01980

[50] M. Li, W. van Keulen, E. Tijs, M. van de Ven, A. Molenaar, Sound absorption measurement of road surface with in situ technology, Appl. Acoust. 88 (2015)
12–21, https://doi.org/10.1016/j.apacoust.2014.07.009.
[51] L. Cao, Q. Fu, Y. Si, B. Ding, J. Yu, Porous materials for sound absorption, Compos. Commun. 10 (2018) 25–35, https://doi.org/10.1016/j.coco.2018.05.001.
[52] T.G. Basten, H.-E. de Bree, Full bandwidth calibration procedure for acoustic probes containing a pressure and particle velocity sensor, J. Acoust. Soc. Am. 127
(1) (2010) 264–270, https://doi.org/10.1121/1.3268608.
[53] F. D’Alessandro, F. Asdrubali, N. Mencarelli, Experimental evaluation and modelling of the sound absorption properties of plants for indoor acoustic
applications, Build. Environ. 94 (2015) 913–923, https://doi.org/10.1016/j.buildenv.2015.06.004.
[54] W.H. Sun, Y.C. Fang, L.S. Wu, Preparation and properties of down feather fibers reinforced cellulose composite aerogel, J. Porous Mater. (2022) 6, https://doi.
org/10.1007/s10934-022-01331-0.
[55] M. Pereir, J. Carbajo, L. Godinho, P. Amado-Mendes, D. Mateus, J. Ramis, Acoustic behavior of porous concrete. Characterization by experimental and inversion
methods, Mater. Constr. 69 (336) (2019), https://doi.org/10.3989/mc.2019.03619.
[56] E.S. Jang, C.W. Kang, Investigation of the pore structure and sound-absorbing capability of the Chinese parasol tree (Firmiana simplex (L.) W.Wight) and
Chinese tulip poplar (Liriodendron chinense) transverse sections as eco-friendly, porous, sound-absorbing materials, J. Porous Mater. 29 (6) (2022) 1791–1796,
https://doi.org/10.1007/s10934-022-01289-z.
[57] R. Bartolini, S. Filippozzi, E. Princi, C. Schenone, S. Vicini, Acoustic and mechanical properties of expanded clay granulates consolidated by epoxy resin, Appl.
Clay Sci. 48 (3) (2010) 460–465, https://doi.org/10.1016/j.clay.2010.02.007.
[58] H.K. Kim, H.K. Lee, Influence of cement flow and aggregate type on the mechanical and acoustic characteristics of porous concrete, Appl. Acoust. 71 (7) (2010)
607–615, https://doi.org/10.1016/j.apacoust.2010.02.001.
[59] N. Hannan, S. Shahidan, Z. Maarof, N. Ali, S.R. Abdullah, M.H.W. Ibrahim, Sound absorption coefficient of coal bottom ash concrete for railway application,
Global Congress on Construction, Material and Structural Engineering (GCoMSE), Johor Bahru, MALAYSIA, 2017. 〈https://doi.org/10.1088/1757-899x/271/1/
012077〉.
[60] M. Pereira, J. Carbajo, L. Godinho, J. Ramis, P. Amado-Mendes, Improving the sound absorption behaviour of porous concrete using embedded resonant
structures, J. Build. Eng. 35 (2021), 102015, https://doi.org/10.1016/j.jobe.2020.102015.

15