Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessArticle

Peer-Review Record

Centrifugal Model Test and Simulation of Geogrid Reinforced Backfill and EPS Interlayer on Bridge Abutment

Sustainability 2022, 14(21), 13908; https://doi.org/10.3390/su142113908

by Zheng Shen^1,* and Lei Fang²

Reviewer 1: Anonymous

Reviewer 2:

Jiayao Chen

Reviewer 3:

De Sena Monteiro Ozelim Luan Carlos

Sustainability 2022, 14(21), 13908; https://doi.org/10.3390/su142113908

Submission received: 26 August 2022 / Revised: 12 October 2022 / Accepted: 24 October 2022 / Published: 26 October 2022

(This article belongs to the Special Issue Resilience and Sustainability in Geotechnical Engineering)

Round 1

Reviewer 1 Report

minor revision

1-whats the effect of geogrid on the backfill

2-whats the effect of Reinforcement on the strength of backfill material

3-whats the effect of Reinforcement on the lateral earth pressure acting on the back of abutment

4- whats the effect of buried depth on the tensile strain of reinforced material

5-whats the effect of EPS on the lateral displacement of the backfill material

6-check the units of all tables.

7-remove "the" in subscription of all figures.

Author Response

Response to Reviewer 1

[General Comment] minor revision:

Response: Thank you very much. We have made considerable efforts to revise the manuscript and try to eliminate these serious flaws.

Some revisions for the authors to consider:

[Comment 1] What^’s the effect of geogrid on the backfill.

Response: Thanks for your question. We replied to your questions as follows [Pg2, Ln59-62]:

“Especially for geogrids, the mutual friction between the geogrid and the soil and the special interlocking effects of the geogrid mesh limit the lateral deformation of the upper and lower soil, improve the shear strength of the reinforced soil, and enhance the stability of the soil.”

[Comment 2] What^，s the effect of Reinforcement on the strength of backfill material.

Response: Thanks for your question. We replied to your questions as follows [Pg2, Ln57-59]:

“Geosynthetic reinforcement exerts its tensile strength, and restrains the lateral deformation of soil by the friction between reinforcement and soil, so as to improve the bearing capacity and shear strength of soil.”

[Comment 3] What^’s the effect of Reinforcement on the lateral earth pressure acting on the back of abutment.

Response: Thanks for your question. We replied to your questions as follows [Pg3, Ln94-103]:

“It can be drawn a conclusion from the above research results, the frictional resistance between the geosynthetics reinforcement and the soil restricts the lateral deformation of the soil and reduces the pressure of the soil on the wall. At the same time, due to the role of the reinforcement, the wall panel and the soil into a whole, thereby increasing the friction between the wall back and the soil, resulting in a corresponding reduction in soil lateral pressure coefficient.”

[Comment 4] What^’s the effect of buried depth on the tensile strain of reinforced material.

Response: Thanks for your question. We replied to your questions as follows [Pg4, Ln123-127]:

“It can be concluded from the above test results that the strain of the reinforcement increases with the increase of the buried depth of the geogrid from the top of the wall, but the increase rate gradually decreases.”

[Comment 5] What^’s the effect of EPS on the lateral displacement of the backfill material.

Response: Thanks for your question. We replied to your questions as follows [Pg5, Ln177-178]:

“The above research results show that embedding the EPS plate behind the abutment can effectively reduce the lateral displacement of the fill behind the abutment.”

[Comment 6] Check the units of all tables.

Response: Thanks for your kind reminders. We have checked the units of all tables.

[Comment 7] Remove "the" in subscription of all figures.

Response: Thanks for your kind reminders. We replied to your questions as follows:

We have removed "the" in subscription of Fig.13. in Ln518.

Reviewer 2 Report

In general, this manuscript is well organized and the corresponding topic is interesting also important. Also, some issues have been raised and then solved in this study. Several comments as follows from which the author can be benefit:

1. The meaning of the abbreviation EPS has never been introduced. Instead, it needs to be explained in the first occurrence.

2. Figures 2 and 5: it is suggested to add the real pictures of centrifugal model test and the established diagram of numerical simulation in the corresponding figures.

3. Page 5: why can numerical calculation be used as a verification method for this study? Please declare the reason.

4. “Calculation results” in the manuscript is incorrect. It is suggested to change them to "numerical results"

5. The Introduction section can be optimized, e.g., add the reason for using centrifugal model test rather than full-scale model test. Also, the following references maybe benefit for stability analysis:

(1) Chen, J., Zhang, D., Huang, H., Shadabfar, M., Zhou, M., & Yang, T. (2020). Image-based segmentation and quantification of weak interlayers in rock tunnel face via deep learning. Automation in Construction, 120.

(2) Chen, J., Zhou, M., Huang, H., Zhang, D., & Peng, Z. (2021). Automated extraction and evaluation of fracture trace maps from rock tunnel face images via deep learning. International Journal of Rock Mechanics and Mining Sciences, 142.

6. The conclusion section can be streamlined since there are too many entries now – resulting in few key points can be grasped.

Author Response

Response to Reviewer 2

[General Comment] In general, this manuscript is well organized and the corresponding topic is interesting also important. Also, some issues have been raised and then solved in this study. Several comments as follows from which the author can be benefit:

Response: Thank you very much.

Some minor revisions for the authors to consider:

[Comment 1] The meaning of the abbreviation EPS has never been introduced. Instead, it needs to be explained in the first occurrence.

Response: Thanks for your kind reminders. We have added an explanation "(expanded polystyrene)" of EPS in Ln149.

[Comment 2] Figures 2 and 5: it is suggested to add the real pictures of centrifugal model test and the established diagram of numerical simulation in the corresponding figures.

Response: Thanks for your kind reminders. We have revised some figures as suggestion. Unfortunately, we forgot to take photos of the real model test during the centrifuge model test, but added photos of the detailed components of the centrifuge used for the test. In addition, the numerical simulation diagram of pile supported abutment is established in Fig.5.

[Comment 3] Page 5: why can numerical calculation be used as a verification method for this study? Please declare the reason.

Response: Thanks for your question. We replied to your questions as follows [Pg6, Ln200-203]:

"By comparing the numerical analysis method with the centrifugal model test results, the validity of the centrifugal model test results can be verified, and the reliability of the numerical method can be verified. Moreover, large-scale centrifuge model tests are expensive, and the use of numerical methods can reduce the cost of engineering design."

[Comment 4]“Calculation results” in the manuscript is incorrect. It is suggested to change them to "numerical results".

Response: Thanks for your kind reminders. We have made revisions accordingly.

[Comment 5] The Introduction section can be optimized, e.g., add the reason for using centrifugal model test rather than full-scale model test. Also, the following references maybe benefit for stability analysis:

Response: Thanks for your kind reminders. We replied to your questions as follows:

We have added the reason for using centrifugal model test rather than full-scale model test in Ln179-Ln181. At the same time, I also have cited the above two references.

[Comment 6] The conclusion section can be streamlined since there are too many entries now – resulting in few key points can be grasped.

Response: Thanks for your kind reminders. We have revised the conclusion as follows [Pg32-33, Ln578-602]:

Geogrid reinforcement decreases the surface settlement of different backfills and reduces the lateral earth pressure acting on the back of abutment.
The reinforcement effect of geosynthetics behind the abutment can be improved by increasing the length of bottom reinforced material.
Centrifuge model tests show that soil pressure of abutment with EPS inclusion is reduced markedly, and the settlement of backfill without geogrid is increased.
The earth pressure on the back of the abutment with both geogrid reinforcement and EPS inclusion is further reduced, and the differential settlement between the abutment and the backfill becomes smaller, and the tensile stress of the geogrid is obviously increased, which is very beneficial to the effect of geogrid reinforcement.

Reviewer 3 Report

The paper presents both centrifuge and FE models of geogrid reinforced backfills and EPS interlayers used on bridge abutments.

Overall, the paper clearly has merit, but major issues must be addressed prior to its acceptance for publication.

General Remarks:

- At first, it is important to highlight that the English level needs to be improved. We advise the authors to seek proofreading services.

- Citations are weird, with initials in the middle of the text. This should be corrected.

Specific Remarks:

The following topics must be addressed:

Abstract

- There is a strange sentence: "The reinforcement improved the strength of the backfill material, thereby reducing the lateral earth pressure of the backfill on the abutment back". Does this really make sense (is the strength improvement responsible for the reduced lateral earth pressure? What about improvements on deformability parameters? How are they related?

Introduction

- There is a strange expression: "phenomenon of emptying."

- The following sentence seems to be lost and unreferenced: "According to the analysis of construction cost, the cost of traditional reinforced concrete abutment may be five times that of optimized reinforced soil abutment”. Why authors support this idea?

- There is a strange sentence "Centrifugal geotechnical model test was to place the model in a special centrifuge and use a 1/n scale model to test in the ng centrifugal acceleration field"

- In my opinion, the authors did not indicate the research questions clearly: ok, there is a lack of centrifuge tests, but what issue was still unclear? Why this study is important? More references are needed to clarify why the work is important.

Centrifugal Model Test design

- What is "SCR electrodeless speed control"?

- Apparently, geogrid mesh size was not considered in the scale studies. What are the possible problems with the interlocking behavior in the correspondent “full size” model?

- Model scale in the abstract is indicated to be 1/20, but Fig.2 and 3 suggest different values. Please clarify.

- Please discuss the consolidation effect on the scale considerations, if any.

- Why no tests were performed for Geogrid+EPS for Clay soil? Why not the same variants for each soil type?

Numerical simulation of the abutment backfill

- Apparently, the geometric model is not "the same of the physical model", but rather that of the correspondent 1/1 model. Please clarify.

- Need to show steps of the simulation. Did it have a geostatic step? What types of finite elements were used? Are there any specific steps were there are geometry changes in the model? What about the mesh? Did the authors perform mesh convergence tests?

- Where did the properties used for each material come from?

- How was the contact modeled? Friction-penalty for tangential behavior? Normal-hard contact? What about the EPS contact? And the contact between geogrid and the rest? This need to be clear in the paper.

- For the 3D model (pile-supported), how were the boundary effects studied? Was this the optimal model size?

- Apparently, there are no "pile" elements in Abaqus. This should be clarified.

Analysis of centrifugal model test results and simulation results

- Figures comparing numerical and experimental results need to have the same scales; x,y ranges and tick marks.

- Are not GA1st and GA2rd (which should be GA2nd), changed? The funnel-shaped curve would be GA1st, no?

- How were these inclinations calculated for 4.1?

- It should probably be GA3rd instead of GA3th.

- The description of the models in 4.2 does not match Fig. 9. See GA3th, for example, which in the text is indicated to have an EPS interlayer, while in the plot seems to be the untreated one. Others also have problems.

- Results from the centrifuge models considerably differ from the FE models in this case. Why? The behavior is different (decreasing settlement in FE model and increasing in the centrifuge?)

- Maybe a graphical representation to Table 5 would make the explanations easier to follow.

- Again, the description of the models in 4.3 does not match Fig. 11.

- Maybe a graphical representation to Table 7 would make the explanations easier to follow.

- Why not compare in Figure 13 case of +0.2m without EPS as well?

Author Response

Response to Reviewer 3

[General Comment] The paper presents both centrifuge and FE models of geogrid reinforced backfills and EPS interlayers used on bridge abutments. Overall, the paper clearly has merit, but major issues must be addressed prior to its acceptance for publication.

General Remarks:

At first, it is important to highlight that the English level needs to be improved. We advise the authors to seek proofreading services. Citations are weird, with initials in the middle of the text. This should be corrected.

Response: Thank you very much.

Some minor revisions for the authors to consider:

[Comment 1] Abstract

Response: Thanks for your kind reminders. We have made revisions of the manuscript [Pg3, Ln102-103] and reply to the question as following:

The first question “Does this really make sense (is the strength improvement responsible for the reduced lateral earth pressure? "

My reply is that the increase of strength has nothing to do with the decrease of lateral earth pressure. It should be that the friction interaction between geosynthetics and backfill materials is the reason for reducing the lateral earth pressure.

The second question “What about improvements on deformability parameters?"

My reply is that the reinforced material increases the lateral modulus of soil and limits the lateral deformation of soil.

The third question “ How are they related?

My reply is that the reinforced material improves the lateral modulus of soil, forms soil arch effect between each layer of reinforced material, and reduces the horizontal lateral pressure of soil.

[Comment 2] Introduction

- There is a strange expression: "phenomenon of emptying."

- There is a strange sentence "Centrifugal geotechnical model test was to place the model in a special centrifuge and use a 1/n scale model to test in the ng centrifugal acceleration field"

Response: Thanks for your kind reminders. We have made revisions of the manuscript and reply to the question as following:

The first question “There is a strange expression: "phenomenon of emptying."

My reply is that It is wrong with emptying, It should be bumping.

The second question “The following sentence seems to be lost and unreferenced: "According to the analysis of construction cost, the cost of traditional reinforced concrete abutment may be five times that of optimized reinforced soil abutment”. Why authors support this idea?”

My reply is that compared with the reinforced concrete retaining wall, the reinforced earth retaining wall is easy to construct, occupies a small area, uses few materials and has low price, and the overall project cost is very cheap. I support the above idea. In order to reduce the complexity of the citation, I cited the cost analysis results in the paper, the specific cost analysis process is not given.

The third question “There is a strange sentence "Centrifugal geotechnical model test was to place the model in a special centrifuge and use a 1/n scale model to test in the ng centrifugal acceleration field".

My reply is that in a centrifuge, the small-scale model is subjected to an acceleration field n times greater than earth’ s gravity acceleration constant, g. Thus, each soil particle in the model weights n times more. As a result, the gradient of body stress within the small-scale model will now be similar to the prototype n times larger, ensuring similarity of effective stress at equivalent depth in model and prototype. This implies that a 1/n scale model will behave like its full-scale prototype if subjected to a centrifugal acceleration of n times g. Thus, gravity-dependant processes are correctly reproduced. Stress and strain in the prototype soil

mass are preserved, and the strain–stress curve of the model is identical to one of the prototype.

The fourth question “In my opinion, the authors did not indicate the research questions clearly: ok, there is a lack of centrifuge tests, but what issue was still unclear? Why this study is important? More references are needed to clarify why the work is important."

My reply is that i: the research issue is how to evaluate the benefits of geogrids and EPS interlayer to mitigate the lateral earth pressures behind the abutment and backfill surface settlement. ii: The reason why this study is very important is that there are two problems in the integral abutment, including excessive backfill surface settlement and high lateral earth pressures behind the abutment, which can cause abutment movement and bridge-head jumping, and we need to take measures to solve these two problems.

[Comment 3] Centrifugal Model Test design

- What is "SCR electrodeless speed control"?

- Apparently, geogrid mesh size was not considered in the scale studies. What are the possible problems with the interlocking behavior in the correspondent “full size” model?

- Model scale in the abstract is indicated to be 1/20, but Fig.2 and 3 suggest different values. Please clarify.

- Please discuss the consolidation effect on the scale considerations, if any.

- Why no tests were performed for Geogrid+EPS for Clay soil? Why not the same variants for each soil type?

Response: Thanks for your kind reminders. We have made revisions of the manuscript and reply to the question as following:

The first question “What is "SCR electrodeless speed control"?

My reply is that SCR refers to Silicon controlled rectifier，which is a new device of automation. SCR electrodeless speed control refers to the control mode for rotary loading of centrifuges.

The second question “Apparently, geogrid mesh size was not considered in the scale studies. What are the possible problems with the interlocking behavior in the correspondent “full size” model?"

My reply is that the characteristics of geogrid-soil interface are obviously affected by the mesh size of geogrid. For standard sand, the geogrid with larger mesh size has higher interface strength parameters and more obvious friction reinforcement effect (Wang et al., 2015). For coarse-grained soils, the greater the ratio of the average particle size of the filler to the mesh size of the grid, the more obvious the interlocking effect, and the greater the peak shear stress and residual shear stress at the reinforced soil interface (Tang et al. 2017). In this study, because the selected backfill materials are clay and aeolian sand, the particle radius of these two materials is small. According to the existing research results, the mesh size of geogrid is 8mm * 8cm.

[1] Wang Yongfu, Zhang Ruiyuan, Tang Xiaosong, & Zheng Yingren. ( 2015 ). The effect of mesh size on the interface characteristics of geogrid-noncohesive soil. Journal of Logistics Engineering College, 31 ( 5 ), 6-10. (In Chinese)

[2] Tang Xiaosong, Zheng Yingren, Wang Yongfu, & Feng Yushi. ( 2017 ). Research on reasonable mesh size of geogrid. Geotechnical mechanics, 38 ( 6 ), 6-11. (In Chinese)

The third question “Model scale in the abstract is indicated to be 1/20, but Fig.2 and 3 suggest different values. Please clarify."

My reply is that it is wrong with model scale of 1/20, and correct model scales were 62.5, 40 for gravity abutment and pile-supported abutment respectively. I have revised it in Ln10-11.

The fourth question “Please discuss the consolidation effect on the scale considerations, if any." My reply is that the time similarity scale between the centrifuge model and the prototype can be derived from the dimensionless time factor T_v in Terzaghi 's one-dimensional saturated soil consolidation equation.

（1）

In the formula (1), c_v, d and t are the consolidation coefficient, drainage path length and consolidation duration of soil, respectively. It is required that the dimensionless physical quantity T_v in the centrifugal model and the prototype must be equal, that is :

（2）

In the formula (2), t_m, t_pare consolidation duration of model soil and prototype soil, respectively. d_m, d_p are drainage path length of model soil and prototype soil; H_m,H_p are thickness of model soil and prototype soil; N is the ratio of inertial acceleration of model to prototype.

The fifth question “ Why no tests were performed for Geogrid+EPS for Clay soil? Why not the same variants for each soil type?"

My reply is that i:because the test project is located in Inner Mongolia, China, where there are many deserts, aeolian sand is the main local filler. In addition, the effect of geogrid reinforced cohesive soil is weaker than that of aeolian sand. Considering the economy of the project, it is more inclined to use aeolian sand as backfill behind abutment in practical engineering, so the EPS interlayer is not set for the clay backfill material behind the abutment. ii: The whole test design idea is to study the reinforcement of the backfill material first, and then consider whether to set the EPS interlayer according to the reinforcement effect. Therefore, the test parameters of each backfill material are not the same.

[Comment 4] Numerical simulation of the abutment backfill

- Apparently, the geometric model is not "the same of the physical model", but rather that of the correspondent 1/1 model. Please clarify.

- Where did the properties used for each material come from?

- For the 3D model (pile-supported), how were the boundary effects studied? Was this the optimal model size?

- Apparently, there are no "pile" elements in Abaqus. This should be clarified.

Response: Thanks for your kind reminders. We have made revisions of the manuscript and reply to the question as following:

The first question “Apparently, the geometric model is not "the same of the physical model", but rather that of the correspondent 1/1 model. Please clarify."

My reply is that the physical model should be the prototype model, in the paper, I have revised it in Ln308-310.

The second question “ Need to show steps of the simulation. Did it have a geostatic step? What types of finite elements were used? Are there any specific steps were there are geometry changes in the model? What about the mesh? Did the authors perform mesh convergence tests?"

i: “ Need to show steps of the simulation".

My reply is that steps of the simulation have been given in the paper in Ln296-300.

ii: “Did it have a geostatic step?".

My reply is that it have a geostatic step in the process of numerical simulation, and an analysis step is created in the Step function module and set the type to Geostatic. In the Load function module, the gravity load for the whole soil mass (Geostatic type) us defined.

iii: “What types of finite elements were used?"

My reply is that the type of finite element in this paper is mechanical element.

ⅳ: “Are there any specific steps were there are geometry changes in the model?"

My reply is that ' Model-1 → Parts → Part-1 → Features → Solid extrude → Section Sketch ' in the tree diagram on the left side of Abaqus is found. Different models will have slightly different Part names.In short, find the sketch options corresponding to the modified Part, double-click or right-click ' Edit ' to change the geometry in the model.

ⅴ: “What about the mesh?"

My reply is that since the abutment is a typical plane strain problem, the plane strain model is used in the modeling calculation. Because there is no large non-uniform deformation of backfill soil, the high-order plane strain 8-node shear-shrinkage integral element CPE8R is used to divide the soil element mesh without considering the influence of mesh distortion. Because the compression of abutment wall and soil foundation is very small, the grid division of abutment wall and soil foundation adopts plane strain 8-node shear-compression integral hybrid element CPE8RH.

ⅵ: “Did the authors perform mesh convergence tests?"

My reply is that we have performed mesh convergence tests in the process of numerical simulation. In the calculation of abutment in this paper, there are several typical contact problems, including the contact between abutment back and filling behind abutment, the contact between abutment structure and soil foundation,the contact between the filler and the soil base and the contact between the reinforcement and the filler. The contact algorithm of ABAQUS is based on the Newton-Raphson iterative technique. ABAQUS checks the state of all contact pairs before each increment starts to determine whether the slave node is open or closed. If the node is closed, it needs to be determined whether it is slipping or sticking. ABAQUS imposes constraints on closed joints, and contact constraints on joints from closed to disconnected ; then ABAQUS iterates and calculates the correction to change the configuration of the model until the iterative results meet the convergence conditions.

The third question “ Where did the properties used for each material come from?"

My reply is that the material parameters are obtained through indoor soil test and other indoor tests, including density test, water content test, compaction test, triaxial test, consolidation compression test, direct shear test, tensile test of geogrid, etc..

The fourth question “ How was the contact modeled? Friction-penalty for tangential behavior? Normal-hard contact? What about the EPS contact? And the contact between geogrid and the rest? This need to be clear in the paper."

My reply is that we have added some descriptions in the paper. Here is a comprehensive answer to the above questions. In the calculation of abutment, there are several typical contact problems, including the contact between abutment back and backfill, the contact between abutment back and soil foundation, the contact between filler and soil foundation, the contact between EPS and abutment back and filler, and the contact between reinforcement and filler. In ABAQUS, * Contact Pair is used to simulate the face to face contact, which can take into account the extrusion, shear and displacement in the corresponding direction during the interaction between faces, and simulate the separation between faces. In order to avoid the transition between contact surfaces and the transition of nodes on the slave surface into the master surface, ABAQUS has set the parameter adjust to adjust the Overclosed Nodes. ABAQUS surface to surface contact is based on Coulomb friction theory with friction coefficient μ To represent the friction behavior between surfaces. When the contact shear stress is equal to or greater than the limit friction force μP, sliding occurs between contact surfaces. It is difficult to simulate the ideal friction behavior. ABAQUS uses the dotted line of the penalty friction formula of "elastic sliding" to approximate, and ABAQUS automatically selects the penalty stiffness (the slope of the dotted line).

The fifth question “For the 3D model (pile-supported), how were the boundary effects studied? Was this the optimal model size?"

My reply is that when the large finite element program ABAQUS is used to calculate the two types of abutments, the plane strain analysis model is established, and the 3D model (pile-supported) is cancelled, which is because the two-dimensional model has met the requirements of calculation.

The sixth question “Apparently, there are no "pile" elements in Abaqus. This should be clarified." My reply is that we have added pile elements in Abaqus for Fig.5.

[Comment 5] Analysis of centrifugal model test results and simulation results

- Figures comparing numerical and experimental results need to have the same scales; x,y ranges and tick marks.

- Are not GA1st and GA2rd (which should be GA2nd), changed? The funnel-shaped curve would be GA1st, no?

- How were these inclinations calculated for 4.1?

- It should probably be GA3rd instead of GA3th.

- Results from the centrifuge models considerably differ from the FE models in this case. Why? The behavior is different (decreasing settlement in FE model and increasing in the centrifuge?)

- Maybe a graphical representation to Table 5 would make the explanations easier to follow.

- Again, the description of the models in 4.3 does not match Fig. 11.

- Maybe a graphical representation to Table 7 would make the explanations easier to follow.

- Why not compare in Figure 13 case of +0.2m without EPS as well?

Response: Thanks for your kind reminders. We have made revisions of the manuscript and reply to the question as following:

The first question “Figures comparing numerical and experimental results need to have the same scales; x,y ranges and tick marks."

My reply is that I have revised the figures comparing numerical and experimental results with the same scales, x,y ranges and tick marks in Fig.7-Fig.9.

The second question “Are not GA1st and GA2rd (which should be GA2nd), changed? The funnel-shaped curve would be GA1st, no?"

i: “Are not GA1st and GA2rd (which should be GA2nd), changed?"

My reply is that GA1st and GA2rd should be changed each in the paper, in addition, it is wrong with ‘GA2rd’, and I have changed ‘GA2rd’ into ‘GA2nd’. All revisions are seen in Ln 386.

ii: “The funnel-shaped curve would be GA1st, no?"

My reply is that the funnel-shaped curve would be GA1st, and the saddle-shaped curve would be GA2nd. I have revised it in Ln 395-Ln401.

The third question “How were these inclinations calculated for 4.1?"

My reply is that the maximum inclination equal the differential settlement of backfill surface divided by length of bridgehead approach slab.

The fourth question “It should probably be GA3rd instead of GA3th."

My reply is that I have changed GA3th to GA3rd in Ln 411.

The fifth question “The description of the models in 4.2 does not match Fig. 9. See GA3th, for example, which in the text is indicated to have an EPS interlayer, while in the plot seems to be the untreated one. Others also have problems."

My reply is that it is wrong with the description of the models in 4.2, and we have made revision, and changed it to “The model GA3th had no treatment measures, and the model GA4th adopted grid reinforcement; The model GA5th contained EPS interlayer, and the model GA6th contains EPS interlayer and adopts geogrid reinforcement, as shown in Tab.1.”

The sixth question “Results from the centrifuge models considerably differ from the FE models in this case. Why? The behavior is different (decreasing settlement in FE model and increasing in the centrifuge?)"

i: “Results from the centrifuge models considerably differ from the FE models in this case. Why?"

My reply is that the main reason may be the constitutive models used for numerical analyses might not have accurately simulated the real behavior of materials including backfill, geogrid, and EPS.

ii: “The behavior is different (decreasing settlement in FE model and increasing in the centrifuge?)"

My reply is that it can be seen from Table 3 that when only EPS interlayer is set at the abutment, the settlement of filler surface increases compared with that without treatment measures, and the results of finite element calculation are slightly larger than those of centrifugal model test (2.87%>1.89%). When EPS interlayer is set at the abutment and geogrid is used for treatment, the settlement of filler surface is reduced compared with that without treatment measures, and the result of finite element calculation is more reduced than that of centrifugal model test (7.41%>1.47%), which may be that the constitutive models used for numerical analyses might not have accurately simulated the real behavior of materials including backfill, geogrid, and EPS..

The seventh question “Maybe a graphical representation to Table 5 would make the explanations easier to follow."

My reply is that I have changed Table 5 to Fig. 14 and Fig. 15.

The eighth question “Again, the description of the models in 4.3 does not match Fig. 11."

My reply is that it is wrong with the description of the models in 4.3, and we have made revision, and changed it to “The model GA7th had no treatment measures; the model GA8th adopted geogrid reinforcement; the model GA9th contained EPS interlayer and adopted grid reinforcement.”

The ninth question “Maybe a graphical representation to Table 7 would make the explanations easier to follow."

My reply is that I have changed Table 5 to Fig. 14 and Fig. 15.

The tenth question “Why not compare in Figure 13 case of +0.2m without EPS as well?"

My reply is that I have added the case of +0.2m without EPS, and compared with the case of +0.2m with EPS.

Round 2

Reviewer 2 Report

All the comments were revised by the authors. It is thus suggested to be accepted for further publishcation.

Author Response

Dear reviewer,

We have checked English language and style. Thank you very much!

Reviewer 3 Report

The authors have addressed the issues raised in the review report.

On the other hand, some of their answers were not incorporated into the manuscript.

In order to be accepted, I adivise to include such answers in the spots mentioned in the first review. After that, the paper can be accepted without the need for a new round of reviews.

Author Response

Response to Reviewer 3

General Remarks:

Response: Thank you very much.

Some minor revisions for the authors to consider:

[Comment 1] Abstract

Response: Thanks for your kind reminders. We have made revisions of the manuscript [Pg3, Ln102-103] and reply to the question as following:

The first question “Does this really make sense (is the strength improvement responsible for the reduced lateral earth pressure? "

The second question “What about improvements on deformability parameters?"

My reply is that the reinforced material increases the lateral modulus of soil and limits the lateral deformation of soil. We have made some revisions in the manuscript. [Pg2, Ln60-64]. The third question “ How are they related?

[Comment 2] Introduction

- There is a strange expression: "phenomenon of emptying."

- There is a strange sentence "Centrifugal geotechnical model test was to place the model in a special centrifuge and use a 1/n scale model to test in the ng centrifugal acceleration field"

Response: Thanks for your kind reminders. We have made revisions of the manuscript and reply to the question as following:

The first question “There is a strange expression: "phenomenon of emptying."

My reply is that It is wrong with emptying, It should be bumping. We have made a revision in the manuscript. [Pg1, Ln28].

My reply is that i: the research issue is how to evaluate the benefits of geogrids and EPS interlayer to mitigate the lateral earth pressures behind the abutment and backfill surface settlement. We have made some revisions in the manuscript. [Pg6, Ln211-214] ii: The reason why this study is very important is that there are two problems in the integral abutment, including excessive backfill surface settlement and high lateral earth pressures behind the abutment, which can cause abutment movement and bridge-head jumping, and we need to take measures to solve these two problems. We have made some revisions in the manuscript. [Pg2, Ln45-46]

[Comment 3] Centrifugal Model Test design

- What is "SCR electrodeless speed control"?

- Apparently, geogrid mesh size was not considered in the scale studies. What are the possible problems with the interlocking behavior in the correspondent “full size” model?

- Model scale in the abstract is indicated to be 1/20, but Fig.2 and 3 suggest different values. Please clarify.

- Please discuss the consolidation effect on the scale considerations, if any.

- Why no tests were performed for Geogrid+EPS for Clay soil? Why not the same variants for each soil type?

Response: Thanks for your kind reminders. We have made revisions of the manuscript and reply to the question as following:

The first question “What is "SCR electrodeless speed control"?

My reply is that SCR refers to Silicon controlled rectifier，which is a new device of automation. SCR electrodeless speed control refers to the control mode for rotary loading of centrifuges. We have made some revisions in the manuscript. [Pg7, Ln228-229]