Numerical Study on Scale Effect of Repetitive Plate-Loading Test
Round 1
Reviewer 1 Report
In this article, the scale effect of 50 and 300mm loading plate was studied. This study based on the experimental study of Jung ae al. 2018. The purpose of this paper was to establish the link between the 50mm and 300mm loading plate test results. The analysis done in this manuscript covered the simulation performed with Plaxis3d software. The simulation was performed on the homogeneous subgrade. The results of the numerical analysis were compared with the field tests. Below are my comments to this manuscript.
1. The literature review is broad but there are not too many references which can sustain the statements.
2. The adhesion term in this manuscript is the soil cohesion? If yes, please change this word to cohesion.
3. Table 1 is unreadable. Please consider the change of its editing.
4. The soil parameters need to be explained more carefully. The direct shear test was performed for all soil types? The shear surface in this test for gravelly soils will be artificial and the friction angle may by over calculated.
5. What was the name of the soils, the gradation curves need to be provided?
6. The dilatancy angle and Poisson ratio are quite venerable parameters which can greatly impact on the numerical analysis results.
7. The subgrade model has a 2m depth. Can you explain in the manuscript why? This thickness may be too low for the 300mm plate test simulation.
8. Table 2 – Elastic modulus. How was it assigned to soil types?
9. The manuscript needs editing. There are a lot of blank spaces.
10. Ev2 – The “v2” part should be in subscript.
11. Figure 8 is not clear. The points are too thick.
12. The subgrade type must be explained before the numerical analysis. Why this type of subgrade? Is field test presented in the previous article was performed on the same type of soil as for example this presented in Eq.3?
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The analyzed article describes the scale effect of different size of the loading plate test for testing layers of unbound subbase and base material of the road.
Below I present some comments and remarks to the manuscript.
In my opinion, the title of the article is misleading. The authors should explain why they use the term “cyclic loading test”, and thoroughly define what means “cyclic loading test”.
It is also not clear in the introduction and in theoretical background. Table 1 and figure 1 need explanations.
Mentioned in table 1 German code DIN 18134 describes static loading test, not cyclic. It is plate loading by dead load and settlement is measured. The procedure of the test includes loading, unloading and repeated loading, probably presented because of the test in figure 5. In result, we calculate primary strain modulus (during loading) and secondary strain modulus (during reloading.
Cyclic plate loading test is not known for me, probably it is a repeated test. In Europe exists Static loading test or Falling weight load test called dynamic loading test. In the result of Falling weight load test, the dynamic strain modulus correlated with secondary strain modulus from static loading test is obtained. In that test, the settlement at the end of the third impact of the hammer is measured.
Term “cyclic” is associated with frequency. I cannot find the results of such tests.
In table 3 there are results of static loading plate but not cyclic plate loading field-test results.
In this paper, authors use statements: cyclic plate loading test, loading plate test, and these terms appear in the text. In my opinion, this should be ordered. The same problem is with tables and figures.
The description of the tested material is not sufficient.
There is not a grain size distribution curve of tested material, gravel, sand, clay. It is not possible to discuss and calculate the coefficients of uniformity of different materials. The results of strain moduli should be recognized in terms of the characteristic of the tested material.
In table 2 some geotechnical parameters of materials are presented. Why Unit density of sand is so high, the same value like for gravel. This are different soils, gradation curves of soils could be useful to understand the parameters. The values of Elastic modulus seems to be strange. 200 Mpa for gravel is ok, for coarse sand also. But for Medium sand and especially for fine sand they are too high. With the value less than 80 Mpa is very hard to discuss, they are to low. For cohesive soils range from 10 to 200 MPa is an abstraction. 200 Mpa for cohesive material is really not possible, 120 MPa for this material means that is very stiff. 10 MPa means that the material is soft. These parameters are not discussed in a manuscript. Why strength parameters were obtained from the direct-shear test? The description of the diameter of tested boxes are not known, are they the same for coarse and fine material? More convenient are triaxial tests in compression conditions. The values of Elastic moduli should be determined within the proper level of deformation, not known for me from the manuscript. Why Poisson’s ratio is the same for coarse and cohesive soils?
We know not too much about the size of particles or grains of the tested material. From literature is obvious that the maximum size of grain should be less 5 to 10 times that size of equipment, in that case of the diameter of the plate. How to compare it with the size of the 50 mm of loading plate.
In conclusion, is very difficult to review the manuscript in the presented version. To do that as best as possible I expect from authors reaction and reference to my comments.
Moreover, the literature review is very tide and need to be increased. The authors of the manuscript should be more open on other scientists, especially on new publications on this subject in foreign literature.
It is not a good idea to refer almost only to your own achievements.
But generally, I find the article interesting and worth further processing.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The paper is generally interesting however it should almost rewire well, considering following emails. When you go through the paper, you will be lost in some places and you will lose the concept which shows that the paper is coherent. Besides, the English should revise by a technical native English speaker to fix lots of grammatical and chronological problems.
1) The title is quite confusing. “loading-plate scale effect of the cyclic plate loading test” doesn’t make sense. It’s recommended to revise it to simple ways, like “A numerical study on the cyclic plate loading test scale effect” or any other simple title.
2) There is a lack of relevant findings in the introduction and theoretical background which should describe more the paper topic. It would be better to reference the following article as well:
a. Abu-Farsakh, M., Hanandeh, Sh., Mohammad, L., Chen, Q., Performance of geosynthetic reinforced/stabilized paved roads built over soft soil under cyclic plate loads, Geotextiles and Geomembranes, Volume 44, Issue 6, December 2016, Pages 845-853,
b. Khalaj, O., Ghotbi Siabil, S.M.A., Moghaddas Tafreshi, S.N., Jirkova, H., “Performance Evaluation of Pavements Constructed on EPS Geofoam Backfill using Repeated Plate Load Test”, World Multidisciplinary Earth Sciences Symposium (WMESS 2018), Prague, Czech Republic, 3-7 September 2018
c. Linb, P., Tang, L., Niad, P., Field evaluation of subgrade soils under dynamic loads using orthogonal earth pressure transducers, Soil Dynamics and Earthquake Engineering, Volume 121, June 2019, Pages 12-24, https://doi.org/10.1016/j.soildyn.2019.03.001
d. Khalaj, O., Moghaddas Tafreshi, S.N., Mašek, B., Dawson, A.R., “Improvement of pavement foundation response with multi-layers of geocell reinforcement: Cyclic plate load test”, Geomechanics and Engineering, 9 (2015) 3, 373-395, DOI: 10.12989/gae.2015.9.3.373
e. Moghaddas Tafreshi, S.N., Khalaj, O., Dawson, A.R., “Repeated loading of soil containing granulated rubber and multiple geocell layers”, Geotextile and Geomembranes, 42 (2014) 1, 25–38, DOI: 10.1016/j.geotexmem.2013.12.003
f. Zhang, Ch., Jiang, G., Buzzi, O., Su, L., Full-scale model testing on the dynamic behaviour of weathered red mudstone subgrade under railway cyclic loading, Soils and Foundations, Volume 59, Issue 2, April 2019, Pages 296-315, https://doi.org/10.1016/j.sandf.2018.11.007
g. Moghaddas Tafreshi, S.N., Khalaj, O., Dawson, A.R., “Pilot-scale load tests of a combined multi-layered geocell and rubber-reinforced foundation”, Geosynthetics International, 20 (2013) 3, 143-161, DOI: 10.1680/gein.13.00008
3) Section 2 (theoretical background), specially 2.1 and 2.2 could be shortened because this information is clear and well known by the other researchers or scientists. It would be enough just to point out the main idea and assumption of each previous which would be related to the present study.
4) (Line 41) ASTM D 1195 indicates that loading plate shouldn’t be less than in 25.4 mm in thickness, and the diameter threshold of loading plate should be between 152 mm to 762 mm. Keep this ins in mind, how the authors statement could justify this contradiction?
5) More clarification needs in lines 45 to 47.
6) Figure 2 (in numerical analysis), generated meshes have an undeniable role in output results. In figure 2, the dimensions of the generated meshes show about 1 meter in dimensions. Sensing the role of mesh dimensions on the output results should have been considered on Authors work.
7) Verification of numerical model by some experimental test is one of the main parts in numerical analysis. I couldn’t find and verification on Authors numerical model. Nonattendance of these verifications could cause doubt on their results since in many parts they don’t have exact reasons for discrepancies between the experimental results and numerical ones.
8) There is no description or clarification in figure 5, what’s the reason for a reduction in settlements of loading surface by decreasing the loading-plate width?
9) In part 4.2, there is no information about the soil types in experimental analyses which could make it more practical.
10) Results from equations 8,9,10 and 11 aren’t the same as equation 7. Would Authors clarify the reason for these discrepancies? How Authors could justify that this trend doesn’t have root in inappropriate numerical analyses?
11) In numerical analyses, Authors mentioned that versatile elastic modulus implemented. In numerical analyses of cohesive soil and sand, by considering the Mohr–Coulomb elasto–plastic model, do Authors believe that changing the only parameter of elastic modulus is sensible? In a line of this fact, how they could justify the discrepancies that have been occurred in Table 7 which have no root in their numerical analyses?
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
Dear Authors,
The problems of the plate loading and cyclic plate loading tests are discussed in the paper. The Authors investigate influence of diameter of the loading plate on the strain modulus for different subgrade conditions. Finally, the Authors propose the values of the scale factor for investigated cases. However, it should be noted, that the values and conclusions given in the paper are derived only from the numerical analyses. Because of this, further verification of the dependencies by the experimental tests is necessary.
Minor and major remarks:
1. Extensive editing of English language and style required
2. I propose to extend the literature review and complete citations in the text.
3. I propose to use italic font for variables
4. Line 121: “The cyclic plate loading test is being commonly used in Europe, especially in Germany.” – Please give reference to the Standards if it is possible.
5. Equation 4 – generally “x” are used for cross (vector) product. I propose to remove such symbols or use dots. Please check, and correct.
6. Lines 129-131: “where r is the radius of the loading plate (mm), Δs is the difference in the settlement amount between the points with 30% and 70% of the maximum stress, and Δσ is the difference in the stress between the points with 30% and 70% of the maximum stress.” – The last phrase is repeated twice.
7. Figure 1: There is a lot of information in the form of comments. Moreover small font is used. Because of this, it is difficult to read. This can be rather used as the graphical abstract. In my opinion the contents of the study should be described in the text. It can be supported by such illustration but it should be significantly shorten.
8. Figure 2: Both pictures are the same, and this fact is given in the text. Much more interesting will be presentation, instead fig2b, pictures of mesh sizes in the normal directions to the contact areas for both sizes of loading plates.
9. Line 223: “Table 2 lists the summary of the analysis cases and applied properties. For the dilatancy angle of the embankment roadbed, the value ψ = φ – 30° proposed by Bolton [16] among the input properties was used.” - What was the reason of application of the above formula?
10. Line 283: “where Ev2,300mm is the strain modulus measured using the 300-mm loading plate and Ev2,50mm” - indexes should be corrected. The same mistake is in line 457
11. Section 4.1: Significant differences for landing plates with diameters 50mm and 300mm are observed in the experimental tests. This can be explained by influence of materials with large particle size. Such experimental tests are compared and investigated with numerical analyses. Were the numerical analyses performed for homogeneous subgrade? Was the subgrade modeled with materials with different particle sizes?
12. Line 339 “where Ev2,d is the secondary strain modulus when the loading plate diameter is d and d is the 339 loading-plate diameter (mm).” – Description should be corrected.
13. Line 364: there is incorrect table in the text. It must be corrected.
Kind Regards,
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 4 Report
Dear Authors,
Remark 1: Verification of the model was made by comparison of FE results with the results of the existing in the literature field-verification tests. However, it is difficult to recognize the regions (A-Q) of field-tests with different elastic modulus E (10-200 MPa) used in numerical analyses and deduce concluding remarks concerning verification of the FE model. There are large differences of Ev1 for 50mm and 300mm loading plate. Moreover, different trends of the strain modulus of the first loading curve Ev1for field-test and FE results are obtained. How such large differences of Ev1 and the assumed method of homogenization influence further analyses of Ev2 and scale factors?
Remark 2: I have no experience with PLAXIS 3D software, but basing on the experience of solving contact problems in other Finite Element Software it can be stated, that in order to obtain acceptable accuracy of solution of contact problems, the model requires using of regular hexahedral elements in contact area and not less than three element in contact per radius. Assuming that I correctly understand figure 2, then both statement are not fulfilled in the FE model. Moreover, the mesh was generated automatically by a solver and loading (rigid?) plate was modeled with very high stiffness. This may results in incorrect distribution of contact stresses at the edge of the loading plate. Due to these facts there are some questions:
Q1:Was the influence of the above problems studied?
Q2: How distribution of contact stresses looks like?
Q3: Did the number of contact elements used in all analyses with different diameters of the loading plates were the same? If yes, how the number of elements in contact area influences obtained results and value of scale factor?
Due to the both above remarks, I am still convinced that experimental verification of the proposed methodology and FE model should be carried before publication of the paper.
Remark 3:Results in Fig 8 are not readable. Different colors should be used in diagrams.
Remark 4: Line 336: grammar mistake “where Ev2,d is the secondary strain modulus when the loading plate diameter is d and d is the loading-plate diameter (mm).”
Kind Regards
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 3
Reviewer 4 Report
Dear Authors,
Thank you for the detailed response.
I have one important remark, which was discussed in the Author response (point 1 of 2nd review), but it was not given in the paper. It concerns lines 256-259:" On the other hand, in the numerical analysis results, the results of the 50-mm loading plate were smaller than those of the 300-mm loading plate, and the deviation was not large for both Ev1 and Ev2. The value of Ev2/Ev1 was also 2.0 for both loading plates, and the correlation coefficient of Ev2 was 0.999, which was higher than that in the field test."
This effect, as well as, the reason of differences in Ev1 (between numerical and experimental tests) should be explained in detail in the paper.
Summary of the paper:
The influence of the scale effect on repetitive plate loading and test are discussed in the paper. The Authors investigate influence of diameter of the loading plate on the strain modulus for different subgrade conditions. Finally, on the basis of the numerical analyses, the values of the scale factors for investigated cases were proposed. The proposed method is interesting, however due to significant differences between numerical and experimental results and assumptions (such as homogeneous ground) in the numerical model it requires further experimental verification.
Kind Regards
Author Response
Please see the attachment.
Author Response File: Author Response.docx
