Significance of Determination Methods on Shear Modulus Measurements of Fujian Sand in Cyclic Triaxial Testing

Round 1

Reviewer 1 Report

1. A smaller revision request to the scale bars in Fig. 2 to make it more visible.

2. L437, it seems this sentence does not make sense.

Author Response

The authors' responses to reviewer's comments, please see attached file "Paper_Response".

Author Response File: Author Response.docx

Reviewer 2 Report

The authors analyze the results of an extensive set of laboratory experiments on Fujian Sand.  The experiments are conducted as undrained cyclic triaxial compression tests (CT).  The complete set of experiments includes various values of the following factors:  confining pressure, the relative density, and the strain amplitude.  The paper applies several procedures for analyzing their laboratory testing data to find G and G_max.  G and the G/G_max relationship are important in applying the equivalent-linear method for predicting ground motions during seismic loading, so it is important to establish proper analysis procedures for determining G and G_max from laboratory data.  The paper investigates four analysis procedures for finding the shear modulus G.  The paper investigates two analysis procedures for finding the low-strain modulus G_max.

 

The authors conclude that the four procedures for finding G give different results, with a difference range of 20% to 30%.  The authors also conclude that the two procedures for finding G_max give different results, but the difference is relatively small.  The authors also find that both G and G_max depend upon the loading cycle that is used for finding these values.

 

The testing appears to have been soundly conducted; the analysis is correct; and the paper is well written.  The Reviewer recommends that the paper is declined for reasons of impact and presentation.

 

1. The paper compares four procedures for finding G from laboratory data.  The differences arise from two origins.  First, the stress-strain is asymmetric.  Second, the stress-strain is not linear (or even bilinear, allowing for the asymmetry).  The first of these origins is the reason for the Reviewer’s recommendation to decline on the basis of impact.  The asymmetry is due to the use of cyclic triaxial (CT) loading in the authors’ experiments, which causes the intermediate principal stress to alternate between the major and minor principal stresses (line 66).  This is an inherent weakness in the testing method.  When we use G to predict wave propagation and ground response, we are analyzing S-waves, in which soil deformation during the wave propagation occurs under cyclic simple shear (CSS) conditions.  Asymmetry does not occur in the field under CSS conditions, but symmetry occurs in the lab when CT tests are used. The important question is the following:  which averaging of CT data will give results closest to CSS conditions?  The paper does not address this question, it only concludes that the four methods give different results, which is not surprising.  The paper does not recommend one or another method for use with the equivalent-linear method.  Hence, the lack of impact.  On line 75, the authors say that “there is no consensus regarding the best method for determining the shear modulus of soils at large strains”.  The authors have left this question unanswered for CT tests.

2. The paper compares two methods for finding G_max from laboratory data.  Both methods are poorly described.  How is the “gamma=1e-6” value determined?  How is the “a” value determined?  Because of this poor presentation, readers will not be able to use the paper’s conclusions. Hence, the deficiency of presentation.

 

Smaller problems are as follows:

 

3. Table 1.  Values of D_50, D_10, and D_60/D_10 should be given for their specimens.

4. Line 223.  The paper should give the order in which the different strain levels were applied.

5. Fig. 7.  The \sigma_c values for figures b, e, and f should be shown in their panels.

6. Line 263.  What is “it”?

7. Line 351.  Did the authors use bender element or resonant column tests?

8. Line 437.  This sentence is confusing.

Author Response

The authors' responses to reviewer's comments, please see attached file "Paper_Response".

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The Reviewer’s comments on the original manuscript have not been fully addressed in the revised document.  Again, the Reviewer regretfully recommends that the paper is declined for reasons of impact and presentation.

1. The authors have explained their reasons for comparing several methods to resolve the asymmetry of stress-strain results from cyclic triaxial tests.  They have not recommended a method that will give results close to those of cyclic simple shear tests.  The authors’ reply says, “This is beyond the scope of this work.”  Because the paper does not recommend one or another method for use with the equivalent-linear method, it continues to lack impact.

2. The paper says “Gmax can be found using these two methods” (line 300).  The revised manuscript still does not adequately explain the two methods. Because of this poor presentation, readers will not be able to use the paper’s conclusions.

Author Response

The detailed response is presented in the file "Paper_Response".

Author Response File: Author Response.docx