Parameter Calibration of Xinjiang Paperbark Walnut Kernels by Discrete Element Simulation

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Thank you for your very careful reply to the modification suggestions one by one.

I hope the author could consider the following issues:

1) you said  that "The walnut was captured in September to October, and we purchased it in the middle of September. Therefore, it can be considered that the walnuts have little difference on physical characteristic of the walnut between the factory and the supermarket.” I'm afraid,  I can't  agree with you. It's not a matter of storage and time. It's a matter of vibration from the place of origin to the supermarket during transportation, so the physical properties have changed, which is contrary to the starting point of this study.

2) It is precisely because vibration is an important factor that changes the walnut kernel physical properties, so it is suggested that vibration test should be conducted in this paper.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report (Previous Reviewer 2)

The changes made by the authors have improved the quality of the manuscript. The manuscript can be accepted in its present form.

Author Response

The changes made by the authors have improved the quality of the manuscript. The manuscript can be accepted in its present form.

We really appreciate your positive comments on this manuscript, which significantly help us improve this manuscript.

Reviewer 3 Report (New Reviewer)

The paper is designed and developed properly. It needs two parts to be improved;

1- A section to explain about the computation cost.

2- A section explaining the mesh dependency of system.

Author Response

The paper is designed and developed properly. It needs two parts to be improved;

1- A section to explain about the computation cost.

2- A section explaining the mesh dependency of system.

Good point. The computation cost and the mesh were added in the text.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report (Previous Reviewer 1)

The revision made based on the comments has significantly improved the manuscript.

The manuscript can be accepted in its present form.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

 

To solve the problem of lacking an accurate model for mechanized transportation and classification of walnut kernels, this paper took the walnut kernel after shell-breaking treatment as the research object and employed the discrete element simulation software EDEM to calibrate the parameters of the discrete element model of walnut kernel fragmentation.

Here are some suggestions for this article：

1）The research object of this paper is Xinjiang paper walnut kernel purchased from the supermarket. Walnuts are transported from the origin to the market through many intermediate links, and they are almost in the hands of the final consumers. Therefore, I believe that if the research goal is to serve the transportation and classification, you should choose the walnut kernel produced in the factory without transportation as the object.

2）  In the article it is mentioned that “The model of the broken walnut kernel shown in Figure 3 was established, and the particle size distribution was specified as a normal distribution. A virtual cylinder was established in the EDEM software as a particle factory (the inner diameter was 300 mm and the height was 150 mm)“,What are the maximum and minimum particle sizes? How many particles are there in the particle factory? Is there only one particle structure in Figure 3?

3）  In this paper, the accuracy of the walnut kernel discrete element model is verified by comparing the walnut kernel crushing simulation test with the physical test. I think it is more appropriate to use the vibration test to simulate the transportation and classification process.

4）   The precision values of parameters in many tables in the paper are inconsistent, and some values have high precision, but they are unnecessary. For example, the Relative error (32.8515625, 23.8984375, 10.46875,  3.0078125, 6.875) in Table 6 has greatly different integers, and the precision is not necessarily high; For example, the relative error in Table 7 and the Maximum normal stress and Maximum sheet stress in Table 9（ 23.3333333 and 76.666667） are the same.

Author Response

Reviewer #1:

To solve the problem of lacking an accurate model for mechanized transportation and classification of walnut kernels, this paper took the walnut kernel after shell-breaking treatment as the research object and employed the discrete element simulation software EDEM to calibrate the parameters of the discrete element model of walnut kernel fragmentation.

We really appreciate your comments and constructive suggestions to improve this manuscript and fully followed suggestions to clarify the unclear points.

Here are some suggestions for this article：

1. The research object of this paper is Xinjiang paper walnut kernel purchased from the supermarket. Walnuts are transported from the origin to the market through many intermediate links, and they are almost in the hands of the final consumers. Therefore, I believe that if the research goal is to serve the transportation and classification, you should choose the walnut kernel produced in the factory without transportation as the object.

Thank you for your suggestion. After harvesting, walnut need to be dried and stored before consuming. Walnut has a stable physical characteristic during storage for the low moisture content. The walnut was harvested in September to October, and we purchased it in the middle of September. Therefore, it can be considered that the walnuts have little difference on physical characteristic of the walnut between the factory and the supermarket.

2. In the article it is mentioned that “The model of the broken walnut kernel shown in Figure 3 was established, and the particle size distribution was specified as a normal distribution. A virtual cylinder was established in the EDEM software as a particle factory (the inner diameter was 300 mm and the height was 150 mm)”, What are the maximum and minimum particle sizes? How many particles are there in the particle factory? Is there only one particle structure in Figure 3?

The maximum and minimum particle size were 0.4 mm and 0.3 mm, respectively. There were 2870 particles in the particle factory. Serval kinds of particles were used in the particle factory (as shown in the Fig. 1). Furthermore, this information was added in the manuscript.

In Fig. 3, the models of broken walnut kernel scanned by the 3D scanner and used in the EDEM were presented to explain the process of the establishment of the discrete element broken kernel bonding model.

3. In this paper, the accuracy of the walnut kernel discrete element model is verified by comparing the walnut kernel crushing simulation test with the physical test. I think it is more appropriate to use the vibration test to simulate the transportation and classification process.

We totally agree with you. It is more appropriate to use the vibration test to simulate the transportation and classification process. However, it is necessary to determine the mechanical properties of walnut kernel before the vibration test. Furthermore, we will conduct the vibration test in our further experiment. 

4. The precision values of parameters in many tables in the paper are inconsistent, and some values have high precision, but they are unnecessary. For example, the Relative error (32.8515625, 23.8984375, 10.46875, 3.0078125, 6.875) in Table 6 has greatly different integers, and the precision is not necessarily high; For example, the relative error in Table 7 and the Maximum normal stress and Maximum sheet stress in Table 9（3333333 and 76.666667）are the same.

Thanks for your suggestion. We changed the precision values of parameters in this article.

 

Author Response File: Author Response.docx

Reviewer 2 Report

This manuscript is a case study where the authors have presented a calibration workflow of DEM particles that is routinely done otherwise. Yes, they have picked walnuts as their system, but such calibration workflows on other systems already exist. In other words, there isn’t much novelty in this work. However, if this journal supports publishing case studies of routine scientific knowledge, then the work may be of interest to some.

I believe that this work is very specific to this current case study. The authors have identified some important parameters and some values of these parameters. They haven’t presented any study to show how these parameters may change with change in the size and shape of the kernels.

a) How will these values change with change in size and shape of the walnut kernels?

b) Do the authors think that the values of the parameters they have identified can be used by others as it is? If yes, what evidence do they have?

c) How is this work useful to other researchers? What is the takeaway message?

 

Unless the above questions are answered, I do not see how this work is valuable to other researchers.

Author Response

Reviewer #2:

This manuscript is a case study where the authors have presented a calibration workflow of DEM particles that is routinely done otherwise. Yes, they have picked walnuts as their system, but such calibration workflows on other systems already exist. In other words, there isn’t much novelty in this work. However, if this journal supports publishing case studies of routine scientific knowledge, then the work may be of interest to some.

Thanks for the critical review on this paper. We considered that this work is very specific. In this paper, we have determined some important parameters, which were changed with the size and shape of the walnut kernel, for a classifying equipment. Furthermore, these results were not found in publications.

1. How will these values change with change in size and shape of the walnut kernels?

The physical and contact parameters are inherent properties and will not change with the shape of walnut kernels. The bonding parameter, is the mechanical parameter between the filled particles, changes with the size and position of the particles.

 

2. Do the authors think that the values of the parameters they have identified can be used by others as it is? If yes, what evidence do they have?

The values of the parameters can be used by others. First, we used the EDEM to calibrate the parameters and verified it through the physical experiments. Secondly, the calibration results were also used in the simulation study of the classifying equipment. The error of the classifying equipment was less than 4 %, which proved that values of the parameters were reliable. Therefore, the values of the parameters can be used by others for the similar experiments. 

3. How is this work useful to other researchers? What is the takeaway message?

The main objectives of this paper were calibrated the walnut kernel parameters and obtained the bonding model of walnut kernel, which presented the mechanical properties of walnut kernel. During processing, packaging, and transportation, some walnut kernels will be break. To avoid the breakage, the researchers need to discuss the mechanical properties of walnut kernel, which determined the breakage ratio of walnut kernels. Then, the methods can apply to reduce the breakage ratio of walnut kernel. Therefore, this study provided useful information for the walnut kernel processing. 

Unless the above questions are answered, I do not see how this work is valuable to other researchers.

Thanks again for the review.

Author Response File: Author Response.docx

Reviewer 3 Report

In the last decades, the discrete element method (DEM) has become widely used in many application fields. This, together with the need to generate new and higher quality data, requires that the calibration of the parameters be increasingly precise. In the manuscript presented for review, the authors studied the Xinjiang paperbark walnut kernel by DEM Simulation. As these parameters are complicated to measure experimentally, they are usually calibrated. This kind of research, not profoundly original but necessary, allows the scientific community to continue advancing in the knowledge of granular materials.

I believe some issues should be addressed but then would have no problems in recommending this article for publication.

Concerns about the presentation:

Principal issues:

1. The conclusions must be improved. In particular (1) and (2) do not seem to be adequate:

1.1-L.296 "measured by the test": what do you mean?

1.2-L.297, does it make sense to use five significant digits?

1.3-L301-302, Measured values have no assigned error!

1.4-L302, "The rolling friction and its square term" or “the static friction and its square t”, seem like a misinterpretation of the statistical results.

Other issues:

2.1- The precision of the values and their errors is not adequate (see for example l.75, l.91, table 7, etc.) and some results have no assigned error (l. 82 - 83).

2.2- In section 2.1.2 the authors claim to have used a cylinder, but in the figure they show a funnel.

2.3- L.89-l.124:Why did you use different lift speed for experiments and simulations. That could affect the results.

2.4- When you present the table 1, there is no explanation of what levels means.

2.5- L. 135: What do you mean by “The modeling process”.

2.6- L. 137: Which is the size of the larger particles (or both).

2.7- L.139: “Then, the model is imported with the key...” What does this sentence mean?

2.8- L.150: “Two-level experimental was designed.” You mean: Two-level experiment was designed? There is no explanation or citation explaining this or the software that was used.

2.9- L.155: The same as before for “Steepest climb test”; “response surface test”, “BBD” and “CCD”.

2.10- L.158: By “the significance was increased” you mean “the significant parameters was increased”

2.11-L.179: “six parameters” or nine?

2.12-L180: The sentence “The simulated results presented in Table 4…” has no verb.

2.13-L187 “packing angle” → repose angle?

2.14- The tables (in particular 6, 7 and 9) have a bad significance digits use.

2.15- Too many data, poor analysis.

2.16- L.232: Which “function of Python”?

2.17- L-281: “The model parameters were set according to Table 12” → Table 11?

2.18- L285: The sentence “As shown in Figure 6,...” is not clear at all.

Author Response

Reviewer #3:

In the last decades, the discrete element method (DEM) has become widely used in many application fields. This, together with the need to generate new and higher quality data, requires that the calibration of the parameters be increasingly precise. In the manuscript presented for review, the authors studied the Xinjiang paperbark walnut kernel by DEM Simulation. As these parameters are complicated to measure experimentally, they are usually calibrated. This kind of research, not profoundly original but necessary, allows the scientific community to continue advancing in the knowledge of granular materials. I believe some issues should be addressed but then would have no problems in recommending this article for publication.

We really appreciate your critical comments on the manuscript. It significantly helps us to improve the article. We really hope the revised manuscript might meet your requirement.

 

Concerns about the presentation:

Principal issues:

1.1. The conclusions must be improved. In particular (1) and (2) do not seem to be adequate:

Thanks for the suggestions. The conclusions were improved.

 

1.2. L.296 "measured by the test": what do you mean?

The confusable word was changed. 

• 297, does it make sense to use five significant digits?

Thanks for the suggestion. We changed the precision values of parameters in this article. 

• L301-302, Measured values have no assigned error!

The missing error was added in the manuscript.

• L302, "The rolling friction and its square term" or “the static friction and its square t”, seem like a misinterpretation of the statistical results.

Thanks for your careful reading. The misinterpretations were changed.

Other issues:

2.1- The precision of the values and their errors is not adequate (see for example l.75, l.91, table 7, etc.) and some results have no assigned error (l. 82 - 83).

Changed as suggested. The missing error was added in the manuscript.

2.2- In section 2.1.2 the authors claim to have used a cylinder, but in the figure they show a funnel.

Changed as suggested.

2.3- L.89-l.124:Why did you use different lift speed for experiments and simulations. That could affect the results.

We used same lift speed for experiments and simulations. The wrong description of the stacking process and simulation process has been corrected.

2.4- When you present the table 1, there is no explanation of what levels means.

The explanations were added in the table 1.

2.5- L. 135: What do you mean by “The modeling process”.

The confusable expression was changed.

2.6- L. 137: Which is the size of the larger particles (or both).

The cylindrical walnut kernel bonding model was constructed by particles with same size. The small particle was a mistake input and we changed it.

2.7- L.139: “Then, the model is imported with the key…” What does this sentence mean?

This sentence mean that the establishment of the model was carried out step by step. First, the bonding bonds were formed, and then the bonding bonds were imported for the simulation.

2.8- L.150: “Two-level experimental was designed.” You mean: Two-level experiment was designed? There is no explanation or citation explaining this or the software that was used.

2.9- L.155: The same as before for “Steepest climb test”; “response surface test”, “BBD” and “CCD”.

The explanations and citations of two-level experiment, Steepest climb test, response surface test, BBD, and CCD were added.

2.10- L.158: By “the significance was increased” you mean “the significant parameters was increased”

Changed as suggested.

2.11-L.179: “six parameters” or nine?

It's nine, not six. We corrected it.

2.12-L180: The sentence “The simulated results presented in Table 4…” has no verb.

The verb was added.

2.13-L187 “packing angle” → repose angle?

Changed as suggested.

2.14- The tables (in particular 6, 7 and 9) have a bad significance digits use.

Thanks for the suggestion. The precision values of parameters were changed.

2.15- Too many data, poor analysis.

Analysis and discussing were added.

2.16- L.232: Which “function of Python”?

Python was used to evaluate the minimum value of destructive power which was added in the manuscript.

2.17- L-281: “The model parameters were set according to Table 12” → Table 11?

2.18- L285: The sentence “As shown in Figure 6,...” is not clear at all.

Changed as suggested.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

I genuinely thank the authors for taking the time to go through my comments. Unfortunately, I am still unclear about the usefulness of this work. The authors have claimed that the values of the DEM parameters they have determined for walnut kernels can be used universally by others, but there is no evidence. They have not shown the validity of these DEM parameter values through experimental validation at large scale processing, packaging or transportation. Moreover, if there's a need-to-know the mechanical properties of the walnut kernels for designing the large-scale processes like packaging, transportation etc., then they can be easily measured in lab. What would one do with these DEM parameter values unless there's evidence of experimental validation?

Moreover, authors said that the physical/contact parameters shouldn't change with shape. The contact parameters, such as some friction coefficients may not change with shape, however the contact pattern may change, such as the point of contact or area of contact. Authors do not provide any discussion about that. Moreover, the walnut kernels are big enough that these physical parameters do not change with size either. However, there are certain powders whose physical properties change drastically with change in the particle size. For example, Acetaminophen. 

I think that the authors should show the applicability of the DEM parameter values they have determined at large scale problems. For example, simulate a transportation or packaging scenario and validate the simulation with experimental data. I think that will be good evidence, which will also improve the information provided in this paper.

 

Reviewer 3 Report

 

I still believe that this type of work, even without being very original, is important. However, the article has serious flaws to correct in order to be recommended for publication. While the aim of the study is clear, the methodology and results are not.

Serious flaws:

1. The article cited to describe the “two-level test” and the “steepest climb test” [24] does not even mention these terms.

2. There are inconsistencies in the data. For the posose angle test, the radius of funnel outlet is 8 mm, but the dimensions of the walnut kernel are 33 mm x 23 mm. How do they flow?

3. Even using the cut particles, whose dimensions are also not detailed (only the length, 10 mm), which is not what is shown in figure 1, it seems difficult for such particles to flow.

4. For the crushing test, the authors cut the particles and obtain cylinders of 10 mm length and 7 mm diameter. These particles also do not flow through a 16 mm diameter hole without clogging.

5. The particles shown in figure 2 do not seem to have the mentioned aspect ratio.

6. The factory used in EDEM was a funnel? or did you designed a funnel which was filled with a particular factory?

7. In fig 5 there appear to be two types of particles of different sizes. What are those sizes?

8. Conclusion 1 shows results not conclusions.

Other issues:

Review nonsense sentences and even without verbs. Some examples:

l.20-21 What do you mean by “actual values”?

l.76 How do you measured the density? Seems to be very accurate.

l.90 “Fill the funnel” sounds like an order.

l.92 What do you mean by “angle of reactivity”?

l.114 The whole paragraph is unclear. In particular, what do you mean by “force process”?

l.124 Which are the normal distribution values

l.184 What do you mean by “according to the formation of...”

l.235 “analyzed by software” which?