Determination of Material and Interaction Properties of Granular Fertilizer Particles Using DEM Simulation and Bench Testing
, Wencheng Wu 1, Xuan Deng 1, Tao Li 1, Hongnan Liu 1, Jinyue Guo 1, Ju Li 1,2, Peixu Zhu 1 and Ke Yang 1Round 1
Reviewer 1 Report
It is a good project, however, i have some observations.
1- In figure 3(d), it is necessary to add scales to both x and y axis in the line graph
2- I have the same observation for figure 5(d).
3- I suggest to add more details about the experiment setup.
Author Response
Dear Editor and Reviewers:
Thank you very much for your letter and for reviewers’ comments concerning our manuscript entitled “Determination of Material and Interaction Properties of Granular Fertilizer Particle Using DEM Simulation and Bench Testing” (ID: agriculture-2496358). These comments are valuable and helpful for revising and improving the paper, as well as the important guiding significance to our researches. We have studied all reviewers’ comments carefully and have made some corrections and looking forward having your approval. The main corrections in the paper and the responds to the reviewers’ comments are as following:
Responds to the reviewers’ comments:
Reviewer #1:
Comment #1: In figure 3(d), it is necessary to add scales to both x and y axis in the line graph
Response: Thank for the reviewer’s comments. We have added scales to both X and Y axis in Figure 3(d).
Comment #2: I have the same observation for figure 5(d).
Response: Thank for the reviewer’s comments. We have added scales to both X and Y axis in Figure 5(d).
Comment #3: I suggest to add more details about the experiment setup.
Response: Thank for the reviewer’s comments. We have added the experiment setup details in the chapter of 2.3.
Reviewer 2 Report
Principally an interesting topic, but the outcome is not very easy to understand. The storyline of the paper needs to be improved. It’s confusing how the own results have been linked to the existing results.
Material and methods
1 You start with the geometrical dimensions OK
2 Coefficient of restitution: You describe the theory and experimental setup: OK
è As you do not describe more, the reader expect then the results of the experiment.
3 Angle of repose is explained. This is ok but finish this topic here, include Fig. 5, Fig. 7 and Table 5. Isn’t the method of fig 9 used in fig 5? Describe the experimental setup how you measured the angle of repose and present the results accordingly.
4 How did you measure the different parameters indicated in Tab. 2 (Density, coefficient of static friction,… rolling friction….) this is not clear.
5 DEM simulation should be a distinct paragraph where you apply the DEM and compare the results with your experimental measurements and literature.
Actually it’s not clear to the reader what exactly YOU have measured and what has been modelled by means of DEM.
Wouldn’t it make sense to describe all proceeded steps of the angle of respose in one paragraph:
- How did you exactly measure it? (3 fertilizers but according of Table 3 you made 4 combinations x1-x4 out of these 3 fertilisers. How did you realise this in the experiment?
A description of the apiece row consistency variation coefficient is missing. How it’s possible to compare this parameter with a fertilizer metering mechanism?
Results
Table 4: Are statistics not relevant? At least an indication of the error would make sense to get an impression of the variability.
Table 5: Why do you only present the single results for stainless steel? Why not using box plots to show the variability of the different materials?
Fig. 11 What are the 4 graphs exactly showing? The term “coefficient of static friction” hasn’t been introduced so far.
You show 5 tests for the angle of repose, then you show it for different plates. For which fertilizer do you show these results. It’s not possible to understand exactly what the message of theses graphs is.
How are the different parameters related? In the conclusion you describe that the angle of repose is influenced by the static and rolling friction of the particles. Please show that in your results.
Discussion
This is too much a repetition of the introduction and some results. Here a profound discussion is needed how well you are capable to simulate experiments and how well the characteristics of fertilizer grains can be modelled by means of DEM. This needs to be revised completely.
Conclusions should not be another abstract…
no specific comments
Author Response
Comment #1: Coefficient of restitution: As you do not describe more, the reader expect then the results of the experiment.
Response: Thank for the reviewer’s comments. The results of the experiment are in the chapter of 3.1.
Comment #2: Angle of repose is explained. This is ok but finish this topic here, include Fig. 5, Fig. 7 and Table 5. Isn’t the method of fig 9 used in fig 5? Describe the experimental setup how you measured the angle of repose and present the results accordingly.
Response: Thank for the reviewer’s comments. The method of Fig 9 is the same as Fig 5, which was introduced in the chapter of 2.1.5.
Comment #4: How did you measure the different parameters indicated in Tab. 2 (Density, coefficient of static friction,… rolling friction….) this is not clear.
Response: Thank for the reviewer’s comments. We have added the measurement method of density, coefficient of static friction in chapter of 2.1.2 and 2.1.4.
Comment #5: DEM simulation should be a distinct paragraph where you apply the DEM and compare the results with your experimental measurements and literature.
Actually it’s not clear to the reader what exactly YOU have measured and what has been modelled by means of DEM.
Wouldn’t it make sense to describe all proceeded steps of the angle of repose in one paragraph:
How did you exactly measure it? (3 fertilizers but according of Table 3 you made 4 combinations x1-x4 out of these 3 fertilisers. How did you realise this in the experiment?
A description of the apiece row consistency variation coefficient is missing. How it’s possible to compare this parameter with a fertilizer metering mechanism?
Response: We have made correction according to the Reviewer’s comments. The DEM simulation method was introduced in the chapter of 2.2.
The density, coefficient of restitution between particles and other materials, and static friction between particle and other materials were determined by bench test. These determined simulation parameters were to model by means of DEM.
The steps of the angle of repose were in the chapter of 2.1.5. The Coding of factors in Table 3 was from Table 2 which was a range from bench test.
The apiece row consistency variation coefficient was revised as coefficient of variation for each row. It was introduced in the chapter of 2.3. The fertilizer feeding performance of a fertilizer metering mechanism can be validated between DEM simulation and bench test.
Comment #6: Table 4: Are statistics not relevant? At least an indication of the error would make sense to get an impression of the variability.
Response: We have added variance analysis in Table 4 according to the Reviewer’s comments.
Comment #7: Table 5: Why do you only present the single results for stainless steel? Why not using box plots to show the variability of the different materials?
Response: Thank for the reviewer’s comments. Because the coefficient of restitution, coefficient of static friction, coefficient of rolling friction of fertilizer particles, and coefficient of static friction of fertilizer-plate material are effect factors, coefficient of restitution, coefficient of static friction, coefficient of rolling friction of fertilizer particles were determined by the single results for stainless steel. The interaction properties between fertilizer particles were the same for different treatments between fertilizer particles and other materials.
Comment #8: Fig. 11 What are the 4 graphs exactly showing? The term “coefficient of static friction” hasn’t been introduced so far.
Response: We have added the introduction of coefficient of static friction in the chapter of 2.1.4 according to the Reviewer’s comments.
Comment #9: You show 5 tests for the angle of repose, then you show it for different plates. For which fertilizer do you show these results. It’s not possible to understand exactly what the message of theses graphs is.
Response: The 5 tests was to verify the accuracy of simulation parameters between simulation and bench test. The Zhonghua fertilizer was the material was added. In Fig. (12a) and (12d), the simulation and experimental results of angle of repose and coefficient of variation for each row were compared. In Fig. (12b) and (12c), the coefficient of static friction between fertilizer particles and aluminum and acrylonitrile-butadiene-styrene copolymer were determined.
Comment #10: How are the different parameters related? In the conclusion you describe that the angle of repose is influenced by the static and rolling friction of the particles. Please show that in your results.
Response: Thank for the reviewer’s comments. The results of angle of repose influenced by the static and rolling friction of the particles was illustrated in Table 6 and chapter of 3.2.
Comment #11: This is too much a repetition of the introduction and some results. Here a profound discussion is needed how well you are capable to simulate experiments and how well the characteristics of fertilizer grains can be modelled by means of DEM. This needs to be revised completely.
Response: Thank for the reviewer’s comments. We have revised the discussion.
Comment #12: Conclusions should not be another abstract.
Response: Thank for the reviewer’s comments. We have revised the conclusions and abstract.
Reviewer 3 Report
Notes are provided in PDF file.
Comments for author File: 
Comments.pdf
Author Response
Comment #1: A lot of attention is paid to angle of repose, but it is not clear how it is related to the work efficiency of fertilizer spreaders, because centrifugal force works in spreaders, and angle of repose is important for fertilizer storage. Therefore, the essence of the work is unclear.
Response: Thank for the reviewer’s comments. The purpose of this manuscript was to determine the simulation parameters of fertilizer material and interaction properties. The angle of repose and fertilizer performance result were utilized to verify accuracy of simulation parameters. This work can improve the accuracy of the simulation model for granular compound fertilizer and aid in design fertilizer spreader.
Comment #2: The term "fertilizer seeders" is incorrect, it should be "fertilizers spreader".
Response: We have made correction according to the Reviewer’s comments.
Comment #3: Incorrect term "varieties of fertilizers", find a more suitable one.
Response: We have made correction according to the Reviewer’s comments.
Comment #4: It is not clear what it is: „Zhushang, Zhonghua, and Sannin“ fertilizer name, manufacturer or something else?
Response: We have added the manufacturer in the chapter of 2.1.1. Zhushang, Zhonghua, and Sannin are fertilizer names according the manufacturer.
Comment #5: If the studies were done with spherical granules, how can "length, width and thickness" by different? And according to what criteria did you determine where the length, width, and thickness are in the spherical granules?
Response: Thank for the reviewer’s comments. The fertilizer particles are ellipsoid granules, rather than spherical granules according to the triaxial dimensions of granular compound fertilizer.
Comment #6: The purpose of parts (b) and (c) of Figure 5 is unclear. Please justify their necessity.
Response: Thank for the reviewer’s comments. The purpose of parts (b) and (c) of Figure 5 was to introduce the method of achieving repose angle from original image.
Comment #7: The authors write: The Hertz-Mindlin (no-slip) model, as shown in Figure 6, was chosen as the particle contact model, and the granular compound fertilizer was modelled as a hard-sphere model in the simulation, but they do not explain why it was chosen. How is it better than others, why is it suitable in this case, etc.
Response: Thank for the reviewer’s comments. We have added the explanation of Hertz-Mindlin (no-slip) model in the chapter of 2.2.1.
Comment #8: The authors write that "The angle of repose was tested using the hollow cylinder method, which consists of a hollow cylinder and a bottom plate", but do not explain why. How does the angle of repose formed by the cylinder model relate to fertilizer spreading?
Response: Thank for the reviewer’s comments. The hollow cylinder method a typical method to measure angle of repose according previous studies. The angle of repose was to verify the accuracy of the DEM simulation parameters. If the DEM simulation parameters are accurate, the simulation of fertilizer spreading can be well simulated.
Comment #9: What is marked with (*) in Table 2 is not explained.
Response: Thank for the reviewer’s comments. We have added the explanation of (*) in Table 2.
- The authors write, “As some parameters were measured or from previous research, the values of uncertain parameters including the coefficient of restitution, the coefficient of static friction, the coefficient of rolling friction between fertilizer particles, and the coefficient of static friction of fertilizer-plate material”, but does not specify in which previous studies.
Response: Thank for the reviewer’s comments. We have added the references of previous studies.
- The authors write: The four factors with three levels of experimental design were carried out using the Box-Behnken method", but it is not explained why such a choice was made, there is no comparison with other suitable models. Are there no other suitable ones?
Response: Thank for the reviewer’s comments. The Box-Behnken method was a feasible and time-saving experimental method.
- In Figure 11 (a) (b) (c) the data is only in a narrow range, i.e., angle of repose varies between 30-40 degrees, and the Y axis is 3 times larger and most of the graph is empty. It needs to be fixed.
Response: Thank for the reviewer’s comments. The X axis in Figure 11 (a) (b) (c) was not the same. Thus, the relative error presentation is small.
- The reference [1] is not appropriate, because it talks about the influence of ammonium sulphate on the granulation process, and not about "crop yield and quality".
Response: We have made correction according to the Reviewer’s comments.
- The reference [24] is not Wang
Response: We have made correction according to the Reviewer’s comments.
- The reference [24] is not Song X. F
Response: We have made correction according to the Reviewer’s comments.
- The reference [23] is not suitable, because there it is written about the angle of repose of granular materials (mainly about soil granulometry), and not about "The coefficient of restitution is a measure of the energy transferred between particles and the surface of impact"
Response: Thank for the reviewer’s comments. As the granular materials were discrete particles, it can be used in the research of fertilizer particle.
Author Response File: 
Author Response.doc
Round 2
Reviewer 2 Report
The structure of the paper is still poor and it’s extremely difficult to follow the logic line of the paper. What do the authors want to tell the reader? On one hand they made several bench tests and on the other hand they made simulations. Unfortunately they mixed both activities and didn’t make a clear separation. It’s not clear what exactly they measured and how well they were able to simulate the characteristics of the three different fertilizers and the simulation of the feeding unit.
The discussion is still another introduction but is not clearly discussing the achieved results.
The paper shouldn’t be published in the present form.
p.8 In the chapter «DEM simulation» you cite Fig. 8 which seems to be a physical experiment with the cylinder. Why do you mix experiment and DEM-simulation? If you need the experiment to test your model then clearly show this and describe the different activities in different chapters. In Table 2 you cite literature values which you need for your DEM model. It’s a mess between DEM model, physical experiment and literature values. Please make a clear logical order.
Under 2.3 you describe again the cylinder test. This is not adequate. Describe the whole thest from A-Z at one place.
One paragraph with one table showing all parameters for the DEM and their source would help the reader to understand which parameters you included into your model.
Then please indicate what exactly you are simulating. Did you simulate the feeding mechanism (Fig. 11)? If yes, then clearly explain this or leve this away.
Simulation results
It’s very difficult to follow your results. What exactly has been measured on test benches, and what exactly has been simulated?
The title is not correct as you present the results of the bench tests too.
Please present first all the results of the bench tests and the in a next chapter the simulation results. Please make clear what exactly you simulated.
3.2 Simulation parameters determination
You simulate the angle of repose and which parameter else? Why don’t you show the measurements and simulations of the different parameters?
Table 5. It seems as you used the factors of Table 3 to make your calculations. It’s not clear why the same test factors don’t result in the same values. Why don’t you include the values of table 3 into table 5…
On p. 12 you argue: “The simulation parameters were optimized to achieve an objective angle of repose of 33.78°, which was obtained from the hollow cylinder bench test. Using numerical optimization analysis, the coefficient of restitution, coefficient of static friction, and coefficient of rolling friction between fertilizer particles were determined to be 0.323, 0.381, and 0.173, respectively. The coefficient of static friction between fertilizer particles and 304 stainless steel was 0.453.”
Don’t you overfit your sample and do not present an independent model? You measured 3 different fertilizers in different samples. Why don’t you present the measurements of the simulated results?
Why did you only do 5 measurements for the angle of repose? This doesn’t give a statistical relevant bases. In table 5 you presented 29 test numbers.
Discussion
The discussion is started like an introduction again. This is not adequate. Discussion means that you discuss your results and don’t repeat an introduction. This is not the case. The discussion is not acceptable like this.
It’s absolutely not clear how you can simulate the result of a feeder unit.
Conclusion
This is still another abstract and not concluding.
What can be simulated in future with your results? What do you tell a fertilizer spreader constructer he can use easier build a machine…
Englisch is OK.
Author Response
Dear Editor and Reviewers:
Thank you very much for your letter and for reviewers’ comments concerning our manuscript entitled “Determination of Material and Interaction Properties of Granular Fertilizer Particle Using DEM Simulation and Bench Testing” (ID: agriculture-2496358). These comments are valuable and helpful for revising and improving the paper, as well as the important guiding significance to our researches. We have studied all reviewers’ comments carefully and have made some corrections and looking forward having your approval. The main corrections in the paper and the responds to the reviewers’ comments are as following:
Responds to the reviewers’ comments:
Reviewer #1:
Comment #1: p.8 In the chapter «DEM simulation» you cite Fig. 8 which seems to be a physical experiment with the cylinder. Why do you mix experiment and DEM-simulation? If you need the experiment to test your model then clearly show this and describe the different activities in different chapters. In Table 2 you cite literature values which you need for your DEM model. It’s a mess between DEM model, physical experiment and literature values. Please make a clear logical order.
Response: Thank for the reviewer’s comments. We have made correction according to the Reviewer’s comments.
Fig. 8 presents the angle of repose test process in the DEM simulation. Then we test the angle of repose using experimental platform. We separate the experiment and DEM-simulation.
In Table 2, we added the parameters origin and the order is from the DEM simulation parameter.
Comment #2: Under 2.3 you describe again the cylinder test. This is not adequate. Describe the whole thest from A-Z at one place.
Response: Thank for the reviewer’s comments. We have combined the cylinder test in the chapter of 2.2.4.
Comment #3: One paragraph with one table showing all parameters for the DEM and their source would help the reader to understand which parameters you included into your model.
Response: Thank for the reviewer’s comments. All parameters for the DEM were illustrated in Table2.
Comment #4: Then please indicate what exactly you are simulating. Did you simulate the feeding mechanism (Fig. 11)? If yes, then clearly explain this or leve this away.
Response: Thank for the reviewer’s comments. We revised the DEM simulation part. The angle of repose for different simulation parameters and fertilization process of a fertilizer metering device were tested. And the simulation process of feeding mechanism was added.
Comment #5: It’s very difficult to follow your results. What exactly has been measured on test benches, and what exactly has been simulated?
The title is not correct as you present the results of the bench tests too.
Please present first all the results of the bench tests and the in a next chapter the simulation results. Please make clear what exactly you simulated.
Response: We have made correction according to the Reviewer’s comments. We have separated the DEM simulation and bench test. Coefficient of restitution and angle of repose of experimental study was introduced in the chapter of 3.1, and parameters determination in DEM simulation was introduced in the chapter of 3.2, and parameters verification between DEM simulation and bench test was introduced in the chapter of 3.3.
Comment #6: You simulate the angle of repose and which parameter else? Why don’t you show the measurements and simulations of the different parameters?
Response: We have made correction according to the Reviewer’s comments. The angle of repose is an evaluation index to assess the accuracy of DEM simulation parameters. The effective factors were coefficient of restitution (x1), the coefficient of static friction (x2), the coefficient of rolling friction (x3) between fertilizer particles, and the coefficient of static friction of fertilizer-plate material (x4).
Comment #6: Table 5. It seems as you used the factors of Table 3 to make your calculations. It’s not clear why the same test factors don’t result in the same values. Why don’t you include the values of table 3 into table 5…
Response: Thank for the reviewer’s comments. Table 3 introduced the simulation design, including effect factors and levels. Table 6 introduced the simulation results. The simulation method and results should be separated.
The reason of the same test factors resulting in different values is the randomness of fertilizer particle movement.
Comment #7: On p. 12 you argue: “The simulation parameters were optimized to achieve an objective angle of repose of 33.78°, which was obtained from the hollow cylinder bench test. Using numerical optimization analysis, the coefficient of restitution, coefficient of static friction, and coefficient of rolling friction between fertilizer particles were determined to be 0.323, 0.381, and 0.173, respectively. The coefficient of static friction between fertilizer particles and 304 stainless steel was 0.453.”
Don’t you overfit your sample and do not present an independent model? You measured 3 different fertilizers in different samples. Why don’t you present the measurements of the simulated results?
Response: Thank for the reviewer’s comments. The optimal parameters were obtained by numerical optimization analysis using Design-Expert version 13.0. In DEM simulation, the simulation model parameters are based on average values, and the constructed model has strong adaptability.
Comment #8: Why did you only do 5 measurements for the angle of repose? This doesn’t give a statistical relevant bases. In table 5 you presented 29 test numbers.
Response: Thank for the reviewer’s comments. 5 measurements were used to verify the accuracy of DEM simulation. In table 5, the optimal simulation parameters should be determined.
Comment #9: The discussion is started like an introduction again. This is not adequate. Discussion means that you discuss your results and don’t repeat an introduction. This is not the case. The discussion is not acceptable like this.
Response: Thank for the reviewer’s comments. We have made correction according to the Reviewer’s comments.
Comment #10: This is still another abstract and not concluding.
What can be simulated in future with your results? What do you tell a fertilizer spreader constructer he can use easier build a machine…
Response: Thank for the reviewer’s comments. We have made correction according to the Reviewer’s comments.
We tried our best to improve the manuscript and made some changes in the manuscript. These changes will not influence the content and framework of this paper.
We appreciate for Editor and Reviewers’ warm work earnestly, and hope that the correction will meet the approval.
Once again, thank you very much for your comments and suggestions.
Sincerely yours
Dr. Xiaolong Lei
Reviewer 3 Report
Komentaras pridedamas docx faile.
Comments for author File: Comments.pdf
Author Response
Dear Editor and Reviewers:
Thank you very much for your letter and for reviewers’ comments concerning our manuscript entitled “Determination of Material and Interaction Properties of Granular Fertilizer Particle Using DEM Simulation and Bench Testing” (ID: agriculture-2496358). These comments are valuable and helpful for revising and improving the paper, as well as the important guiding significance to our researches. We have studied all reviewers’ comments carefully and have made some corrections and looking forward having your approval. The main corrections in the paper and the responds to the reviewers’ comments are as following:
Responds to the reviewers’ comments:
Reviewer #2:
Comment #1: In Figure 11 (a) (b) (c) the data is only in a narrow range, i.e., angle of repose varies between 30-40 degrees, and the Y axis is 3 times larger and most of the graph is empty. It needs to be fixed.
Response: Thank for the reviewer’s comments. We have made correction according to the Reviewer’s comments.
We tried our best to improve the manuscript and made some changes in the manuscript. These changes will not influence the content and framework of this paper.
We appreciate for Editor and Reviewers’ warm work earnestly, and hope that the correction will meet the approval.
Once again, thank you very much for your comments and suggestions.
Sincerely yours
Dr. Xiaolong Lei
