An Attempt to Model the Surface Pressures of Apples Using the Finite Element Method
Round 1
Reviewer 1 Report
In this paper, the authors used the Finite Element Method (FEM) to develop a simplified FEM model of an apple (‘Golden Delicious’ variety). The FEM model was validated against previously published impact test results. In general, it is an interesting and well-presented paper. However, some minor corrections are needed.
In the introduction, please indicate the significance of this study. For example, in which applications the experimentally validated FEM model will be used?
Please add a picture of the FEM model and explain why the apple was modelled as an ideal sphere and not of a more ‘realistic’ shape?
In the legend of Figures 2 – 4 please make more clear which data correspond to the experimental tests for concrete substrate.
Please add values at the contour plots of Figures 12, 14, 16 and 18 to make easier the comparison between FEM and experimental data.
Please correct the unit in the X-axis of Figure 17.
At the conclusions, please make clarify which are the findings of this study and mention possible future research.
Author Response
We would like to thank the Reviewer for valuable comments. Our answers are presented below.
„In this paper, the authors used the Finite Element Method (FEM) to develop a simplified FEM model of an apple (‘Golden Delicious’ variety). The FEM model was validated against previously published impact test results. In general, it is an interesting and well-presented paper. However, some minor corrections are needed”.
- Question: „In the introduction, please indicate the significance of this study. For example, in which applications the experimentally validated FEM model will be used?”
Answer: Lines 135-136; „The experimentally validated FEM model in this study will be used to determine the appropriate conditions related to the transport, storage and processing of apples”.
- Question: „Please add a picture of the FEM model and explain why the apple was modelled as an ideal sphere and not of a more ‘realistic’ shape?”
Answer; We have added a picture. It is the first step in modeling surface pressures, therefore it was decided to simplify the model to verify whether during such research it would be possible to obtain results similar to the real findings. It should be noted that structure of the apples as well as other agricultural products are different and very often there are significant differences in their shape (i.a. diameter, radius and other curvatures) between them. While modeling them with the reflection of real shapes (which includes the future plans on the basis of current research), it will be necessary to decide how to determine a shape of the apple as similar to real shape.
- Question „In the legend of Figures 2 – 4 please make more clear which data correspond to the experimental tests for concrete substrate.”
Answer The changes were implemented in the manuscript.
- Question „Please add values at the contour plots of Figures 12, 14, 16 and 18 to make easier the comparison between FEM and experimental data.
Answer The changes were implemented in the manuscript.
- Question. Please correct the unit in the X-axis of Figure 17.
Answer The changes were implemented in the manuscript.
- Question. At the conclusions, please make clarify which are the findings of this study and mention possible future research.
Answer: Lines: 446-457; Presented research which inlcudes a modeling surface pressures can replace long-lasting and time-consuming experimental studies, but what is more important, determined FEM numerical model can be used for analysis and design a new devices, technologies or systems, e.g. during transport, processing in order to reduce not only the risk of damage, but real deformation of the tissues of plant origin at the stage of harvesting, transport, storage and processing them, which will result in quality and price of them.
Next step of presented research will be a creation of model which geometry will be very similar to real object. Future plans inlcude research on another apple varieties and other types of agricultural products, for which it is possible to define the damage of flesh tissue on the basis of surface pressures. The created models will be used to analyze the processes of harvesting, transport and storage in order to reduce the occurrence of damage and to design new technological lines.
Reviewer 2 Report
The paper presents a numerical analysis of the impact between apples and different substrates. Numerical contact pressure, contact area, and contact force records were compared against experimental results. In my view, the paper still needs some work before it can be considered for publication. The authors should address the following points:
Line 59: the expression "..in the structure of the internal structure of..." is not clear; please clarify.
Line 94: please specify what the acronym "MES" means
Line 154: A figure showing the details of the numerical model setup and the boundary/initial conditions used in this analysis needs to be included somewhere in this section.
Line 163: what do the authors mean by "...all the elements were adjusted to non-linear calculations in terms of material and geometry..." Please specify what adjustments were made.
Line 166: The sentence: "The choice of material was so difficult that it had to be suitable for studying dynamic phenomena." is not clear. It seems that the authors are trying to connect a high level of difficulty with the fact that the material model needs to address dynamic phenomena. Please rephrase or clarify in the paper.
Line 211: "On the basis of their observations, they assumed the hypothesis that..." Not sure to whom the authors are referring in here. Also, the rest of this sentence seems very disconnected with the rest of the paragraph - please double check.
Line 220: this is the first time that corrugated cardboard is mentioned in the text. The usefulness of the sentence that starts with "Comparison of substrates..." is not clear since no results are presented.
Lines 245-247: this paragraph should be deleted!
Line 237: please include in the paper the formula used to calculate the "model fitting error". Without that, it is not possible to assess what a 10% model fitting error actually means.
Lines 238-240: Figure 2 shows the average pressure obtained from the numerical results as a function of those obtained from the experimental results. Ideally, if the model were to replicate the experimental data exactly (considering the same initial and boundary conditions, i.e., drop height, apple size, etc.), then the slope of the regression curves would be equal to one, i.e., the model would yield exactly the same value of surface pressure as the one reported by the experiments. However, in Figure 2, the slope of the regression curves are quite far away from the value of one, 0.3579 for the wood substrate and 0.4497 for the concrete substrate. Please explain why this is so in this work. A similar situation is observed in figures 3 and 4, although figure 3 shows a much better comparison mainly for the wood substrate where the slope is very close to one.
Line 259: The legend of the axis is wrong, it should read "Surface Area" instead of "Surface pressure".
Line 275: how is the value of the surface pressure obtained both for the experiments and for the numerical modeling? Is it an average pressure across the contact area? Please explain in the paper.
Line 284: How do these results correlate to those presented in Figure 2? Please clarify in the paper.
Line 294: How do these results correlate to those presented in Figure 3? Please clarify in the paper.
Line 302: How do these results correlate to those presented in Figure 4? Please clarify in the paper.
Line 333/Figure 12: while the numerical results are very symmetric (as it is to be expected, since the shape of the apples is assumed to be spherical), the experimental results are not (see Figure 12a). The authors are comparing the pressure profile along a horizontal line only (which seems to be the most favorable direction); what happens if the comparison is made along a vertical line? From a visual inspection of Figure 12a it is evident that the results will be quite different. This needs to be included into the paper along with the corresponding discussion on the azimuthal differences. This same comment is valid for Figures 13/14, 15/16, and 17/18.
The conclusions are incomplete/disconnected from the paper. What are the "open methods" that the authors refer to in line 424? This is the first time that it is brought up in the paper without much explanation. Additionally, the authors refer to a "three-layered model" which was not properly explained in the paper: how thick is each layer? what are the properties used for each layer? etc. Finally, there is no proper summary of the findings of the research on the conclusions.
Author Response
We would like to thank the Reviewer for valuable comments. Our answers are presented below.
„The paper presents a numerical analysis of the impact between apples and different substrates. Numerical contact pressure, contact area, and contact force records were compared against experimental results. In my view, the paper still needs some work before it can be considered for publication. The authors should address the following points:”
- Question; Line 59: the expression "..in the structure of the internal structure of..." is not clear; please clarify.
Answer: This sentence was improved.
- Question Line 94: please specify what the acronym "MES" means.
Answer: It was a mistake, it should be described as „FEM”.
- Question Line 154: A figure showing the details of the numerical model setup and the boundary/initial conditions used in this analysis needs to be included somewhere in this section.
Answer: We added additional figure (fig.1).
The additional information mentioned in the review was deliberately not included because it represents the know-how of the model's construction and, due to its commercial application (cooperation with industry is ongoing), we must protect our own intellectual property, which we treat as an industrial secret of financial value.
- Question Line 163: what do the authors mean by "...all the elements were adjusted to non-linear calculations in terms of material and geometry..." Please specify what adjustments were made.
Answer: Data which, the Reviewer indicated, are very important part of the model construction as well as there are a part of Author’s know-how and therefore will not be presented in this paper, because the model can be used commercially.
- Question Line 166: The sentence: "The choice of material was so difficult that it had to be suitable for studying dynamic phenomena." is not clear. It seems that the authors are trying to connect a high level of difficulty with the fact that the material model needs to address dynamic phenomena. Please rephrase or clarify in the paper.
Answer: The choice of material was difficult as it had to be suitable for studying dynamic phenomena and to match the viscoelastic nature of biological material.
- Question Line 211: "On the basis of their observations, they assumed the hypothesis that..." Not sure to whom the authors are referring in here. Also, the rest of this sentence seems very disconnected with the rest of the paragraph - please double check.
Answer: Stopa et al. in their observations, assumed and confirmed the hypothesis that permanent deformations indicated a occurrence of the first signs of damage in the parenchyma tissue and it followed an increase of the impact energy corresponding with initial phase of maximum surface pressures stabilisation.
- Question Line 220: this is the first time that corrugated cardboard is mentioned in the text. The usefulness of the sentence that starts with "Comparison of substrates..." is not clear since no results are presented.
Answer: In this aspect, this sentence is related to paper (Komarnicki, P., Stopa, R., Szyjewicz, D., Kuta, Ł., Klimza, T. Influence of Contact Surface Type on the Mechanical Damages of Apples Under Impact Loads. Food Bioprocess Technology 2017, 10, 1479–1494. DOI 10.1007/s11947-017-1918-z) in which we analyzed four types of substrate (among others cardboard). In that paper, we explained why we did not carry out model tests for the cardboard substrate.
- Question Lines 245-247: this paragraph should be deleted!
Answer: This sentence was deleted.
- Question Line 237: please include in the paper the formula used to calculate the "model fitting error". Without that, it is not possible to assess what a 10% model fitting error actually means.
Answer: The average model fitting error for each substrat was determined on the basis of average value of surface pressure +/- 5% observations around trend line (p<0,95).
- Question: function of those obtained from the experimental results. Ideally, if the model were to replicate the experimental data exactly (considering the same initial and boundary conditions, i.e., drop height, apple size, etc.), then the slope of the regression curves would be equal to one, i.e., the model would yield exactly the same value of surface pressure as the one reported by the experiments. However, in Figure 2, the slope of the regression curves are quite far away from the value of one, 0.3579 for the wood substrate and 0.4497 for the concrete substrate. Please explain why this is so in this work. A similar situation is observed in figures 3 and 4, although figure 3 shows a much better comparison mainly for the wood substrate where the slope is very close to one.
Answer: Dear Reviewer, we would like to explain that the differences in these data can result from the construction of the model as well as its simplifications for model purpose. Additionally, it should be noted that it follows from material model, on which the Authors still work. Moreover, it should be emphasized that the Abaqus program is used especially for engineering purposes and not for modeling agricultural tissues because there is no solutions in this area.
- Question Line 259: The legend of the axis is wrong, it should read "Surface Area" instead of "Surface pressure".
Answer: This figure was corrected.
- Question Line 275: how is the value of the surface pressure obtained both for the experiments and for the numerical modeling? Is it an average pressure across the contact area? Please explain in the paper.
Answer: In our studies we took, an average value of surface pressure, and maximal were taken from cross-sections of the modeled objects.
Question 13. Line 284: How do these results correlate to those presented in Figure 2? Please clarify in the paper.
Answer Figure 3 shows the maximum values that occurred on the contact surface, and those at first showned on line 284 present the average values of surface pressures for each height.
Question 14 Line 294: How do these results correlate to those presented in Figure 3? Please clarify in the paper.
Answer Figure 4 shows the maximum values that occurred on the contact surface, and those at first showned on line 294 present the average values of surface pressures for each height.
Question 15 Line 302: How do these results correlate to those presented in Figure 4? Please clarify in the paper.
Answer Figure 5 shows the maximum values that occurred on the contact surface, and those at first showned on line 302 present the average values of surface pressures for each height.
Question 16 Line 333/Figure 12: while the numerical results are very symmetric (as it is to be expected, since the shape of the apples is assumed to be spherical), the experimental results are not (see Figure 12a). The authors are comparing the pressure profile along a horizontal line only (which seems to be the most favorable direction); what happens if the comparison is made along a vertical line? From a visual inspection of Figure 12a it is evident that the results will be quite different. This needs to be included into the paper along with the corresponding discussion on the azimuthal differences. This same comment is valid for Figures 13/14, 15/16, and 17/18.
Answer: Dear Reviewer we explain that we took the cross-sections so that they move through the point / points with the maximum local values. Yes, if cross-sections will be passed through other points, the graphs will look different, but in the adopted method, the cross-section should be drawn through local maximum values and in such a way that it passes through the center of the surface while object's contact with the ground.
Thanks to such presentation it is possible to clearly mark where they have been exceeded the maximum local stress values in the pulp tissue of the tested objects / model, which is associated with the occurrence of damage.
Question 17 The conclusions are incomplete/disconnected from the paper. What are the "open methods" that the authors refer to in line 424? This is the first time that it is brought up in the paper without much explanation. Additionally, the authors refer to a "three-layered model" which was not properly explained in the paper: how thick is each layer? what are the properties used for each layer? etc. Finally, there is no proper summary of the findings of the research on the conclusions.
Answer: Presented research which inlcudes a modeling surface pressures can replace long-lasting and time-consuming experimental studies, but what is more important, determined FEM numerical model can be used for analysis and design a new devices, technologies or systems, e.g. during transport, processing in order to reduce not only the risk of damage, but real deformation of the tissues of plant origin at the stage of harvesting, transport, storage and processing them, which will result in quality and price of them.
Next step of presented research will be a creation of model which geometry will be very similar to real object. Future plans inlcude research on another apple varieties and other types of agricultural products, for which it is possible to define the damage of flesh tissue on the basis of surface pressures. The created models will be used to analyze the processes of harvesting, transport and storage in order to reduce the occurrence of damage and to design new technological lines.
Reviewer 3 Report
The paper deals with an interesting topic - a simplified model of an apple for analysis by FEM.
Remarks:
- physical representation of apple is missing (characteristic cross section and view)
- it is necessary to improve the quality of the drawing Fig 1,
- it is necessary to present a sketch with the method of testing surface pressures obtained during experimental tests (sketch of test device)
- in drawing 12, 14, 16, 18 under c) - show the contours of the apple (with the designation of the contact stress domain)
Author Response
We would like to thank the Reviewer for valuable comments. Our answers are presented below.
- Question; physical representation of apple is missing (characteristic cross section and view).
Answer:. We have added a cross-sectional model drawing on page 4.
- Question; it is necessary to improve a quality of the drawing Fig 1,
Answer: Fig.1 was prepared by one of the author on the basis on private source. Unfortunately, we do not have a better copy of the presented figure, but we believe that it is possible to analyze it precisely.
- Question: it is necessary to present a sketch with the method of testing surface pressures obtained during experimental tests (sketch of test device)
Answer: In our opinion, it is unnecessary to provide full description of experimental test station because it is not the subject of this paper and could lead to unnecessary confusion about it. For this reason, this work is presented as development of our earlier work, the source was given, in which all the answers regarding experimental research were described. In addition, the experimental station does not influence a model research.
- Question: in drawing 12, 14, 16, 18 under c) - show the contours of the apple (with the designation of the contact stress domain)
Answer: We understood, that the reviewer asks for the presentation of entire modeled object in the above-mentioned drawings. We believe that this is unnecessary procedure because can cause a deterioration in perception, due to size of the figure. The presented drawings are an approximation of the contours of surface pressures, what a core of this work and also represent the designation of the contact stress domain. Below we present a drawing of the full model view to confirm our opinion that such a presentation format will have poor quality.
Number of figure has been change (+1, e.g. fig. 12 is now Fig 13) due to add additional figure.
Author Response File: 
Author Response.pdf
Round 2
Reviewer 2 Report
Dear Authors,
Thanks for revising the manuscript.
You argue that "The additional information mentioned in the review was deliberately not included because it represents the know-how of the model's construction and, due to its commercial application (cooperation with industry is ongoing), we must protect our own intellectual property, which we treat as an industrial secret of financial value." and then you go on to say that you cannot explain what adjustments were made to the non-linear calculations in terms of material and geometry because "Data which, the Reviewer indicated, are very important part of the model construction as well as there are a part of Author’s know-how and therefore will not be presented in this paper, because the model can be used commercially." This means that many of the details of the model setup, material models used and "tricks" implemented to obtain these results have been left out on purpose because of IP concerns. That is very understandable, but at the same time, it must be considered that the purpose of open publishing is to provide the research community in general with reproduceable advances that others can use in their work. If the advances are not completely explained, for whatever reason, then the work is not reproduceable and therefore begs the question: would this work be more suitable for a patent application instead of open publishing? In any case, you may want to explore the avenue of pursuing a provisional patent to protect your IP before publishing...
Regarding point #10 of the review, the authors argue that the discrepancies between the numerical model results and the experiments are due to material model features, which are not explained in the paper and the authors are still working on, and on the fact that the numerical tool used for this is not designed to model agricultural tissue. These factors result in a consistent deviation of the numerical results from the experimental results which is evidenced by the differences in the slopes of the curves shown in Figures 3 and 5 (refer to comment on previous round of review for more details on this). In my view, this needs to be explicitly mentioned in the manuscript to avoid confusion. It is not enough to have a high coefficient of determination for the regression curves, but it also needs to be stated that there is a consistent shift in the numerical results with respect to the experimental results because of the reasons presented by the authors. Please clarify this point on the paper.
Author Response
We would like to thank the Reviewer for valuable comment. Our answer is presented below.
- Question; Regarding point #10 of the review, the authors argue that the discrepancies between the numerical model results and the experiments are due to material model features, which are not explained in the paper and the authors are still working on, and on the fact that the numerical tool used for this is not designed to model agricultural tissue. These factors result in a consistent deviation of the numerical results from the experimental results which is evidenced by the differences in the slopes of the curves shown in Figures 3 and 5 (refer to comment on previous round of review for more details on this). In my view, this needs to be explicitly mentioned in the manuscript to avoid confusion. It is not enough to have a high coefficient of determination for the regression curves, but it also needs to be stated that there is a consistent shift in the numerical results with respect to the experimental results because of the reasons presented by the authors. Please clarify this point on the paper.
- Answer: Dear Reviewer there are our pilot study. In fact, a high coefficient of determination does not always guarantee a perfect fit of model to the experimental data. Beside, we can mention about random factor. In this paper we showed that there are discrepancies between the experimental and model data, which gives the basis for further studies, to make thid model more precise. The differences may results, among others, from properties of the material, conditions of fruit ripening, conditions under which the tests are performed. These premises influence the differences in contact surfaces and surface pressures in the analyzed figures 3-5.
