Optimizing Roller Design to Improve Web Strain Uniformity in Roll-to-Roll Process

Round 1

Reviewer 1 Report

This paper combines physical experiments and numerical simulations to improve the strain distribution of the web during the roll-to-roll process by optimizing the roller design. The authors firstly adopted finite element analysis (FEA) and physical experiments to show the non-uniform strain distribution of the web during the static tension tests and to validate that FEA is a valid design tool.  A novel design with a barrel shape roller is proposed to alleviate the strain deviation. A parametric study was carried out to probe the design space and resulted in the best combination of design parameters using FEA. The optimized roller was manufactured and tested, and the results show better performance on the strain distributions for the web. This work is novel and has value in practice. Yet, significant improvement is needed before it can be published in Applied Science. Detailed comments are:

• English is informal with lots of typos and grammar errors. For example, the statements on the results should use present tense, not past tense. It is recommended to have an editing service or a professional to rewrite the whole manuscript.
• The literature review in the Introduction should be enriched with the technologies adopted in the industry to handle the uneven strain distribution. Readers cannot distinguish the proposed optimization method from existing methods.
• In Figure 1, the red line on the boundary with applied force should be deleted given that the surface is applied uniform tension, not fixed displacements.
• Symbols for the distance between the roller and the print lines, as well as the width of the web, should be used and add those symbols for the legends in figure 2&3. The current ones are misleading.
• Mark the direction of MD and CD in Figure 1.
• A quantitative comparison between the FEA prediction and the experimental measurements is recommended for the validation in Section 2.3. Two different figures of different quantities (strain deviation v.s.  relative displacement) with a similar pattern do not provide a strong validation. The authors should plot them in the same figure with the same quantity.
• The sensitivity analysis in Section 3.3 needs further clarification on the logic and results. What does the number represent in the columns of AB, AC, and BC in Table 3?
• The last column should be 9, not 11.
• How are the results in Figure 9 obtained? Does the maximum and minimum strain difference include all scenarios when one design parameter is fixed? The same question goes for Figure 10.
• The acronym in Table 5&6 needs to be explained either in the context or in the title.
• Given the proposed roller is termed as barrel roller, the authors should keep the term consistent throughout the manuscript. The interchange between compensation roller and barrel roller is confusing to readers, e.g., Figures 15 & 16.
• Section 4.2 is misleading. Directly compare the performance using the experimental data is more convincing. Also, lots of info is missing for the FEA. The material input parameters for the two rollers? Two lines (distance 50 mm and 150 mm) are investigated? The title of Figure 17 is wrong. The tension is fixed as 5MPa.
• I believe the y-axis of Figures 2, 3, and 17 should be the absolute value of strain, not the strain deviation.

Author Response

Please find the attached. The authors replied in it.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper describes the application of DOE and Taguchi methodologies to the design of web used in the roll-to-roll process. This process is gaining technical interest because it can be used to produce films for the new generation of mobile technology devices. The paper is well written, the methodology phases follow in a meaningful way, and a set of pictures illustrates the main issues described in the paper in a quite comprehensive way. The English is correct and understandable without efforts by the reader. 

The paper is strong in the methodology that is “case study driven”, but can be applied in other fields too. The weakest point of the paper is that a reader not well acquainted with the knowledge of DOE or Taguchi methods could be disoriented being the manuscript without a description of these tools.

Therefore, I suggest the following minor improvements in the paper:

• avoid the use of abbreviations in the Abstract
• In the FE model, did you check the results convergence with change in number of nodes? Such a study is usually carried out to be confident in the final result.
• I would add in the section two a scheme showing the position of the cameras respect to the web, together a description of the cameras features.
• I suggest to add a short description of what is DOE and Taguchi methods: some lines to summarize how it work plus a pair of references where the reader is addressed to better understand both the methodologies;
• I would describe how the probability distribution that is used in ANOVA (F-Fisher) can be used for statistical inference considerations; also in this case some lines plus a reference could be useful for the reader.

Would it be possible to carry out a 1-factor ANOVA with the most impacting factor to evaluate its importance?  You say in the text that there is one factor whose importance is higher by far than others. The output of a process can be approximated in such a way: medium value + Treatment effect + noise (see Montgomery book on DOE for instance). I would be interested in understanding the effect of treatment effect respect to medium value in numerical way.

Author Response

Please find the attached. The authors replied in it.

Author Response File: Author Response.pdf

Reviewer 3 Report

The article deals with an interesting and current topic and is beneficial for practice. The used methods are acceptable and the results are realistic. The method of data processing and commenting is fine. The cited references are corresponding to the solved topic.

There are several formal shortcomings in the article, these are highlighted in the appendix.

Comments for author File: Comments.pdf

Author Response

Please find the attached. The authors replied in it.

Author Response File: Author Response.pdf