Development of Closed-Form Equations for Estimating Mechanical Properties of Weld Metals according to Chemical Composition

Round 1

Reviewer 1 Report

The manuscript reported a data analysis using ANN approach to investigate the effect of chemical composition on mechanical properties. Overall, it is not good. I suggest a major revision, and the following concerns must be addressed before publication:

(1) A careful reading of the text should be done to suppress language errors. A lot of language errors have been found in the text. Please check carefully. Also, please reduce the use of “many, better, some, various” in scientific writing.

(2) “Abstract” and the last paragraph of “1. Introduction” both mentioned that “impact toughness” was studied, but it was not actually studied in this manuscript.

(3) “1. Introduction” is too long and does not fit the topic. Please rewrite it. For example, paragraphs 4, 5, and 6 are obviously repetitive.

(4) The expressions are not consistent. For example, the element symbol (Mn) is used in “1. Introduction”, while the full name (Manganese) is used in “2.1. Data collection”.

(5) I have doubts about “2.2. Data augmentation”. Various nonlinear laws were guessed by just a few points (specially 4 points), as shown in Fig. 3. An ANN was used to reproduce the guessed nonlinear laws. For example, the ANN reproduction data corresponding to Fig. 3 is displayed in Fig. A1. The various nonlinear laws have been guessed. Is it necessary to reproduce them? Please give a reasonable explanation.

(6) Please rewrite "3.1. ANN Model", especially paragraphs 1 and 2. The patchwork trace is obvious, and the connection is problematic.

(7) Please check the figures and tables are correctly described. For example, it is mentioned that "The models for all eight cases, as summarized in Table 1, are…" Table 2, not Table 1, should be correct here.

(8) The symbols should be explained in equations. Otherwise, the readers will not understand the meaning. Especially for Eq. (3). By the way, are Eq. (1) and Eq. (3) the same? If so, please emphasize the relationship between them.

(9) It is suggested to plot the four curves in each figure (i.e., Fig. 8, Fig. 9, Fig. 10, Fig. A1–A7) in one graph for easier comparison.

(10) Please rewrite “5. Conclusions”.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Reviewer comments:

1. Introduction should address the clear research gap and objectives at the end of the introduction based on the literature survey. The authors should convey to the readers how your work is unique compared to other literatures. The last paragraph in the introduction section summarizes mainly the methods. In the same last paragraph, it is required to add research gap and objectives. The author should also express the novelty of the work.
2. Use “ANN Model” instead of “ANN’ or “ANN technique” wherever required throughout the paper.
3. The data used for data augmentation and then as an input to the ANN model is taken from another literature article. It is required to provide a table with all the training data inputs and outputs, test data and outputs.
4. From the section 2.1, the reviewer could understand that the authors have manufactured some electrodes by varying the chemical compositions and also performed welding operation using SMAW. The major important details are missing in the manuscript.
   1. The pictures of the manufactured electrode samples, welded samples, manufacturing unit and welding setup should be added.
   2. The grade of mild steel (base material) and the chemical composition should be added to the manuscript.
5. Since the electrodes are manufactured by the authors, the manufacturing process of the electrodes has to be added to the manuscript.
6. Details about the SMAW welding power source is missing.
7. How is the interlayer temperature measured? The instrument or the methodology has to be added.
8. The authors have stated that the welding speed is set in such a way that the heat input is fixed to 1kJ/mm. I assume the process is a manual SMAW process. How was it ensured the heat input be fixed at 1kJ/mm in a manual SMAW process?
9. The images of the Charpy test specimens are missing in the paper.
10. The author has mentioned that two 20 mm thick plates were welded by three passes. Figure 1(a) shows more number of passes (greater than 3). Instead of adding and citing the image from the paper, the authors are suggested to draw their own schematic sketches.
11. The procedure used for estimating the chemical compositions of the electrode should be mentioned in the methods.
12. Line 129: The standard “ISO 2560:1973” is currently not inactive status. Please revise the ISO standard with updated code. https://www.iso.org/standard/35626.html. The standards should be cited.
13. Line 136: Mention the dimension of the root gap maintained in the weld design.
14. Figure 2 is not clear.
15. Section 2.2: Non-linear regression analysis is used for predicting more input data (as ANN requires more data input) for Mn of 0.4, 0.6, 1.0 and 1.5C. Please clearly explain how accurate the predicted data are? ( R and R-sq values of regression analysis)
16. Line 204: Please change “traditional method” to “statistical method”.
17. Section 3.1: Development of ANN model is good along the optimizing the hidden neurons. Please add how many input data (experimental runs) are provided for ANN model. Also mention how many data for training, testing and validation in number (along with percentage).
18. Line 280: remove “s” from “and”.
19. Section 4: The comparison of predicted results from ANN and the experimental runs has been made. The discussion was made only about the ANN model whereas the technical point of view is not explained i.e. why there is variation in the mechanical properties due to changing the composition of Mn and other components?
20. Please rewrite the conclusion part. Please enhance the conclusion with the results obtained in the study. For example, at point 1, the performance of the ANN model will definitely increase if the numbers of input data are high.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Authors

I have read the Development of Closed-Form Equations for Estimating Mechanical Properties of Weld Metals according to Chemical Composition send to Metals. The article is interesting. The subject is very interesting. I have some suggestions to consider before final publication. In the introduction, the Authors did not specify the purpose of the research. What is the utilitarian aspect of research.

Author Response

Thank you for your valuable comment. We have modified the introduction section accordinly.

Round 2

Reviewer 1 Report

It is acceptable.

Author Response

Thank you for your comment.

Reviewer 2 Report

The authors have not conducted any experimentations. Instead, the data has been taken from the work of Evans and Bailey. The data has been augmented. However, the data is still can only be considered as a small set even after augmentation. There are no efforts to validate the augmented data by experimentations. The methodology is misleading giving an impression to authors that some experimentations have been conducted. The authors failed to provide informations that have been sought from point 3 to point 8 in the last reviewer comments. 

  The reviewer's decision is to reject the manuscript.

Author Response

Please see the attached file.

Author Response File: Author Response.pdf