Optimization of Laser-Induced Hybrid Hardening Process Based on Response Surface Methodology and WOA-BP Neural Network

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

Please find enclosed my comments.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

 

The manuscript, " Optimization of Laser-Induced Hybrid Hardening Process Based on Response Surface Methodology and WOA-BP Neural Network," primarily focuses on two simulation methods for predicting the depth and hardness of a steel sample after the Laser-Induced Hybrid Hardening Process is performed on the sample. The models are: 1. Response Surface Methodology (RSM) and 2. back propagation neural network optimized by the whale optimization algorithm (WOA-BPNN). The authors have compared the efficiency of these 2 models in predicting the experimental results and have concluded that WOA-BPNN model outperforms RSM in stability and generalization, achieves higher accuracy, and has superior generalization and multi-objective prediction capability.

The manuscript has a clear structure. While the presentation is generally clear, I believe in some sections of the paper addition of more explanation/clarification is necessary. The introduction provides an overview and the background of the subject and concludes with a summary of the authors' approach. The experimental setup section provides insights into the experimental setup, but addition of more clarification and in some cases, explanation will be beneficial for the readers.  In the 3^rd section of the paper the authors present and discuss their results, this section is generally clear and coherent, but please refer to comments below for areas that I think can be clarified and/or improved.

 

General Comments:

·       I believe the addition of some references to the paper will improve the paper.

1.     In the introduction the authors mention they have used Box-Behnken Design (BBD), Response Surface Model (RSM), and Whale Optimization Algorithm (WOA-BPNN). Although these are well-known in the field, I still think it will be beneficial if authors add references which can provide general introduction/fundamentals of these methods and models. I should mention that the WOA is referenced in line 253, but I think it should also be referenced in the introduction.

2.     In section 3 the authors mention F-test and p-value which are statistical terms, it will be beneficial if the authors also add references which can provide general introduction/fundamentals of these statistical methods.

 

·       Tables 4, 5, 6, and 7:

In these tables the column “Significance” needs to be updated. Stating “yes” and “No” for each of the parameters separately. For example, in table 4, parameters AB and AC seemingly do not have significant effect based on the text and their P-value, but the way the table is represented, reader can assume that all parameters from row 1 to 10 have significant effect on the depth of hardened layer. The same issue is present on the other 3 tables.

 

·       Subsections 3.1 and 3.2:

In both subsections by comparing the data in the tables 4-5 and 6-7 which correspond to IF=10 and 15 kHz, it seems that some of the parameters that have significance effect vary between the case of IF=10 kHz and 15 kHz.  Do authors have any explanation for this observation?

 

·        Subsection 3.5:

In this subsection the results from previous subsections are compared.  The authors present the parameters R2, and R2adjust for RSM model and RMSE, MAE, and MAPE after tables 2,3,4, and 5 in the text of the paper. Also, in lines 309-315 similar results are reported in text for the WOA-BPNN model.

I believe it will be beneficial for the reader (To facilitate easier comparison of the reported results) if the authors add two tables summarizing these results in subsection 3.5.

 

Specific Comments:

·       Lines 90-93:

Can the authors please specify if 50 mm or 55 mm is the dimension corresponding to the depth (the dimension that is parallel to the incidence laser beam)?

·       Lines 100-102:

I think this sentence needs to be clarified. Is this centerline correspond to the center of sample? Which surface authors are referring to when mentioning “0.5 mm from the upper surface”?

·       Lines 112-124:

Can the authors please clarify for each of the 17 data points (for both sets corresponding to IF= 10 and 15 kHz), what was their system for choosing other parameters (P, V, and IP)?

·       Lines 122-124:

The authors mention the addition of datapoint form 20 experiments that are systematically generated using MATLAB. Can the authors please add more details on these datapoints? From the sentence it is not clear for me if these datapoint are generated via experimental measurements or via MATLAB through simulations? For these datapoints do the authors split them uniformly between IF= 10 kHz and 15 kHz like the previous 34 experiments?

·       Lines 150-152:

In the text it is mentioned that quadratic term A² has a notable effect with p-value smaller than 0.05 (p<0.05), but in table 4 the p-value corresponding to this parameter is 0.1436. Can the author please clarify this discrepancy?

·       Lines 185-191:

The authors mention that the terms B² and AB² and 3 interaction terms (AB, AC, and BC) have significant effect with P<0.05. I just want to mention that in table 6 term B² has a value of 0.0550 which is not smaller than 0.05. I believe that while the mentioned value for B² in table 6 can still represent possible significance, I think it is better if the sentence is corrected to represent that fact that B² is slightly higher than 0.05.

Also based on the data from table 6 it seems that parameter A² also has p- value equal to 0.0321 <0.05. Is there any specific reason why the authors do not mention its significance?

·       Figure 2 and 3:

In the caption of the figure authors mention labels (a) to (e), but these labels are not presented in the figures themselves.

·       Lines 298-306:

Initially in line 300 the authors mention the training target of 0.0001. Then in line 306 the authors mention “hardness fitness met the training target”. However, based on the reported values of Fitness in figure 6(b), the final Fitness value achieved is close to 0.06. Can the authors please clarify this?

·       Lines 328-336:

The authors mention that MATLAB is used to randomly generate 10 sets of process parameters.  Can authors please add an explanation about this process. As in figure 7 the authors also have represented experimental results, are these 10 random sets chosen from the previously mentioned sets in table 3?

·       Figure 7:

On the horizontal axis the authors have used label “Number of Samples”. I believe the term “Sample Number” would be more appropriate to describe these axes.

·       Figure 8(a):

To have a better understanding of the represented image, Can the authors please modify the image, by adding more details? What do horizontal and vertical directions in this figure represent? Does this image need a color map?

Is the vertical axis in this image correspond to depth (which corresponds to horizontal axis of figure 8(b))?

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper entitled “Optimization of Laser-Induced Hybrid Hardening Process Based on response Surface Methodology and WOA-BP Neural Network” employed a Box-Behnken Design (BBD) experiment to construct a 63 response surface model (RSM) and a back propagation neural network optimized by the 64 whale optimization algorithm (WOA-BPNN) model.

The paper is well-written, objective, and the methods appear to be applied appropriately. However, I am sending some corrections to further improve its quality, as outlined below:

·         Line 46, Page 2: The expression "test-improve" appears repetitively. It is recommended to revise the wording to avoid redundancy.

·         Table 1 – Chemical Composition of the Material: The chemical composition of 42CrMo steel is presented. If this composition was determined experimentally, it is essential to specify the equipment and method used for the analysis. Otherwise, the source from which the composition was obtained should be cited.

·         Table 1 Caption: The notation “wt./%” should be corrected to “wt.%,” which is the standardized way to express weight percentage.

·         On line 95, page 3, it is stated: 'A light-absorbing coating is applied to the treatment surface to optimize laser absorption, achieving over 85% absorption by the pre-treated specimen.' The nature of this coating should be mentioned.

·         Line 99, Page 3 – Microhardness Testing: The current text states that hardness measurements were conducted using a fully automated Vickers microhardness tester (model HMV-2) with a vertical load of 2.942 N for 10 seconds. However, it is crucial to specify the brand of the equipment in addition to the model to ensure accuracy and traceability of the results.

·         Hardness Measurement Distribution: The study mentions that “Hardness values were recorded at five points near the center line, 0.5 mm from the upper surface, and averaged. The depth characteristics of the hardened layer were determined according to the user's requirements, and it is a heat-treated area with a hardness higher than 480 HV0.3.”. Regarding the description of this procedure, some doubts remain, and a diagram would most likely help to better understand the process. Why was a value of 480 HV0.3 defined for the hardened layer? Were all hardness measurements taken at 0.5 mm from the surface and repeated five times?

·         Figure 2: The different sub-images in Figure 2 are not labelled. They need to be identified as a, b, c, d, e, and f.

·         Figure 3: The same applies to Figure 2.

·         Lines 310, 311, and 315, page 12: pay attention to significant digits (e.g., 0.99534, 0.99584, and 0.7867%).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I want to thank the authors for their detailed responses to my comments and also for modifying the manuscript. I think the authors have addressed all my comments. 

Reviewer 3 Report

Comments and Suggestions for Authors

After carefully reading the authors' response to the reviewers, it was possible to conclude that the suggested modifications were implemented and the paper is now ready for publication.