Predicting High Temperature Flow Stress of Nickel Alloy A230 Based on an Artificial Neural Network

Round 1

Reviewer 1 Report

Review of the paper «Predicting high temperature flow stress of Nickel alloy A230 based on artificial neural network», for publication in MDPI Metals.

 

The paper "Predicting High Temperature Flow Stress of Nickel Alloy A230 2 based on Artificial Neural Network" deals with the plastic flow analysis of a Ni superalloy, namely A230. The flow curves are obtained experimentally, for the temperature range while an artificial network artificial intelligence (AI) technology, which is known to be effective in modeling complex relationships. As a result, it was confirmed that the flow stresses modeled by the artificial neural network (ANN) used in the simulation of the present paper could be effectively compared to the Arrhenius model. The paper is interesting, logically organized and well written, even if some revisions are required. The publication of the paper in the MDPI journal Metals is compatible with the topic and the development of the subject; major revisions are envisaged before the publication is possible.

 

The used English is suitable for the scientific level of the manuscript. A thorough correction is needed, as typos/minor errors are present. For example, line 166 and following: «The input layer has three nodes. And a normalized temperature, a normalized strain rate and a strain are inputted to each node. The reason for normalizing the temperature and strain rate values is to ensure that each input data has the same sensitivity in deriving a result value by making the numerical range of each data the same. The output layer has one node, and it correspond to normalized stress values.» "And" should not be used at the beginning of a sentence. "correspond" is a third person, a "s" is required.

 

lines 78-79: «The cut specimens were mounted using conductive carbon powder, and then mechanically polished.» Please, provide a better description about the sample preparation procedure.

 

line 81: «The polished surface was observed through the EBSD measurement.» The authors should provide a better explanation of the used techniques, of the used parameters and of the sample preparation method for EBSD.

 

lines 93-95: «After the strain reaches to the critical strain, DRX occurs actively so that the flow stress is gradually reduced. As a result, a steady-state in which work hardening and stress softening are balanced is observed. » The author should point out clearly in the graphs where this happens.

 

formula 1: Z is a specific parameter, well known to specialists in the field, that should be clearly named.

 

formula 1: the used symbol for deformation rate, έ, has an apostrophe instead of a dot on the top. It should be modified, for it is not correct in the present form.

 

lines 162-190: «To overcome these shortcomings, a flow stress modeling using ANN model was applied in this study.... » This section contains an amount of useful information for the identification of the features used in the definition of the ANN parameters. This section should be put in the materials and methods section, as it is a procedural section. It is not in the appropriate place, because it is an important parenthesis inside the Result section.

 

line 164: «multilayer perceptron (MLP)» the author should comment their choice on the basis of similar works

 

As a general comment, the author wrote an interesting work, but the comparison between the present work development (obtained data, processing, comparison of same method for same alloys, or different method for the same alloy, etc.) is a bit scarce. The author should improve this part.

Author Response

First of all, we greatly thank for the reviewer’s valuable comments. We tried to give the best answers to the questions.

 

Comment 1: line 166 and following >> "And" should not be used at the beginning of a sentence. "correspond" is a third person, a "s" is required.

Answer.

→ As reviewer’s recommend, the modification were made. [lines 84 ~ 88]

 

Comment 2: lines 78-79 >> Please, provide a better description about the sample preparation procedure.

→ The entire paragraph was deleted because it is not related to the subject. (we did not analyze microstructure (EBSD))

 

Comment 3: line 81 >> The authors should provide a better explanation of the used techniques, of the used parameters and of the sample preparation method for EBSD.

Answer.

→ The entire paragraph was deleted because it is not related to the subject. (we did not analyze microstructure (EBSD))

 

Comment 4: lines 93-95 >> “After the strain reaches to the critical strain, DRX occurs actively so that the flow stress is gradually reduced. As a result, a steady-state in which work hardening and stress softening are balanced is observed.” >> The author should point out clearly in the graphs where this happens.

Answer.

→ As reviewer’s recommend, corresponding points (critical strain, peak strain, steady-state) were presented in Figure 2(d)

 

 

Comment 5: formula 1: Z is a specific parameter, well known to specialists in the field, that should be clearly named.

Answer.

→ As reviewer’s recommend, Z (Zener-Hollomon parameter) was clearly named. [line 127]

 

Comment 6: formula 1: the used symbol for deformation rate, έ, has an apostrophe instead of a dot on the top. It should be modified, for it is not correct in the present form.

Answer.

→ As reviewer’s recommend, the modification were made. [Eq. 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], [lines 127, 140, 144, 149, 155]

 

Comment 7: lines 162-190: «To overcome these shortcomings, a flow stress modeling using ANN model was applied in this study.... » This section contains an amount of useful information for the identification of the features used in the definition of the ANN parameters. This section should be put in the materials and methods section, as it is a procedural section. It is not in the appropriate place, because it is an important parenthesis inside the Result section.

Answer.

→ As reviewer’s recommend, position of the paragraph was modified. [Section 2.2]

 

Comment 8: line 164: «multilayer perceptron (MLP)» the author should comment their choice on the basis of similar works

Answer.

→ MLP and ANN is similar meaning, so MLP expression was deleted. [line 82]

→ Therefore, the reason for the choice of ANN was delivered based on the similar work. [lines 62 ~ 64]

 

Comment 9: As a general comment, the author wrote an interesting work, but the comparison between the present work development (obtained data, processing, comparison of same method for same alloys, or different method for the same alloy, etc.) is a bit scarce. The author should improve this part.

Answer.

→ As reviewer’s recommend, background literatures for flow stress prediction using AI technique for other materials were added. [lines 54 ~ 64]

Reviewer 2 Report

Dear Authors, 

I appreciate the effort you have put in preparing this manuscript through a genuine investigation. The results are interesting and novelty is significant. I have read the paper thoroughly and I don't see any technical flaw or mistake in the paper however, some minor spell checks are required. 

Author Response

First of all, we greatly thank for the reviewer’s valuable comments. We tried to give the best answers to the questions.

 

Comment 1: some minor spell checks are required.

Answer.

→ As reviewer’s recommend, the manuscript was delicately reviewed.

Reviewer 3 Report

The paper presents the application and optimization of an artificial neural network for predicting the flow stress for a part made of A230 nickel alloy.

The paper structure is generally good, although the description of the ANN employed in the paper should have been maybe placed in paragraph 2 with the description of the employed material and methods and not only in paragraph 3 (Results and discussion).

The researches described in the paper are thorough and well described.

The English language employed in the paper is generally good, although there are some minor problems - e.g. in line 192, instead of "To easy the comparison..." there should have been "To ease the comparison" or even "To facilitate the comparison...", or in line 217 "a tendency to variability of the error level" (?)

 

Author Response

First of all, we greatly thank for the reviewer’s valuable comments. We tried to give the best answers to the questions.

 

Comment 1: The paper structure is generally good, although the description of the ANN employed in the paper should have been maybe placed in paragraph 2 with the description of the employed material and methods and not only in paragraph 3 (Results and discussion).

Answer.

→ As reviewer’s recommend, position of the paragraph was modified. [Section 2.2]

 

Comment 2: in line 192, instead of "To easy the comparison..." there should have been "To ease the comparison" or even "To facilitate the comparison...",

Answer.

→ As reviewer’s recommend, the modification was made. [line 192]

 

Comment 3: in line 217 "a tendency to variability of the error level" (?)

Answer.

→ As reviewer’s recommend, the paragraph has been rewritten [line 216 ~ 219]

Round 2

Reviewer 1 Report

Review of «Predicting high temperature flow stress of Nickel alloy A230 2based on artificial neural network» (METALS-1497128)

Just a comment

→ The entire paragraph was deleted because it is not related to the subject. (we did not analyze microstructure (EBSD))

I think that this is relevant, because in any case this information is shown to the reader and presented for revision. This is the same for everything submitted to peer review, the accuracy of all the elements in the paper needs to be checked and evaluated.

The paper can however be published.

Thank you.