Multi-Objective Optimisation of Welding Parameters for AZ91D/AA6082 Rotary Friction Welded Joints

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The meaning of this work is low.

1. No joint cross sections were given.

2. No microstructure or IMC was studied.

3. No joint defect was studied.

4. No fracture behavior or fracture morphology was studied.

Comments on the Quality of English Language

This english should be improved.

Author Response

Comments

The meaning of this work is low.

1. No joint cross sections were given.
2. No microstructure or IMC was studied.
3. No joint defect was studied.
4. No fracture behavior or fracture morphology was studied.

 

Response to comment

Thank you for your comments. The aim of the article was to show an effective method of genetic algorithms for determining the optimal parameters of the welding process to obtain the appropriate (maximal) tensile strength and minimal shortening of the joint. The developed GA method allows obtaining a permanent joint of a magnesium alloy with aluminum without time-consuming attempts to select welding parameters. In further studies, the authors will investigate the microstructure and IMC layers at the weld interface. We have included this information in the conclusions (see page 16, lines 482-485 in the Conclusion chapter).

Reviewer 2 Report

Comments and Suggestions for Authors

1. The topic addressed in the article is of current interest and consistent with the journal profile.

2. The authors used modern research equipment, methods, and software.

3. It is not recommended to use abbreviations in the Abstract of an article, to facilitate a quick understanding of its content. The authors also used an incorrect abbreviation in the second sentence of the Abstract (FRW instead of RFW). Following current conventions, the abstract must include at least information regarding the research problem, the research method used, and the research results. The latter aspect is less presented in the current version of the abstract.

4. In line 42, the concept “percent” seems to be misspelled.

5. The title of Table 2 is misspelled. The title seems to be a sentence that could be included in the text of the article. For concepts listed in the first column of Table 2, the second word doesn't need to start with a capital letter. For abbreviated words, it is useful to include a period in the end (e.g., “St. dev.” instead of “St dev,” etc.).

6. The wording “After that, all contacting surfaces of samples were polished using grit silicon carbide papers to eliminate the effect of oxide films and surface roughness.” is confusing. The effect of the surface roughness could be reduced and to a lesser extent eliminated.

7. It is necessary that the first time a symbol is used, its meaning be included immediately. The meanings of the symbols “ID,” “MSE valid.” in Table 3, etc. are not mentioned.

8. More extensive explanations are needed to justify the inclusion of equations (1)-(12) and to analyze the results obtained as a result of using these equations. How were the equations corresponding to the various constraints developed?

9. How were the corresponding values ​​for metal loss determined/measured?

10. What arguments were used for the selection of ANN architecture?

11. The formulation “To find the best relations between the inputs (friction force, friction time and upsetting force) and outputs (tensile strength and metal loss) and tan-sigmoid, log-sigmoid and pure-line functions and the number of neurons (6, 9 and 12) in the hidden layers were proved (see Table 3).” is confusing. What arguments were taken into account for the use of tan-sigmoid, log-sigmoid, or pure-line function approximations?

12. In the last part of the Conclusions chapter, the possibilities for further research in the future could be mentioned.

13. A schematic representation of the way the rotary friction welding process takes place would be useful. It is not clear from Figure 1 whether both samples or only one of them are rotated, how the axial force is exerted, how the specimens were located and clamped, etc.

14. The title of the article is too general and does not mention that the research refers only to rotary friction welding.

15. The authors do not seem to have mentioned all their previous works on topics related to the subject addressed in the analyzed article. For example, the work from the web page https://www.e3s-conferences.org/articles/e3sconf/pdf/2019/58/e3sconf_polsita2019_01028.pdf, Radosław Winiczenko, Andrzej Sibicki, Paweł Skoczylas, and Jędrzej Trajer, Multi-objective optimization of the process parameters for friction welding of dissimilar metals, E3S Web of Conferences 132, 010 (2019),

https://doi.org/10.1051/e3sconf/201913201028 was not included in the bibliographic reference list.

16. The content of Figure 3 in the analyzed paper appears to be identical to that of a component of Figure 2 in reference 51 (Winiczenko, R.; Skibicki, A.; Skoczylas, P. Optimization of friction welding parameters to maximize the tensile strength of magnesium alloy with aluminum alloy dissimilar joints using genetic algorithm. Processes. 2021, 23, 385−391). The authors seem to have modified only the material symbols inscribed on the two figures. It would be preferable to use a different photograph and not the same photograph with insignificant modifications in two different scientific papers.

17. It is necessary for the authors to pay more attention to the editing of the article and the expression in English. In addition to the editing errors mentioned above, others can also be highlighted. Thus, in line 51, the tool rotation speed is expressed in min-1, but without “-1” being positioned as an exponent. For such situations, it is customary to use rev/min or rpm as units of measurement.

It is not necessary to write information in italic fonts in the title of a table (Table 1). In the title of a table, after the table number, a period is placed (for example, in the case of Table 1). A period should also be placed at the end of the legend of a figure (for example, in the case of Figure 3).

At the beginning of a paragraph, a blank space is left, by using the TAB key; in the article, in many cases, this blank space was not used.

In line 148, the word "conventional" appears to have been misspelled ("convectional").

Italic fonts could be used for symbols throughout the article (for example, in line 213, in the case of the symbol "R").

Comments on the Quality of English Language

See the comments for the authors.

Author Response

Comment 1

The topic addressed in the article is of current interest and consistent with the journal profile.

Response 1

Thank you for your comment.

Comment 2

The authors used modern research equipment, methods, and software.

Response 2

Thank you for your comment.

Comment 3

It is not recommended to use abbreviations in the Abstract of an article, to facilitate a quick understanding of its content. The authors also used an incorrect abbreviation in the second sentence of the Abstract (FRW instead of RFW). Following current conventions, the abstract must include at least information regarding the research problem, the research method used, and the research results. The latter aspect is less presented in the current version of the abstract.

Response 3

Thank you for your comment. The abbreviations have been removed from the Abstract. The sentences “RFW” have been improved in the text. The research results have been added to the abstract (see page 1, lines 26-29).

Comment 4

In line 42, the concept “percent” seems to be misspelled.

Response 4

Thank you for your comment. The comment has been considered (see page 2, line 42).

Comment 5

The title of Table 2 is misspelled. The title seems to be a sentence that could be included in the text of the article. For concepts listed in the first column of Table 2, the second word doesn't need to start with a capital letter. For abbreviated words, it is useful to include a period in the end (e.g., “St. dev.” instead of “St dev,” etc.).

Response 5

Thank you for your comment. The title of table  and text have been improved (currently, Table 5 on page 7).

Comment 6

The wording “After that, all contacting surfaces of samples were polished using grit silicon carbide papers to eliminate the effect of oxide films and surface roughness.” is confusing. The effect of the surface roughness could be reduced and to a lesser extent eliminated.

Response 6

Thank you for your comment. We agree. The text has been modified (see line 113 on page 3 in the text).

Comment 7

It is necessary that the first time a symbol is used, its meaning be included immediately. The meanings of the symbols “ID,” “MSE valid.” in Table 3, etc. are not mentioned.

Response 7

Thank you for your comment. We agree with this comment. The symbols have been explained (see line 268, page 7)

Comment 8

More extensive explanations are needed to justify the inclusion of equations (1)-(12) and to analyze the results obtained as a result of using these equations. How were the equations corresponding to the various constraints developed?

Response 8

Thank you for your comment. This text has been modified (see page 10, lines 307-334).

Comment 9

How were the corresponding values ​​for metal loss determined/measured?

Response  9

Thank you for your comment. We agree with this comment. The metal loss has been explained (see page 5, lines 198-201).

Comment 10

What arguments were used for the selection of ANN architecture?

Response 10

Thank you for your comment. The text has been improved (see page 7, lines 262-268, page 8, lines 275-281).

Comment 11

The formulation “To find the best relations between the inputs (friction force, friction time and upsetting force) and outputs (tensile strength and metal loss) and tan-sigmoid, log-sigmoid and pure-line functions and the number of neurons (6, 9 and 12) in the hidden layers were proved (see Table 3).” is confusing. What arguments were taken into account for the use of tan-sigmoid, log-sigmoid, or pure-line function approximations?

Response 11

Thank you for your comment. The text has been improved (see page 7, lines 262-268). These functions reproduced the relationships between input and output data well.

Comment 12

In the last part of the Conclusions chapter, the possibilities for further research in the future could be mentioned.

Response 12

Thank you for your comment. Further research in the future has been added in the Conclusions chapter (see page 16, lines 479-482).

Comment 13

A schematic representation of the way the rotary friction welding process takes place would be useful. It is not clear from Figure 1 whether both samples or only one of them are rotated, how the axial force is exerted, how the specimens were located and clamped, etc.

Response 13

Thank you for your comment. We agree with this comment. A schematic representation of the way the rotary friction welding process has been added (see Fig.1b on page 4). Additionally, Figure 1a has been improved.

Comment 14

The title of the article is too general and does not mention that the research refers only to rotary friction welding.

Response 14

Thank you for your comment. We agree with this comment. The tittle of the article has been improved (see page 1, lines 2-3).

Comment 15

The authors do not seem to have mentioned all their previous works on topics related to the subject addressed in the analyzed article. For example, the work from the web page https://www.e3s-conferences.org/articles/e3sconf/pdf/2019/58/e3sconf_polsita2019_01028.pdf, Radosław Winiczenko, Andrzej Sibicki, Paweł Skoczylas, and Jędrzej Trajer, Multi-objective optimization of the process parameters for friction welding of dissimilar metals, E3S Web of Conferences 132, 010 (2019),

https://doi.org/10.1051/e3sconf/201913201028 was not included in the bibliographic reference list.

Response 15

Thank you for your comment. The author's articles have been added to the literature (see. 39 and 40 in the references)

Comment 16

The content of Figure 3 in the analyzed paper appears to be identical to that of a component of Figure 2 in reference 51 (Winiczenko, R.; Skibicki, A.; Skoczylas, P. Optimization of friction welding parameters to maximize the tensile strength of magnesium alloy with aluminum alloy dissimilar joints using genetic algorithm. Processes. 2021, 23, 385−391). The authors seem to have modified only the material symbols inscribed on the two figures. It would be preferable to use a different photograph and not the same photograph with insignificant modifications in two different scientific papers.

Response 16

Thank you for your comment. We agree with this comment. A new figure has been added. (Currently, it is Fig.3 on page 5).

Comment 17

It is necessary for the authors to pay more attention to the editing of the article and the expression in English. In addition to the editing errors mentioned above, others can also be highlighted. Thus, in line 51, the tool rotation speed is expressed in min-1, but without “-1” being positioned as an exponent. For such situations, it is customary to use rev/min or rpm as units of measurement.

-It is not necessary to write information in italic fonts in the title of a table (Table 1). In the title of a table, after the table number, a period is placed (for example, in the case of Table 1). A period should also be placed at the end of the legend of a figure (for example, in the case of Figure 3).

At the beginning of a paragraph, a blank space is left, by using the TAB key; in the article, in many cases, this blank space was not used.

In line 148, the word "conventional" appears to have been misspelled ("convectional").

Italic fonts could be used for symbols throughout the article (for example, in line 213, in the case of the symbol "R").

Response 17

Thank you for your comment. We agree with this comment.  The expressions in English have been improved in all text. The unit of RS has been changed (see page 2, line 54)

 The above comments have been considered in the text.

The blank space has been used in all the articles.

The word ("convectional") has been corrected (see page 4, line 181).

Italics fonts “R” have been included in the all text.

Reviewer 3 Report

Comments and Suggestions for Authors

2.1. Materials

-Table 1 should be split into two.

-Delete the abbreviation descriptions in the table description and place the description in the table along with the units.

 

-HB means Brinell hardness, not an abbreviation of hardness. The other designations are also wrong. Add information about the measurement base for elongation, e.g. A80, A50.

-Please state the technological state of the aluminum alloy.

-Provide the friction coefficients for the material and the coefficient determined for the pair of materials. In this type of process, the friction coefficient is not constant. Are the authors aware of this? This should be described in the context of other works.

 

2.2. Welding parameters

-The title of the chapter should be changed to: sample preparation.

-The text from the last paragraph should be moved to chapter 2.2.

-A table with variable parameters for preparing joint samples should be presented.

-Please describe how a constant pressure force was obtained during the joint. Control was probably performed by feed on the machine tool?

-What were the dimensions of the samples. Why are they different in photo 2?

2.4. Testing

-Please correct the designation: (100-kN).

-The same designation of materials should be used throughout the work.

-Photo 3 shows a handle with pins. How was the grip of round samples in the testing machine? The method of attachment affects the transfer of load to the tested material.

 

3 Neural network modeling

-Before chapter 3, a chapter with the results from the basic experiment should be developed.

-Chapters 3.1. Network architecture, and 3.2. Validation and learning process should be added

-Table 2 incorrectly describes the force "Friction Force". How was the friction force determined?

-In line 212: is the correlation coefficient or R2 described? R represents the value of the Pearson correlation coefficient, which is used to mark the force and direction among variables, while R2 represents the coefficient of determination, which determines the strength of the model.

-First, the structure of the network should be presented, then its data and description.

 

5 Results and discussion

-The results should be developed better. Add more detailed analysis.

 

4. Formulation of optimization task

- the objective function should be better defined and described.

- this chapter can be part of chapter 3: 3.3. Formulation of optimization task.

6. Conclusions

Should be improved. Quantitative and qualitative evaluation of research results should be made.

 

7. Appendix A

- it is not terribly extensive. It can be placed in an appropriate place in the publication.

Author Response

Comment 1

2.1. Materials

-Table 1 should be split into two.

-Delete the abbreviation descriptions in the table description and place the description in the table along with the units.

-HB means Brinell hardness, not an abbreviation of hardness. The other designations are also wrong. Add information about the measurement base for elongation, e.g. A80, A50.

-Please state the technological state of the aluminum alloy.

-Provide the friction coefficients for the material and the coefficient determined for the pair of materials. In this type of process, the friction coefficient is not constant. Are the authors aware of this? This should be described in the context of other works.

 

Response 1

Thank you for your comment. We agree with this comment.

-Table 1 has been split into two (see Tables 1 and 2 on page 3).

-The abbreviation descriptions in the table description have been deleted. Moreover, the description in the table with the units and abbreviation of hardness and elongation have been revised. (see descriptions in Tables 1 and 2).

-The technological state of the aluminium alloy has been added (see the line 113 in the text).

-The authors have used a neural model to optimise the welding parameters using genetic algorithms (as the direct effect of the parameters on the bond strength and shortening of the joint). Therefore, in our case sudy, the friction coefficients of the materials are omitted because the model does not take into account thermal and plastic phenomena as is the case of FEM modelling.

 

Comment 2

2.2. Welding parameters

-The title of the chapter should be changed to: sample preparation.

-The text from the last paragraph should be moved to chapter 

-A table with variable parameters for preparing joint samples should be presented.

-Please describe how a constant pressure force was obtained during the joint. Control was probably performed by feed on the machine tool?

-What were the dimensions of the samples. Why are they different in photo 2?

 

Response  2

Thank you for your comment. We agree with this comment.

-The title of the chapter has been changed to “sample preparation” (see page 3, line 126 in the text).

-The text from the last paragraph has been moved to chapter  2.2.

-A table with variable parameters for preparing joint samples has been added (see Table 3, page 4).

- Agree. Control was performed by feeding on the machine tool. The text has been modified (see page 3, lines 129-131, lines 134-137).

- The dimensions of the joint samples in photo 2 differ due to various process parameters. Moreover, dimension markers have been added, and photos have been improved (see Fig.2 on page 5).

 

Comment 3

2.4. Testing

-Please correct the designation: (100-kN).

-The same designation of materials should be used throughout the work.

-Photo 3 shows a handle with pins. How was the grip of round samples in the testing machine? The method of attachment affects the transfer of load to the tested material.

 

Response 3

Thank you for your comment.

-The designation has been corrected (see page 5, line 209).

-The same designation of materials has been used.

- This photo has been revised (see Fig.3 on page 5).

 The samples were mounted on the machine by means of a pin driven through a hole drilled in the gripping part of the sample. In the case of this method of loading the samples with a tensile force, according to Saint Venant's rule, a uniform stress distribution is assumed at a distance from the pin equal to the maximum transverse dimension of the sample.

 

 Comment 4

3 Neural network modeling

-Before chapter 3, a chapter with the results from the basic experiment should be developed.

-Chapters 3.1. Network architecture, and 3.2. Validation and learning process should be added

-Table 2 incorrectly describes the force "Friction Force". How was the friction force determined?

-In line 212: is the correlation coefficient or R2 described? R represents the value of the Pearson correlation coefficient, which is used to mark the force and direction among variables, while R2 represents the coefficient of determination, which determines the strength of the model.

 -First, the structure of the network should be presented, then its data and description.

 

Response 4

Thank you for your comment. We agree with this comment.

-A chapter with the basic experiment has been developed (see page 5, lines 226-228, Table 4, and lines 236-240 on page 7). Thank you for pointing this out. We have accordingly revised the structure of the chapters.

-The friction force was with the correlation of air pressure (see Fig 1c on page 4).

-R-value is the coefficient of correlation. It shows how well the predicted outputs are matching with real outputs

-Agree. The text has been revised (see page 7).

 

Comment 5

5 Results and discussion

-The results should be developed better. Add more detailed analysis.

 

Response 5

Thank you for your comment. The results and discussion have been more developed (see page 12, lines 377-383).

 

Comment 6

4. Formulation of optimization task

- the objective function should be better defined and described.

- this chapter can be part of chapter 3: 3.3. Formulation of optimization task.

 

Response  6

Thank you for your comment.

-The objective function has been better defined and described (see page 10, lines 307-335 in the text).

-Thank you for pointing this out. Thank you for your comment. We agree with this comment.

 

Comment 7

6. Conclusions

Should be improved. Quantitative and qualitative evaluation of research results should be made

 

Response 7

Thank you for your comment. The conclusions have been improved (see page 16, lines 477-479).

 

Comment 8

7. Appendix A

- it is not terribly extensive. It can be placed in an appropriate place in the publication.

 

Response 8

Thank you for your comment. The tables have been inserted into the main text. (Currently, there are Tables 4 and 9 on  pages 6 and 11, respectively).

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The text has been corrected taking into account the guidelines. The final decision is made by the editor-in-chief. I wish you success in your further scientific work.