A Data-Driven Approach for Cutting Force Prediction in FEM Machining Simulations Using Gradient Boosted Machines

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper addresses a challenging subject and has a good scientific level. However, some remarks have to be done:

- The research is pure theoretical. It concerns a very simple, ideal, cutting process (the free orthogonal cutting). There is not any connection with a real cutting process. No physical validation!

- The research methodology is not properly introduced.

- The research is more about modeling the material behavior than about the cutting process / cutting force prediction, which is a sample of application.

- We have a table with material properties, but this material is not specified, it is an immaginary one?

- Cutting force prediction in real cases requires to consider the cutting regime, here we have no mentions about it.

- The discussion is about FEM simulation (??? - unusual association between "FEM" and "simulation") or, perhaps, about FE modeling or, maybe, FEA?

- Fig. 3 is not referred in the text.

- Industry 4.0 concept would require a direct data-driven prediction of the cutting force rather than the application of FE modeling to do this.

Comments on the Quality of English Language

The English is good. However, there are some inappropriate expressions - e.g. "the Finite Element Method of machining processes", "improve the material property estimation using cutting processes" (???) etc.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

A very interesting article with an interesting approach on research issues. The methodology of conducting them is correct in my opinion and the results are correctly discussed. Despite the fact that the J-C model is widely used, the authors showed a non-standard approach in determining the cutting forces using the python language and its numerous libraries. I rate the value of the article highly. Nevertheless, I miss the given value of the friction coefficient (the authors write about it in the summary) and also the given value of the damage initiation parameters in the model.

As the authors noted, the thermo-mechanical aspects of the workpiece material, technological parameters and others also influence the result.

In conclusion, the article is very interesting bringing a lot of useful knowledge to the area of machining modeling.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The article topic is timely as it uses FE simulations together with several machine learning approaches with the aim to reduce time to simulate cutting and thrust forces. Five machine learning models are trained and compared. It is found machine learning is suitable for the task, can make predictions with a good accuracy with the Ligth Gradient Boosting Machine method being the best performer under the given conditions. 

The article is written-well. Its structure and language are both judged to be excellent. The problem formulation, descriptions are sufficiently detailed. Findings are evaluated correctly and credibly. Publication is advised.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf