Modeling of Cutting Force in the Turning of AISI 4340 Using Gaussian Process Regression Algorithm

Round 1

Reviewer 1 Report

The manuscript does not follow the standard structure of a typical metal cutting manuscript.  The follwoing needs to be corrected prior to re-submission:

1. Materials and Equipment/Apparatus section needs to be created;
2. Computation section needs to be created to explain the methods used:
3. Experimental Methods section needs to be added;
4. Results of experiment and computation needs adding;
5. Discussion followed by conclusions plus a well stocked refernce section of at least 50 references in this well researched field.

The manuscript also needs a thorough check of grammar and spelling too.

Author Response

Dear respected reviewer, 

Thank you for your time and efforts in reviewing our manuscript. A point-by-point response is attached for your kind review. 

Regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

The knowledge of cutting forces is key to nearly all machining operations. Thus, cutting forces, especially for turning processes have been researched for decades. This paper does not add anything significant to the state of the art. Furthermore, it does not adhere to established naming conventions, making it very hard to follow even for an experienced reader.

It is absolutely unclear, why this well-researched field requires the application of machine learning algorithms to characterize the cutting forces. As the authors state themselves "GPR is feasible for small size dataset, nonlinear, complex and high dimensional regression problems". In most cases, cutting force modeling can be related to linear models that are neither complex nor of high dimensionality. Thus, no sophisticated calculation techniques are required. Moreover, GPR is the only technique that is (briefly) explained, leaving out details about SVN and ANN that are used for comparison.

Nevertheless, even assuming that the approach was valid, the research design is completely inadequate, as it completely relies on data of other researchers (which is not accessible to the reviewer) and does not provide any information of relevant parameters or setups. Diagrams are few and lack vital information (axis labeling?).

Due to these serious flaws, the paper is rejected.

Author Response

Dear respected reviewer, 

Thank you for your time and efforts in reviewing our manuscript. A point-by-point response is attached for your kind review. 

Regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

1.- Line 61: A full stop is missing at the end of the line.

2.- Line 56: Start with lowercase "Feed" and "Tangency".

3.- Line 57: Start with lowercase "Cutting" and "Velocity".

4.- Where is the model of equation (2) obtained from? Reference the source or explain where it is obtained from.

5.- Explain or detail references [16] to [19].

6.- Describe and detail the SVM and ANN models used in the predictions: configuration, equation, structure, functions used, etc.

7.- Why are the experimental data of reference [23] used, and not your own?. Do you have permission to use these data?.

8.- Expand the references in the introduction section, for example:

a) Surface finish monitoring in taper turning CNC using artificial neural network and multiple regression methods, Procedia Engineering 63, 2013, pp. 599-607

b) Multi-sensor data fusion for real-time surface quality control in automated machining systems, Sensors 18 (12), 2018, pp. 4381

9.- ¿Computational cost of each method?.

Author Response

Dear respected reviewer, 

Thank you for your time and efforts in reviewing our manuscript. A point-by-point response is attached for your kind review. 

Regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 4 Report

Fig.1 (line 51) - the picture looks a little distorted.  what does that mean v? and what is the difference between v and workpeace rotation?

line 57 - there should be v_c not V_c

line 155 - Results of what?

line 174 - the units are not listed in the table 1, it would be top beneficial to understand the text above. Trials are mentioned in the text above the table, but there is not described what such a trial looks.

line 175 - the units are not listed in the table 1, it would be top beneficial to understand the text above. 

line 182 and 184 - Graphs do not describe axes, units are not listed. The selected spectrum is not suitable, because the graphs are confusing due to this. It is not clear from the text why the forces change so much within the individual trials.

Based on previous interpretation, the graphs are incomprehensible.

The article is interesting and current, only some descriptions are unclear and confusing and the reader gets lost in the text.

Author Response

Dear respected reviewer, 

Thank you for your time and efforts in reviewing our manuscript. A point-by-point response is attached for your kind review. 

Regards, 

Authors

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Acceptable

Author Response

Dear respected reviewer, 

Thank you for your efforts and time in reviewing the manuscript. 

We believe your comments have improved our manuscript substantially. 

Thank you once again, 

Regards, 

Authors

Reviewer 2 Report

Although the authors added some detail on the methodology and experimental data they used for their study, the manuscript still has serious flaws (e.g. naming convenctions, use of non-SI units, to name but a few).

Moreover, the reviewer still sees no relevance in applying this methodology to predicting cutting forces in turning.

Author Response

Dear respected reviewer, 

Thank you for your efforts and time in reviewing the manuscript. 

We believe your comments have improved our manuscript substantially. 

Attached is a point to point response. 

Thank you once again, 

Regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

• Figure 1 needs to be improved. The large cylinder will be closed behind with a circumference. The reference lines of dimension D2 must be tangent to the cylinder. Put "Fr" next to "Radial force". The angular velocity is "n", "Vc" is tangential to the cylinder contour. The cylinder perspective is not correct, it seems squashed. Improve the figure, the dimensions with arrows, and put all the variables.
• Line 13: "suport vector machine" and "artifical neural networks" must be in lowercase, and and the abbreviation in capital letters.
• Introduce some reference in SVM and ANN where the method description can be reviewed.
• Describe the ANN structure (number of neurons and layers) used. 
• Check the numbering of tables and figures: old table 1 is now table 3, and old table 2 is now table 4. Write correctly the references of tables and figures in the text.
• Line 56:  "cutting speed" is more correct than "cutting velocity".
• Line 56: Describe that Ff has the same direction that the tool feed rate (Vf).
• In line 57, line 58 and equation 2 there are three abbreviations to refer to the cutting speed: vc, V, and v. Unify them, all the same.
• Line 57: Describe Fr in radial direction to cylinder.
• In table 2 column headings, second words must be lowercase: "Cutting speed", "Surface roughnes", etc.

Author Response

Dear respected reviewer, 

Thank you for your efforts and time in reviewing the manuscript. 

We believe your comments have improved our manuscript substantially. 

Attached is a point to point response. 

Thank you once again, 

Regards, 

Authors

Author Response File: Author Response.pdf

Reviewer 4 Report

The article after the repair is much clearer. The processed topic is interesting and current, and certainly beneficial for practice.

Author Response

Dear respected reviewer, 

Thank you for your efforts and time in reviewing the manuscript. 

We believe your comments have improved our manuscript substantially. 

Thank you once again, 

Regards, 

Authors