A Novel Method for Tool Identification and Wear Condition Assessment Based on Multi-Sensor Data

Round 1

Reviewer 1 Report

General observations

The paper content fits the aims and scope of the journal.

The references appear to be appropriate.

The language is generally good in the paper

Verify that the author names are correct and that the use of capital letters is OK "Yirong liu, Jiechao Lv and Xiaoli Wang"

In the body text use only family names of the authors for referencing

It is surprising that not all authors approve the publication

Specific requests

Line 97 Provide the model and make of the sensors used "Since the tool must be detected quickly and the acquisition frequency of the sensor is required 97 to be high, the sensor adopts three-axis force sensor 1303m-160, three-axis vibration sensor ct100slc 98 and acoustic emission sensor ae144sa40 BNC."

Line 103: The sampling frequency used 100kHz is low to catch many of the fracturing events in metal. This must be motivated better. A normal range for the AE sensors are 50 to 800 kHz and at 100 kHz, the full range is not utilized.

Figure 8 The x-axis title is confusing. For instance 2e-5s???? with the scaling 10e-5. The exponent does not match for all cases and the factor 2 does not make sense.

Author Response

Reply to Reviewer’s Comments

Reply to Reviewer #1

    Firstly, we would like to thanks the reviewer for the positive and constructive comments. According to your comments, we have checked our manuscript carefully. Some grammatical & language errors and other inexact expressions in the manuscript have been corrected. The important changes in our revised manuscript have been marked “in Red”. Thanks again.

Reviewer’s Comments:

General observations:

The paper content fits the aims and scope of the journal.

The references appear to be appropriate.

The language is generally good in the paper.

Verify that the author names are correct and that the use of capital letters is OK "Yirong liu, Jiechao Lv and Xiaoli Wang".

In the body text use only family names of the authors for referencing.

It is surprising that not all authors approve the publication

Response: Thank you for your approval and comments. All of our authors acknowledge and agree that the article is published in this publication. It is possible that some co-authors did not reply to the confirmation email due to the failure of the school's email system at the time of submission. If necessary, we can further confirm at any time.

Specific requests:

Point 1: Line 97 Provide the model and make of the sensors used "Since the tool must be detected quickly and the acquisition frequency of the sensor is required 97 to be high, the sensor adopts three-axis force sensor 1303m-160, three-axis vibration sensor ct100slc 98 and acoustic emission sensor ae144sa40 BNC."

Line 103: The sampling frequency used 100kHz is low to catch many of the fracturing events in metal. This must be motivated better. A normal range for the AE sensors are 50 to 800 kHz and at 100 kHz, the full range is not utilized.

Response 1: Thank you for your professional review opinion. There is indeed an error in our original draft. The acoustic emission signal of tool breakage is about 80 ~ 400kHz, so the AR sensor model we proposed is ae144sa40 BNC, and its sampling frequency is 50 ~ 500KHz. The signal acquisition card adopts ni-pci-mio-16e-1 data acquisition card, and its maximum sampling frequency is 1.25MHz. We have amended and supplemented it on line111 of the revised version. Thanks again!

Point 2: Figure 8 The x-axis title is confusing. For instance 2e-5s???? with the scaling 10e-5. The exponent does not match for all cases and the factor 2 does not make sense.

Response 2: Thanks for the comments and valuable suggestions. The x-axis title 2e-5s in Figure 8 of the original manuscript represents 0.00002 seconds. As you said, it is really confusing to use 2e-5 to represent 0.00002. We have changed the x-axis title in Figure 8 and changed the picture to a higher definition one.

At last, special thanks to you for your comments again!

Author Response File: Author Response.pdf

Reviewer 2 Report

Review of applsci-770768-peer-review-v1: “A Novel Method for Tool Identification and Wear Condition Assessment Based on Multi-Sensor Data”

The subject of the paper is relevant with the topics of the journal. Its significance with respect to their industrial value is very good. The number and the quality of the references selected are very good.

I would suggest the authors to incorporate the following in a revised version of the paper:

• Line 38, CNN should be clarified
• In figure 1, rectangles should be used for the processes. There is not any decision making that needs other geometrical shapes.
• Figures 1 and 2 can be included in one figure, with the appropriate time axes
• The descriptions on top of figures 2, 3, 4 and 5 should be increased for better readability. The labels of x-axes and y-axes should be increased too, for better readability. Use vector software if necessary.
• The labels of x-axes, y-axes and the data descriptions should be increased, for better readability in figure 8, 9, 10, 12, 14 and 15. Use vector software if necessary.
• Figure 13 is missing.
• In conclusions, a comment about the performance of the proposed methodology in a commercial application could be included. Can this methodology be applied and work commercially, and under which circumstances?

My proposal to the editor is to accept the paper with minor revisions

Author Response

Reply to Reviewer’s Comments

Reply to Reviewer #2

Firstly, we would like to thanks the reviewer for the positive and constructive comments. According to your comments, we have checked our manuscript carefully. Some grammatical & language errors and other inexact expressions in the manuscript have been corrected. The important changes in our revised manuscript have been marked “in Red”. Thanks again.

Point 1: Line 38, CNN should be clarified

Response 1: Thank you for your careful review. CNN means revolutionary neural networks. We have revised it on line 39 of the revised version.

 

Point 2: In figure 1, rectangles should be used for the processes. There is not any decision making that needs other geometrical shapes.

Response 2: Thanks for the comments and valuable suggestions. According to your suggestion, we have corrected Figure 1.  

 

Point 3: Figures 1 and 2 can be included in one figure, with the appropriate time axes.The descriptions on top of figures 2, 3, 4 and 5 should be increased for better readability. The labels of x-axes and y-axes should be increased too, for better readability. Use vector software if necessary.

 Response 3: Thanks for the comments. Figure 2, 3, 4 and 5 show the preprocessing process of the spindle current data. The most significant feature is that the characteristics of the original data are retained while the number of data is reduced. According to your opinion, we have replaced the picture with the '.emf' format picture that is more suitable for Word, which improves the readability of the article.  

 

Point 4: The labels of x-axes, y-axes and the data descriptions should be increased, for better readability in figure 8, 9, 10, 12, 14 and 15. Use vector software if necessary. Figure 13 is missing.

 Response 4: Thank you for your careful review. First of all, figure 13 is caused by our numbering error, which we have fixed in the revised manuscript. The x-axis in Figure 8 represents the sampling time, and its unit is 0.00002s. We use 2e-5 to represent 0.00002 in the first draft, which is really confusing. We have revised the x-axis label in Figure 8 to make it easier to understand.  The x-axis in Figure 9 indicates that the wear degree of the tool is gradually deepened from 1 to 315, which is an abstract representation, so there is no unit. The y-axis indicates the value after normalization of variance and average value to the [- 1,1] interval.The x-axis in Figure. 10 has the same meaning as that in Figure. 9, which indicates the wear degree of the tool. The y-axis of Figure 10 represents the wear abstraction flute value, so there is no unit.Figure 12 shows the training process of Elman neural network, and the x-axis of both figures shows the number of iterations. In Figure. 12 (a), the y-axis of the three graphs shows the gradient value, generalization ability and learning rate from top to bottom, and Figure. 12 (b) shows the training error.Figure 14 (Figure 13 in the revised version) and Figure 15 (Figure 14 in the revised version) of the original show the change of wear amount flute with tool wear degree.We have added a description of the above pictures in the revised version and marked them in red. Thanks for your suggestions again. 

 

Point 5: In conclusions, a comment about the performance of the proposed methodology in a commercial application could be included. Can this methodology be applied and work commercially, and under which circumstances? 

Response 5: Thanks for the valuable suggestions. The tool wear evaluation method proposed in this paper is very suitable for workpieces that need to be processed with a variety of tools. It can monitor the tool wear in real time and avoid the damage of workpieces caused by tool breakage or excessive wear or even the production accidents caused by the damage of machine tools. At the same time, it can avoid the machine stop caused by the excessive wear of the cutting tool, so as to improve the utilization rate of the machine tool and the overall production efficiency. In the light of your comments, we have added new information about the business implications of this approach in the conclusions.  

At last, special thanks to you for your comments again! 

Author Response File: Author Response.pdf

Reviewer 3 Report

I would like to suggest you improve the introduction and try to use some new research works and add a summary of new methods which have been used for tool wear monitoring in the machining process.

The authors investigated a novel method for tool identification and wear
3 condition assessment based on multi-sensor data. The abstract is unmatched with the conclusion.

Author Response

Reply to Reviewer’s Comments

Reply to Reviewer #3

Firstly, we would like to thanks the reviewer for the positive and constructive comments. According to your comments, we have checked our manuscript carefully. Some grammatical & language errors and other inexact expressions in the manuscript have been corrected. The important changes in our revised manuscript have been marked “in Red”. Thanks again.

 

Point 1: I would like to suggest you improve the introduction and try to use some new research works and add a summary of new methods which have been used for tool wear monitoring in the machining process.

Response 1: Thanks for the comments and valuable suggestions. According to your opinion, we have added references [13] and [14] in the introduction to enrich the introduction of the latest research on tool wear monitoring in this paper.

  

Point 2: The authors investigated a novel method for tool identification and wear 3 condition assessment based on multi-sensor data. The abstract is unmatched with the conclusion.

Response 2: Thanks for the comments. We have rechecked the abstract and conclusion, and corrected the mistakes.  

At last, special thanks to you for your comments again!

Author Response File: Author Response.pdf