Investigation of the Applicability of Acoustic Emission Signals for Adaptive Control in CNC Wood Milling

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents an interesting approach to the application of acoustic emission signals in CNC wood milling. The structure of the paper is satisfactory, and the thought process can be easily followed. The findings are also interesting. Addressing the following comments should further improve the study:

1. Line 35-60. These statements are not supported with references. Please add some citations.
2. Line 61. This line itself contains 11 references. These should be evaluated separately.
3. Line 161-162. It is not clear, how these parameters varied.
4. Line 174-177. Interesting, can you explain further?
5. Line 202. Some clarification should be added, why are these chosen?
6. Line 218. Phase naming and sensor naming are same. (cont. in Comment 11)
7. Figure 6. To help the easier understanding, above the columns could be written the two types of feed rate, left to the rows could be written the two types of spindle speed.
8. Line 239. The meaning of these markings should be explained further.
9. Figure 8. 1 and 2 could be marked as 3 and 4, since they are related to Figure 6.
10. Table 4 should be moved to 2.3, since it clarifies the cutting setup. The introduced markings of the setups could be used later in Figure 6-8 for example.
11. Furthermore, the Sensors, the measurement positions and the setups are all marked with latin capital letters (A, B…), which is confusing. Consider using different markings for these three things (S1 and S2 for the sensors, I, II, III for the measurement positions perhaps).
12. Line 258. If the tool revolutions are fixed, how the rotation speeds are different?
13. Line 284. The authors mention four of the six possible comparisons. This arises the question, why is the C- D comparison not showed?
14. I think, Section 4 should be separated as: 4. Discussion, 5. Conslusions. The conclusions should contain a short summary of the study, which length should not exceed 1 page. The discussion section should contain all the evaluation of the results in Section 3. Also, this should be supplemented by the description of the effect of the changing cutting speed and feed as well on the AE. Now, it is mainly discussed, that there is an effect, but the exact nature of this effect is not detailed.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study investigates the applicability of acoustic emission signals for adaptive control in CNC wood milling, analyzes the relationship between acoustic emission signals and cutting parameters through experiments, and explores the role of signal acquisition in improving machining efficiency and tool life as well as its limitations. The obtained results helps to improve machining efficiency and extend tool life.

The following comments should be addressed before considering of publication:

1. In Section 2.1, the description of the method of fixing the workpiece is missing from the manuscript.
2. What is the exact location of the AE sensor and does the location have an effect on the measurement results?
3. In Section 2.3, experiments were conducted with only two rotational speeds and feed speeds, although this parameters can show the trend of the AE signals as a function of the parameters, it cannot fully reflect the actual machining conditions.
4. Please explain in the text what the red circles marked in Figure 4 represent.
5. The different processing states are marked in Figure 7, what are the evaluation criteria for significant/ weak/ negligible impacts?
6. The paper lacks how to use the acquired AE signals and verify the correctness of this method through machining
7. The article does not discuss the stability and immunity of the AE signal, which is critical in real industrial environments.
8. Please make the conclusion section a separate section.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper explores the use of acoustic emission (AE) signals for real-time monitoring and adaptive control in CNC medium-density fibreboard (MDF) wood milling. It shows that AE signals are altered with changing spindle speed and feed rate, indicating potential for process monitoring. The two-channel sensor test arrangement with a 5-axis machine provides useful information, though the study is compromised by limited parameter range and potential material and sensor variations. While potential adaptive control applications are suggested, further work needs to be done in overcoming current limitations and verifying the method on broader industrial scales. However, the following elements needs consideration in order to improve the quality of the manuscript:

• In the Introduction section, research contributions are lost. The research contributions should be highlighted (for instance using bullet points) for the readers to understand the overall contribution to the existing body of knowledge.
• An explicit novelty statement is missing from the introduction section. An explicit novelty statement in clear and concise manner help the readers understand the novel element of the research.
• What are the advantages of side milling over other milling approaches for AE signal assessment?
• How does the CNC machine's rotational frequency and spindle power influence AE signal features?
• Why were two AE sensors used instead of one, and how does their spatial distribution influence data quality and signal interpretation?
• What is the significance of applying FFT on 20 ms windows of signal in this context, and how does it allow cutting-edge engagement to be distinguished?
• Why is MAX_sig chosen as the main AE parameter over others, and what does it reveal about tool condition or material interaction?
• How is the nature of the AE signal pattern (i.e., discrete bursts vs. constant noise) connected with the physical behavior underlying wood milling?
• How can AE data be used to quantify tool wear or cutting quality at different feed rates and tool revolutions?
• How does the mounting and calibration process (pen-test) influence the reliability and comparability of AE data from one trial to another?
• What are the challenges in analyzing AE signals in wood-based materials compared with metals, mainly with regard to repeatability and noise?
• How may the split of cutting into stages A, B, and C serve to enhance AE signal progression understanding during tool engagement?
• With the huge difference in AE signal sum values among conditions (e.g., conditions A and C have ~2× B's and D's amount of data points), how was comparability of conditions preserved despite different sample sizes?
• How was validation of AE signal normalization performed? Was there risk involved in changing signal characteristics during normalization, particularly given the radically reduced variance (e.g., SD ~0.006 after normalization)?
• The study uses both non-parametric (Kruskal–Wallis, Dunn's test) and parametric (ANOVA, Tukey HSD) methods. Given that the Shapiro–Wilk test showed non-normality, why were parametric ANOVA and Tukey HSD utilized without transforming data or employing a full non-parametric analogue?
• Were effect sizes (e.g., η² for ANOVA or r for Dunn's test) estimated to be reported with statistical significance? Without them, how do we assess the practical significance of measured AE signal difference?
• The argument that halving speed causes the cutter to engage all cutting edges is based on visual trends in AE signals. Was this hypothesis quantitatively confirmed (e.g., number of peaks, frequency analysis), or is it qualitative?
• How were stages A, B, and C separated from one another in the AE signal logs? Were these separations on the basis of real process events, or random time periods?
• Were AE sensor gains, background noise levels, and machine vibration profiles the same in all 24 measurements (6 conditions × 4)? If not, how was measurement bias avoided?
• How were overlapping AE signals between successive milling steps handled in analysis to ensure that envelopes measured do indeed correspond to the contribution of a single cutting edge?
• Was any interaction analysis (e.g., two-way ANOVA or MANOVA) performed to determine whether speed and feed rate both influence AE signal behavior, rather than independently?
• Because AE measurements are highly sensitive, how was the assurance made that "dominant packages" in AE envelopes are genuine cutter–material interactions rather than machine noise, resonance, or backlash?

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I am satisfied with the corrections made.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have addressed all  comments thoroughly. The revised manuscript meets the journal's requirements, and I recommend it for publication.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed all the comments.