Vibration Resonance Spectrometry (VRS) for the Advanced Streaming Detection of Rotor Unbalance

Round 1

Reviewer 1 Report

Dear authors,

Your paper is interesting and the topic presented follows current trends. The article, however, has some faults. Furthermore, some editing is required.
The research and its design are not fully clear and some significant problems were detected. The methods (experimental) used were described but not in detail.
Therefore, some additional corrections must be done for the article to be published as professional scientific work.

1.    The article is not always clearly written and easy to follow. Suggest, if possible to introduce some more visual help in the subchapters with a lot of text information.
2.    The authors give relevant references which are linked to their study. However, the number of references and especially evaluation in the introduction is limited and not fully presenting the state of the art of the field.
3.     The abstract is well written introducing the basic overview of the paper. It is also written in a way that even a person not familiar with the topic can understand what the authors are proposing in their research. The reviewer would suggest not including any abbreviations in the abstract for clarity reasons(ML, MSET, VRS). Moreover, it is suggested to point out the novelty of the study in a strong manner.
4.    The introduction requires some improvements. The purpose of the introduction is to present the problem of the article and clearly present the overview of the state of the art in case of the topic and presented later and methods. Some problems noticed:
a.    First paragraph- there are multiple works referenced [1-6] as examples of use the of accelerometers. Some of them are by the same author. Suggest to live multiple examples but give examples from different fields showing that this is a universal way of performing such measurements. Missing fields:  application in cooling and ventilation especially- e.g. the influence of mechanical ventilation and cooling systems on vibrations of high precision machines; application in quality control (example also in point b); data acquisition for simulation (FEM) models and in material science for non-invasive measurements of ultra-lightweight composite materials etc.
b.    The background of the paper in the introduction is not sufficient. More than half of the introduction is taken by presentation what is author's solution and some information which should be placed in materials and methods, not in the introduction. Some way to improve- after introducing accelerometer measurements and problems associated with acquisition parameters authors go straight to the introduction of background for their solution. It is advised also to present alternative measurement techniques, especially none contact ones, with may be better suited for such measurements e.g the use of 3D laser vibrometers in case of machines' dynamic state evaluation or quality control (e.g DOI: 10.3390/s23031263) and video motion amplification (https://doi.org/10.1016/j.measurement.2022.112218, https://doi.org/10.1016/j.ymssp.2016.08.041). It must be pointed out that those non-contact methods are also especially well-fitted for rotational equipment measurements.
c.    Please introduce what is your aim, the scope of the study and especially the novelty in a comprehensive and informative way. Currently, it is not fully clear.
5.    Chapter 2 some comments: Fig 1 and 3 suggest moving the legend outside the graph on the left side so it is not masking the plots. This way the figure will be much more clear. No additional request to the text and information
6.    Chapter 3 is acceptable.
7.    Chapter 4. Comments:
a.    Fig.5 “testlab signal collector” This phrasing is not fully professional. Use rather “signal acquisition system”.
b.    No information on the experimental setup. What are the sensors and their parameters? The microphone and accelerometer look like PCB ones but what models and parameters, what is the weight of the tested system and the accelerometer (mass may influence the dynamics of the system)? What is the acquisition system (looks similar to the Muller BBM MQ system)? What is the software? What are signal acquisition parameters? How was the accelerometer attached? No information on speed control design. Etc…That information are necessary in case of someone wants to cross-check the results.
This is basically the materials and methods chapter and it needs to be improved significantly. Requires major revisions. This is the weakest element of the paper.
8.    The results presentation and explanation are adequate. When using a similar style of graphs (e.g. excel ones) please unify the sizes, fonts etc. This chapter is also a little messy in case of editing (different font sizes etc.).
9.    Conclusions are acceptable. However, an emphasis on novelty would be profitable. 

Conclusions
The article has the potential to be interesting but after corrections. I have marketed it for major revision especially due to the very basic introduction and even more problematic lack of “materials and methods” chapter with detailed information on experimental setup and equipment.
In the introduction, additional state-of-the-art analysis of modern modal analysis testing techniques has to be incorporated in the introduction.
 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

In this paper, a combination of VRS pre-processing and an ML signal-processing technique called Multivariate State Estimation Technique (MSET), is applied for the detection of the unbalance state of a fan model.

The paper is well organized and written and it does show some merits in the innovation of anomaly detection. The reviewer recommends it for publication but suggests the authors consider the following remarks and make necessary revisions accordingly.

 

1.         There are some merits is this combined technology, however it does not show excellence in rotor unbalance detection, compared to conventional tracking 1x peak in frequency domain. Using this example to demonstrate the “new” technology is not convincing enough. Rotor unbalance, especially for a single fan, can be easily detected and quantified, even life-predicted.

2.         The superiority of the derived approach over tracking 1x peak lies in the predictive monitoring. In this regard, the reviewer suggests the authors to include in more detail about those metrics of TTL, CF, and their calculations etc.

3.         As far as prognostics is addressed, people is expecting some descriptions about failure threshold, defect severity estimation and finally RUL, etc. The present paper is more likely about unbalance detection rather than prognostics because no time factor is included in the health monitoring. The authors may need to modify the title to suitably reflect the paper’s content.

4.         The illustrated example has 4 sets of measurements, tri-accelerometer and one microphone. Are they equally important? It seems to the reviewer that the radial vibration may provide information enough for unbalance detection and analysis. The authors need to analyze the weighting of these 4 sets of signals and how the authors used them in the process.

5.         The caption of Fig. 7 is too lengthy.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear Authors,

At this time I do not have any further suggestions. The paper can be considered for publication.