Research on a Novel Flux-Switching Permanent Magnet Motor with Adjustable Torque Ripple Using an Auxiliary Rotor

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors propose a novel FSPM machine with an auxiliary rotor for torsional vibration to adjust torque ripples. The manuscript content is well presented in the background. The authors carried out systematic finite element simulation calculations and used experiments for validation. The overall content is logical and well organized. The reviewer has the following suggestions.

1. In finite element calculations, mesh quality has a certain influence on the results. It can be seen from the cloud diagram in the paper that the cloud variation is not continuous. The authors can provide additional explanation in terms of mesh size and quality.

2. In the experimental tests, it is suggested that the authors perform several experiments to exclude accidental experimental errors.

3. In line 287 to 297, is Figure 10 misspelled, shouldn't it be Figure 11?

4. In Figures, it is recommended that units be added to the legend. For example, in Figure 5a, Dri = 86 mm to maintain consistency with the presentation in the text.

Author Response

Replies to Reviews

We would like to thank the reviewers for the constructive suggestions. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our research. The careful and professional attitude of the reviewers moves us very much. They are models of our scientific research. Thanks again for the professional work of the reviewers. The summary of the changes and responses to reviewers’ comments are listed below.

1. In finite element calculations, mesh quality has a certain influence on the results. It can be seen from the cloud diagram in the paper that the cloud variation is not continuous. The authors can provide additional explanation in terms of mesh size and quality.

Response: Thanks for your question. In the FEM calculations, the length-based refinement was used. And the maximum element lengths were set as 0.5mm for rotors and stator, 1mm for magnet and 5mm for other parts. In considering the size of the section of the machine, the element size is suitable. We try to change the element size of rotors and stators in the range of 50%, the calculation results have around 5.0% difference with the original value. The distribution of the flux density has nearly no difference. But the smaller element caused much more time in analyze. So that, the origin mesh settings are utilized in the FEM analysis.

As you mentioned, the cloud diagram in the paper that the cloud variation is not continuous. That was mainly caused by the rendering options in the post-processing. In this paper the rendering model is set as “Fringe”, the color in the diagram have sharped edge and the color block can be easily matched with legend. The rendering model can also be set as “Gourard”, the edges of color are blurred, and the cloud variation is continuous.

 

2. In the experimental tests, it is suggested that the authors perform several experiments to exclude accidental experimental errors.

Response: Thanks for your suggestion. Actually, the experiments have taken operation at least three times in each part. And the experiment results are averaged to exclude the accidental experimental errors. The corresponding expression are added in manuscript.

 

3. In line 287 to 297, is Figure 10 misspelled, shouldn't it be Figure 11?

Response: We apologize that we made a mistake in the manuscript. The corresponding content was revised in the manuscript. And the whole manuscript was checked to avoid the error.

 

4. In Figures, it is recommended that units be added to the legend. For example, in Figure 5a, Dri = 86 mm to maintain consistency with the presentation in the text.

Response: Thanks for your suggestion. The units were added in the legend of figures. The corresponding content are revised in the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

1. The manuscript is lacking an introduction that places the work in context and provides motivation, along with a proper conclusion, which is necessary to summarize the work and discuss the outlook.

2. It is more convincing to compare the FSPM motor before and after the improvement.

3. The paper lacks some theoretical knowledge, so it is suggested that the author add some relevant theories.

Comments on the Quality of English Language

It is noted that your manuscript needs careful editing by someone with expertise in technical English editing paying particle attention to English grammar, spelling, and sentence structure, so that the goals and results of the study are clear to the reader.

Author Response

Replies to Reviews

We would like to thank the reviewers for the constructive suggestions. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our research. The careful and professional attitude of the reviewers moves us very much. They are models of our scientific research. Thanks again for the professional work of the reviewers. The summary of the changes and responses to reviewers’ comments are listed below.

1. The manuscript is lacking an introduction that places the work in context and provides motivation, along with a proper conclusion, which is necessary to summarize the work and discuss the outlook.

Response: Thanks for your suggestion. According to your advice, the last paragraph of introduction and conclusion were revised to summarize the work and discuss the outlook. The modification is marked as red in the manuscript.

 

2. It is more convincing to compare the FSPM motor before and after the improvement.

Response: Thanks for your suggestion. When the a-rotor and m-rotor rotate simultaneously, the motor output are nearly the same as the conventional FSPM motor. The speed difference is zero. So that the two V-shape ridges of harmonic compound in the Figure 4 are not appeared in the case of conventional FSPM motor. This is the biggest difference in motor performance before and after improvement.

 

3. The paper lacks some theoretical knowledge, so it is suggested that the author add some relevant theories.

Response: Thanks for your advice. For the reasons of the motor torque output changes caused by magnetic circuit characteristics, we have made a supplementary theoretical explanation, and added in the revised manuscript. At present, this study focuses on the qualitative and trend studies of this new idea, and some quantitative studies are lacking in calculation. This is indeed the deficiency of the current research. We will carry out more refined research in the follow-up study. Thank you for your valuable advice.

Reviewer 3 Report

Comments and Suggestions for Authors

 

In the attached file, I have some questions and observations.

Comments for author File: Comments.pdf

Author Response

Replies to Reviews

We would like to thank the reviewers for the constructive suggestions. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our research. The careful and professional attitude of the reviewers moves us very much. They are models of our scientific research. Thanks again for the professional work of the reviewers. The summary of the changes and responses to reviewers’ comments are listed below.

1. In line 16 of the summary, where it mentions “In some cases, the amplitude of torque...”, it is necessary to specify what those cases are for greater clarity and precision.

Response: Thanks for your suggestion. When the speed difference of two rotors is six times of the m-rotor speed, the torque ripple waveform overlapped. In this case, by control the phase difference between two rotors, the amplitude of torque ripple can be adjustable. The modification is marked in red in the revised manuscript.

 

2. Compared with conventional methods used to generate vibrations superimposed on the main torque, what are the main advantages of the proposed method from different perspectives, such as motor complexity, cost, among other aspects?

Response: Thanks for your question. The conventional methods to generate vibrations superimposed on the main torque is to install a vibration generator on the end of motor axis or actuator. And an energy converter is used to drive control the vibration generator. In the aspect of structure complexity, the conventional method has the bigger size in the axial direction, all functional part is connected in series. The prototype in this paper has smaller size in axial direction with a little wider in the radial direction. Overall, the method proposed in this paper is slightly more expensive because it requires a servo motor to control the a-rotor and more complex control methods. From the perspective of flexible manufacturing, the method proposed in this paper can provide more flexible vibration output to support different manufacturing needs.

 

3. The use of torque harmonics with the proposed system: Could it cause excessive wear on motor components, such as bearings, or induce harmful resonances?

Response: Thanks for your question. Unlike the vibration in axial direction, the torque harmonics is generated in the rotating direction, and this degree of freedom of bearing is free, it would not be harmful for the bearing. This direction of torque harmonics may cause resonances in some cases. The torque ripple is an exciting source, when it is consistent with the natural frequency of the structure, it will arouse resonance. So, it is necessary to understand the modal characteristics of the structure in advance through calculation or testing. The users could control the motor generate the torque ripple at the appropriate frequency to avoid resonance.

 

4. In line 142, it is mentioned: "The stator and m-rotor geometric parameters in Table 1 are the optimized results without a-rotor." Which specific magnitude or magnitudes are referred to when talking about "optimization"?

Response: Thanks for your question. The optimization is carried out on the parameters of rotor salient width, stator salient width and permanent magnet thickness. The purpose of this optimization is to reduce the amplitude of torque ripple. After optimization the amplitude of torque ripple cause by m-rotor is about 5% of the rated torque. The optimization introduced here to show that the motor original size is appropriate structure. Based on this structure, the rotor was transformed into dual rotor co-axis placement to perform adjustable torque ripple. In fact, it is also practicable without optimization, and the amplitude of the torque ripple will be a different value.

 

5. In section 3, "Influence of geometrical characteristics of a-rotor on torque ripple," the analysis focuses on how geometry affects the amplitude of the harmonics. Is there a specific combination of geometric parameters that optimizes torque ripple? Has any particular optimization algorithm been used for this purpose?

Response: Thanks for your question. The effect of structural geometry on harmonic amplitude is a multi-parameter coupling optimization problem. Through reasonable parameter configuration, the harmonics amplitude can be nearly the same value. Further, the phase reversal of the two harmonic waveforms is realized through the phase control of the two rotors, so as to achieve the harmonics eliminating. This involves the multi-parameter coupling optimization problem, which is exactly what we are studying. Thank you very much for helping us to clarify the direction of our research.

 

6. In lines 291, 292, 293, and 296, where figure 10 is mentioned, should it instead refer to figure 11?

Response: We apologize that we made a mistake in the manuscript. The corresponding content was revised in the manuscript. And the whole manuscript was checked to avoid the error.

 

7. In figure 11, there is a harmonic at 30º whose amplitude does not depend on the speed difference. What is the reason for this?

Response: Thanks for your question. The harmonic at 30 º depends on the m-rotor speed. It is the 12^th order of m-rotor speed in electrical angle. And it is one of the main parts of the torque ripple generated by m-rotor. When the m-rotor stops, this harmonic does not exist.

 

8. Is there a missing temporal analysis in figure 12? When the initial angle is 40 degrees, does the amplitude of the torque of the 6th harmonic remain constant over time? Does the same happen for the rest of the initial conditions?

Response: Thanks for your question. In the analysis of Figure 12, the speed difference of two rotors is always six times of the m-rotor. The speed of the two rotors is constant during test. Therefore, the periodic characteristics of all torque ripple is the same for all test. The main variation is the amplitude of the harmonics. Therefore, its periodic frequency is not the focus of our analysis. We mainly analyze the harmonic amplitude of torque ripple, which is the result of Figure 12.

When the initial angle is 40 degrees, the amplitude of the torque of the 6th harmonic remains constant over time. The initial angle is the start condition of two rotors. Before start, the m-rotor is aligned to the coil A1, the initial angle is the angle between the a-motor salient and m-rotor salient. After start, the speed difference of two rotors is always six times of the m-rotor speed. So that, the common cycle of two rotors is equal to the m-rotor cycle. Their periodic character is stable. Therefore, the amplitude of their torque ripple is also stable. And the torque ripple characteristics is also stable for the rest of initial conditions.

 

9. For a given geometry, the amplitude of the harmonics, according to figures 11 and 12, mainly depends on the speed difference and the initial angle. Are these two variables independent of each other, or could they be functionally related in some way

Response: Thanks for your question. The speed difference and initial angle are two independent variables. Speed difference is the variable between m-rotor and a-rotor. The two rotors could rotate independently. In the prototype, the m-motor was driven by the winding in stator, the a-rotor was driven by a servo motor. The initial angle was used to define the phase difference between the two rotors before start. When the torque ripple generated by speed of m-rotor and speed difference by two rotors are not overlapped, the initial angle has nearly no influence on the torque harmonics. When the torque ripple waveform overlaps, the initial angle could influence the amplitude of the overlapped harmonic waveform.