Investigations on System Integration Method and Optimum Design Method of Electro-Mechanical Actuator System

Round 1

Reviewer 1 Report

The manuscript from Zheng and coauthors described an integrated electro-mechanical 10 actuator module (IEMM) with multiple structural forms. A comprehensive literature review was included and all details regarding the design and results of the IEMM were presented. Overall, the manuscript is well-written and present the novelty. However, I feel some of the content could be removed to the supplementary materials rather than the main manuscript. For example, the thermal network model and the flow chart of optimization design. Also, figure captions could be expanded to include more details and Figure 3 was not correctly labeled. Please go over the manuscript and correct all format issues.

Author Response

Thank you very much for the reviewers' recognition of the research content. We will continue to make efforts to carry out research on IEMM.

We request that the thermal network model and the flow chart of optimization design be kept in the text to ensure the smooth reading of the manuscript. At the same time, we also checked the figure in the manuscript and made a correction to Figure 3.

Thanks again for the reviewer's comments.

Reviewer 2 Report

Comment 1: 

 The nature of this paper presents a work to optimal design of the actuators by considering all relevant aspects. Such a work includes a lot of efforts, but no new knowledge generated. In fact, the Idea of large-scale optimization design model was proposed long time ago for robots with the phrases such as "concurrent design," etc. More recently, such a kind of work was performed, see the thesis “Sudipto Shekhor Mondol, 2019. DESIGN OPTIMIZATION FOR A SUPPORT FOR THE STORAGE RING QUADRUPOLE MAGNET IN A SYNCHROTRON RADIATION FACILITY”. Master Thesis, University of Saskatchewan. Therefore, the authors need to carefully explain the generic knowledge available to the reader of this journal. 

 Comment 2: 

 Accuracy is 13% is pretty bad. Likely it is related to the optimization of off-the-shelf components with parameter estimation, discrete variable optimization problem. Please explain it. 

Author Response

Answer to question 1:

"DESIGN OPTIMIZATION FOR A SUPPORT FOR THE STORAGE RING QUADRUPOLE MAGNET IN A SYNCHROTRON RADIATION FACILITY" designed a stent for CLS 2.0 quadrupole magnet, and the natural frequency and weight of the frame are taken as the optimization objective to optimize design. The design concept and related technology have reference significance for this manuscripts. We quote it in the revised version.

But it is different from the research object and optimization goal of our manuscript. This manuscript mainly proposed variety of structures of IEMM, and a multi-objective optimization design model was built with light weight, low power loss, and high level of integration. There is no need to rebuild the whole system model under different demand indexes. The model helped achieve better comprehensive performance in the early design phase of IEMM, thus avoiding the manually iterative design and improving the optimization design efficiency. It also provides a new design idea for researchers to design IEMM.

Answer to question 2:

We thank the reviewer for pointing out this problem. The most incorrect parameter in parameter estimation is "Rotational inertia", but we didn't use it in the optimization process, so ignored it for the first time. We have deleted this parameter and corrected this part in the revised manuscript. At the same time, we will carry out robust design for the estimation method in the follow-up study, so as to estimate other parameters more accurately.

Reviewer 3 Report

This submission is well in line with the journal topics.

Globally speaking, I am impressed by the amount of work that is reported in this submission.

Unfortunately, it appears that the authors want to address too many things in a single paper. As a consequence, each point addressed looks being not sufficiently detailed against the state of the art, the proposals made or the results obtained (see the comments in the attached pdf file).

I would recommend to address mainly the part(s) that provide(s) a scientific advance and to put aside the parts that are more relative to an exert engineering work.

I feel that:

- the part about architectures is potentially interesting but it ignores two much the already published work, and the comparison does not address sufficient aspects (load path, friction, backlash/compliance, etc.). Geometrical integration could have also considered differents way to link the housing to the supporting frame with consequences on the load path (not only eye end to eye end).

- the part on the combined 3D/1D modelling and simulation is interesting. I find however its description too fuzzy to clearly identify the scientific steps forward enabling the combined use of 3D and 1 D simulation.

- the part concerning the global modelling does not show sufficiently the assumptions made, in particular for the variability or  the uncertainty of the parameters (e.g. the loss models are two simple and not related to temperature). The other means to get representation models (e.g. data mining or scaling laws) must be into consideration and compared with the proposed polynomial approach.

- the part concerning the optimization needs to be clarified. The objective, parameters vector and constraints must be clearly and explictely listed.

- the part related to the control design involves a model that is too simplistic to conclude on the closed-loop performance.

As a conclusion, I would recommend:

- to present the submission with a title and content showing that it concerns the process and its implementation for the preliminary design and decision making.

- to make a choice about the topic(s) that are addressed in detail, with a more exhaustive review of the state of the art

- to not intend to address everything in a single paper.

Finally, I am convinced that this work is really interesting and worth being disseminated, after revision to make it more understood, useful and re-usable by the readers.

This is why I suggest a major revision that mainly concerns the way to present the work and not the work itself.

Comments for author File: Comments.pdf

Author Response

Thanks for the reviewer’s insightful comments.

We have carefully read the reviewers' suggestions and have modified the manuscript accordingly. We hope that the modified manuscript can be approved by the reviewers.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comment 1:

In the title, “ ... electro-mechanical actuator module ...”, “module” should be changed to “system”.

Comment 2:

Please compare your actuator with the category of actuators based on the piezoelectric and inertia-friction actuators, see “development of a two degree ....” (Review of Scientific Instruments 77, 035112 (2006); https://doi.org/10.1063/1.2185500) in terms of accuracy, and long-range stroke.

Comment 3:

It seems that the authors use a Coulomb’s model for friction. Please explain the reason for choosing this model, given the fact that in micro-motion systems, there may be other friction models, e.g., LuGre model, that may be more appropriate, see the reference “A Novel Methodology for Comprehensive Modeling of the Kinetic Behavior of Steerable Catheters," in IEEE/ASME Transactions on Mechatronics, vol. 24, no. 4, pp. 1785-1797, Aug. 2019, doi: 10.1109/TMECH.2019.2928786.”

Author Response

Answer to question 1:

Thanks for the reviewer’s insightful comments. We have revised the title.

Answer to question 2:

Piezoelectric materials have a series of advantages such as simple mechanical structure, small dimension, multifunction, and high displacement resolution. Piezoelectric rotary-linear actuator is a kind of small and high-precision actuator. We have listed the manuscript as a reference in section 1.

If the motor part of IEMM uses sensor with the precision of 0.5°, reducer with the reduction ratio of 120 and PRS with the lead of 3mm, the motion accuracy of GIERM can reach 0.004° and the motion accuracy of GIELM can reach 35nm. The accuracy of piezoelectric rotary-linear actuator is 0.02° and 26nm. The rotation angle of GIERM can reach 360° and the linear motion stroke of GIELM is no more than 2m (mainly related to the length of PRS). However, the rated output torque and rated output speed of IEMM are much greater than that of piezoelectric rotary-linear actuator.

In addition, we believe that these two kinds of actuator are applicable in different places. Piezoelectric rotary-linear actuator is mainly applied in the field of cellular robots while IEMM is mainly applied in the field of industrial robots.

Answer to question 3:

Thanks for the reviewer's insightful comments. In this paper, multiple structural forms of IEMM and multi-objective optimization design method are proposed. The friction torque of the reducer is calculated by general formulas. When calculating the friction torque of PRS and PTRB, various types of friction torques (e.g. friction torque caused by elastic hys-teresis, friction torque caused by spin sliding and friction torque caused by differential sliding and so on) are considered. The methods and models mentioned in article "A Novel Methodology for Comprehensive Modeling of the Kinetic Behavior of Steerable Catheters" are beneficial to improve the accuracy of friction torque calculations, we have listed the manuscript as a reference in section 4.3.

Reviewer 3 Report

Although I do not agree with some replies (e.g. BS and RS have same efficiency that is proven to not be true bysuppliers catalogs), I consider that a sufficient effort has been placed by the authors to accept the revision as it is.