Towards System-Level Simulation of a Miniature Electromagnetic Energy Harvester Model

Round 1

Reviewer 1 Report

This paper presents a system-level co-simulation of a miniature electromagnetic energy harvester model. This kind of research is useful to this field. However, the reviewer suggests the authors to make clear their contributions to the field.

1. From 20 years ago, many publications reported system-level modeling. Even in the past 3 years, several advances have been reported as listed below,

[1] Circuit Simulator Compatible Model for the Ring-Dot Piezoelectric Transformer, J. Microelectromech. Syst., vol. 32, no. 1, pp. 103-116, 2023.

[2] An Efficient Macromodel Extraction Strategy for Packaged MEMS Thermal Wind Sensor Considering Multiphysics, IEEE Sens. J., vol. 23, no. 6, pp. 5879-5891, 2023.

[3] Finite element compatible matrix interpolation for parametric model order reduction of electrothermal microgripper, J. Comput. Des. Eng., vol. 8, no. 6, pp. 1622–1635, 2021.

What is the major difference between them? What is the advantage of this presented method, compared with them? Can the presented method be used for the modeling of other MEMS devices?  

2. In Fig. 17, the magnet height is in the range of 1-1.3 mm. This range is the selection of matrix interpolation, thus it usually has higher accuracy. Could you show more results out of the range?

3. Are the FOM results of 3D structure obtained by scaling the mesh elements? If so, please show the re-meshed FOM results of 3D structure.

4. The title of the manuscript is about system-level simulation, but it may be only reflected in Section 3. Could you provide more information about system-level simulation between the energy harvester and interface circuits, especially in pROM model of Section 5?

5. In page 15, line 343, the table ?? is table 5?

Some minor errors should be corrected.

Author Response

Please see the attachment. Thanks!

Author Response File: Author Response.pdf

Reviewer 2 Report

The article presents the technique of using pMOR model approach to significantly reduce time of computations of the FEM. Although the article is  interesting and in general well-written, I have a few comments/suggestions:

1. The article can be very useful for those using the specific software ANSYS. However, some parts of the article, especially those having specific names not existing in other FEM softwares, can be difficult to understand and confusing. In my opinion, the article should be written more general to be useful for wider audience. Now, unfortunately, it looks like the advertisement of ANSYS or like an ANSYS guidebook. 

2. The title "Towards System-level Simulation of a Miniature Electromagnetic Energy Harvester Model" suggests the development of a new energy harvester. However, the authors do not present any new design. They just prepare faster model. To be honest, the same procedure can be used for completely different model not only for miniature devices. 

3. The last part of the article (chapter 5) shows the comparison between FOM and pROM, presenting the relative error. This is a comparison between two simplified models. In my opinion, the comparison should be given in respect to the "traditional" FEM model given in the chapter 2. 

4. I do not understand the goal of preparing time-dependent simulations of the harvester. Even the equation 1 can be treated as harmonic. The steady-state harmonic simulation is much faster than the time-dependent. So, what is the goal?

Some detailed comments:

Abstract: I suggest removing references from the abstract.

Line 41: Finite Element Method (FEM) is more often used than FE.

Line 108, Fig.4 and the rest of the article: I would suggest avoiding of using too many names of the inventor "Beeby", "from Beeby" etc. In my opinion, just "ref. []" is enough.

Fig. 3 and the rest of the article: There are many different softwares aiming FEM which reader can use. In my opinion your article should be more general to be interesting not only for ANSYS users. Moreover, it can be not fully understandable for those who are not familiar with ANSYS. 

The begining of the chapter 3: 125k elements with about 18 times points computed within 135 minutes looks too long. Firstly, you should provide the mesh plot. How the mesh was optimized? You should use adaptive meshing technique to adjust the mesh elements number in a few iterations (10.1016/j.ijepes.2021.107737). Moreover, in your case we can cut the model and use the symmetry plane. Secondly, why do you use time-domain solver? You can assume the rotation is constant so you can use harmonic-steady state solver to compute it within a few minutes. 

Lines 122-123: "Electrostatic", "Magnetostatic" do not mean anything for non-ANSYS users. You should provide the description of the solvers and what simplifications in physics the software consider (preferably by giving equations). 

Table 2: you should provide comparison of the results between models. For instance, what is the power calculated by two models in a period. 

Eq. 15: Does the star mean conjugate?

Author Response

Please see the attachment. Thanks!

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Thanks the authors for their reply. Just minor comment for the future: when the magnets rotate, the mesh associated with some parts can rotate with them. In another words, some parts of the mesh can be copied due to it doesn't change during rotation.