Analysis and Design of Novel Axial Field Flux-Modulation Permanent Magnet Machines for Direct Drive Application
Round 1
Reviewer 1 Report
Page 2 line 81: ...N and N-pole (NN), shown in Figure 1(b)
Usually, in NN AFPM machines, to optimize the winding flux linkage, the winding coil axes are tangentially oriented rather than parallel to the rotation axes (see figure inserted at page 2 in the attached file). Authors should explain why they choose the same orientation of the NS configuration, since results confirm, as might be expected, a worst performance for this configuration.
Page 2 line 88 and following: The AF-FMPM machines proposed exhibit the following advantages: ...
and what about the drawbacks? (Huge axial forces, complicated bearings, difficult cooling of the inner end-windings, cumbersome lamination manufacturing, etc.)
Page 4 line 130: Table 1. Constant parameters of the proposed machine.
No detail about material properties for stator cores and PM back irons. In this kind of machines, where m.m.f. harmonics (and even sub-harmonics) play a relevant role, magnetic saturation may represent a heavy concern, as well as iron losses. Transient non-linear 3D-FEAs must be adopted to take into account these phenomena, resulting in very high calculation time. Again, no information provided about these issues. Authors should give details about magnetic properties and FEA characteristics (i.e., what kind of lamination is adopted or, alternatively, if iron cores are made of soft sintered magnetic materials and specify their properties; if FEAs are they magnetostatic or transient; what angular step is adopted to define the torque profile and corresponding average value. If magnetostatic FEAs are adopted, some consideration on the iron losses and their determination is anyway necessary.
Page 4 lines 136-138: It is worth noting that during the process of researching optimal value of each individual parameter for maximizing the
machine average torque, other parameters of the proposed machine should be fixed.
According to this statement, the authors present a sequence of distinct independent one-dimensional parametric analyses starting from a predefined arbitrary initial design. This approach gives no guarantee of success in finding an effective optimal point, as demonstrated by plenty of literature on design optimization, much less does it allow for general design considerations, as might be expected in a paragraph entitled "Influence of Critical Parameters on Machine Performance". At the very least, a multidimensional line search strategy should have been adopted.
A further general remark is about the influence of parametric variation on the machine performance: authors focus on torque average value and torque ripple, but geometry variations affect efficiency and heating as well, and the mere torque increase may not in itself represent a performance improvement, should it result in an unacceptable overheating. Also these design aspects should be considered.
Page 6, Figures 3-6: A generally erratic/non monotonic dependence of torque on geometric parameters is rather suspect and suggests it can be related to the influence of the mesh refinement or to an inadequate angular step for the the determination of the average torque. Authors should investigate on this issue.
Page 6, line 178: 3.4. Influence of Slot Depth
Since the deeper the slot, the thinner the stator yoke, the flux weakening (and consequently the torque decrease) is obvious in the NN type. Rather surprisingly, however, the torque increases as slot depth increases from 14.5 to 14.75. Authors should give some explanation about that (see also previous remark).
Page 7, line 186: 4. Design Optimization
General remark: the procedure presented in this paragraph can be actually defined a design optimization strategy. The authors should explain why don't they adopt a similar approach in Section 3. Influence of Critical Parameters on Machine Performance.
Page 7, lines 192-194: In addition, the sinusoidal characteristic of open-circuit phase back EMF is also important to the machine performance.
This kind of machines is generally supplied by dedicated power converters and a moderate back e.m.f. harmonic content generally represents a very minor concern. Authors should explain why they focused on this issue.
Page 8, lines 209-211: Optimization goal is to simultaneously minimize the 5th-order harmonic of phase back EMF and the cogging torque.
Apparently, the most relevant performance aspects, i.e. average torque, fundamental harmonic e.m.f., losses and efficiency are not taken into account in this analysis. At the very least, authors should compare their values at the initial point and at the optimized one.
Further remark: the optimization considers the absolute values of phase back e.m.f. 5th harmonic and peak (I suppose) cogging torque. In my opinion, their relative values should instead have been considered, with respect to the fundamental back e.m.f. and the rated average torque, respectively. In fact, it is not worth decreasing the former ones if the latter ones decrease by the same way (or even more).
Page 11, Figure 15: Authors should explain why the torque for NN type is not null at 90° while it is null at -90°. This asymmetry is not justified by the electromagnetic configuration.
Page 12, lines 302-305: 6. Conclusions
In this paper, an in-depth investigation of the machine configurations with different PM stator position and its flux-modulation principle of the novel AF-FMPM machine are carried out.
Since in the introduction the authors enumerate several drawback of the configurations presented in literature, in the conclusions they should compare the results related to the solution they proposed with the literature ones, to show the benefits they achieved, for instance as for torque density, relative cogging torque and so on.
Comments for author File: 
Comments.pdf
Author Response
The responce to the reviewer is attached
Author Response File: 
Author Response.docx
Reviewer 2 Report
The abstract is good, However, it does not discuss the existing work and its shortcomings,
The introduction meets the requirements and can add recent works.
The methodology session discussed the Design of Novel Axial Field Flux-Modulation Permanent Magnet Machines,
The GA optimization algorithm should be included in the revised version. Not enough information about GA.
Results and discussion should be improved from a technical point of view of content.
The title shows in the application drive, which does not reflect the results and discussion.
Author Response
The responce to the reviewer is attached.
Author Response File: Author Response.docx
Reviewer 3 Report
Greetings to colleagues!
I believe that describing the optimization of electric motors without any mention of the way they are controlled (there is not a single motor switching circuit in the article) is not entirely correct and justified
Best wishes to all.
Author Response
The responce to the reviewer is attached.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Although not entirely agreeing with your optimization approach, on the whole your paper is well conceived and some beneficial insights make it worthy of publication.
Reviewer 3 Report
Agree with revisions
