Bionic Stiffener Layout Optimization with a Flexible Plate in Solar-Powered UAV Surface Structure Design
Round 1
Reviewer 1 Report
In general, it was very difficult to understand how the design procedure works and more explanations of the methods are needed. This reviewer has understood the paper as follows in the next paragraph, and subsequently outlines recommendations for improving the manuscript.
A method for optimizing the orientation of stiffeners for thin-walled panels is described in this paper, motivated by the need for lighter wing structures for solar-powered unmanned aerial vehicles. The synthesis procedure involves multiple stages starting with a ground structure method for initialising the design, followed by homogenization procedure and a Lindemayer-system (L-system) for partitioning zones of stiffeners. The optimization procedure is driven by a genetic algorithm which updates the seed weights upon which the ultimate design is based. The method has been applied for two buckling cases, namely uniaxial compression and biaxial compression and out-of-plane buckling factors along with mass were computed and formed the objective functions in separate cases.
This manuscript needs to be thoroughly checked and revised for its English Please combine Sections 2 and 3 into a more coherent single section. Describe Fig. 4 sequentially in complete detail to make it clearer. Questions that remain include: What is equation 1? There is no description of any of the variables Why has this sequence been adopted: ground-structure method followed by homogenization followed by the L-system partitioning? It is not clear why this is necessary, more explanations are needed How is the ground structure method performed in detail? How many design variables are there? (It is only stated that “a small number of input parameters” are required. What is the population size for a generation? It is mentioned that 50 seconds are required for an iteration, is that for a particular individual or a generation? What computational power (memory etc) was needed for this? It was not clear how exactly the design variables (seed weights) correspond to a final topology? Page 8 Line 198, what is Eq. 6? Why was only buckling investigated? For lightweight aircraft structures, issues such as bending, twist, shear, etc are also necessary. How would the composite layup look like in more detail? It is mentioned that a sandwich construction is used, with a honeycomb core and (assuming) face-sheets. If this assumption is correct, how would the stiffeners be attached to the face-sheets? The paragraph on page 9 (lines 230-236) could contain more detail regarding the empirical relations and the overall numerical modelling. For the results, why are the final topologies the way they are? Can any insight into what is physically happening in terms of load-transfer between the stiffeners? Whilst the results are compared with traditional +/- 45 degree grid-stiffeners, can these results be compared with designs specifically for buckling? For example, results from gradient-based topology optimization methods such as SIMP for buckling of thin-shells? What is the definition of “relative buckling load RBL”? Figure 12 a and b: the arrow indicating maximum relative buckling load does not seem to be attached to the points that are furthest to the right along the horizontal axis. What does this mean? From Table 2, the results for the “Optimal RBL” for biaxial compressions shows lower buckling load factors for modes 2 and 3. Why is this the case? Figure 16b, the topology contained in the results does not seem to look like any of the topologies in Figure 14. Are these results then valid? This clearly needs investigation. Recent literature has shown other uses of L-systems, such as the algorithm “SPIDRS” (“Spatial Interpretation for the Development of Reconfigurable Structures”). It would be worthwhile citing some papers on this method.Overall the synthesis procedure appears to be quite interesting and novel however many details are missing which, in the present state of the manuscript, limits the value for publication. Therefore, this reviewer recommends major changes as outlined above.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The point of peer-reviewed article publication is the dissemination of impactful and fruitful scientific and engineering research studies – new methods, innovative ways of accomplishing problems we have in science, medicine, and engineering. If a study is presented so vaguely and limited that even an experienced technical person cannot recreate, or cannot reasonably design a similar study themselves, then there isn’t much point to publish it. This paper is founded on an interesting idea, albeit not novel or new, that could lead to different ways of designing lightweight structures with comparable or better performance. However, the presentation is so riddled with grammatical and spelling errors, and basic shortcomings like not defining variables or providing information about CAE tools and knobs, parameters employed, that’s it neither contributes to the field nor backs the claim that the claims are even meaningful. Even if these issues are aggressively addressed, it is of minor impact because the actual numerical examples are brief and leave a multitude of unanswered claims. Before submitting a document for peer-review, have the article reviewed for proper and clear language. It only wastes the time of expert reviewers and diminishes the impact of the study. There are no definitions of the variables used in the paper, either in a nomenclature table nor within the text – so the reader has no idea what the objective function is minimizing. Fundamentally, the uni-axial compression problem should not suggest any out-of-plane variability. If variability is observed in the design process, then it’s likely due to machine round-off error, or inconsistencies within the framework. There are no mesh refinement studies conducted, which is even more pertinent for a grid/fem optimization study. In general, the grids are very coarse, and could provide some major benefit if it can be shown that the same results are obtained when the ground structure is much finer, for example. More studies are needed to claim the method provides better performing designs, and further, that it’s not just an exercise in academic, unrealistic applications. The words in the title are literally mentioned one time in the introduction, and the actual technical aspects of the paper have little to do with the claimed application. How does the method perform with true solar-UAS loading scenarios? If the language and clarify were significantly boosted, then at best, it would be a “stub” without much support or scientific backing; albeit a brief, somewhat interesting, prelude to a more in-depth, elaborate study, representative of a conference proceeding (not a journal publication).
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Whilst more details regarding the optimization procedures and methods have been provided that do improve the paper, the paper needs to be heavily revised in terms of its English usage. This was requested in the first review but has not been adequately addressed.
This reviewer suggests revising this paper with a native English speaker, or with the use of a professional translation service if possible. It is this reviewer's intention to provide technical comments that are as helpful as possible however this reviewer is unable to do since the paper remains quite difficult to understand. This would also be the case for the readers in the scientific community. This comment is not intended to devalue the scientific quality, rather is a necessary remark to ensure that the work sees a greater impact.
Author Response
The article has undergone English language editing by MDPI.
