Modelling Aero-Structural Deformation of Flexible Membrane Kites
Round 1
Reviewer 1 Report
1. What is the novelty of the paper?
2. The paper should show the effect of wind direction to the speed and direction of the kite.
3. Please show the aerodynamic characterization of the kite designed.
4. The paper doesn't provide the power generated from the designed kite.
5. The paper doesn't provide the formula of wind harnessing of the designed kite.
6. The paper doesn't provide a suitable wind penetration power for the designed kite.
6. What is the future work of the paper?
Author Response
Dear reviewer,
Thank you for taking the time to review our manuscript and providing feedback. As there were some comments about things not being thoroughly explained and/or details missing; we have rephrased and added lots of new text and figures. I hope those efforts satisfy your questions. Below a detailed response to each one of your questions is present.
1. What is the novelty of the paper?
Response: Within the leading-edge inflatable (LEI) kite simulation world, a robust, fast, available and accurate aero-structural coupled model does not yet exist. In a conventional aero-structural model, FEM and CFD are used, which requires large computational costs. Furthermore, the dynamic mesh coupling causes several hard-to-resolve problems, making the handful of successful efforts highly specified and prone to instabilities, i.e. running only under a strict number of conditions, simplifications and sometimes even requiring calibrations. This paper presents a novel structural model that resolves the difficulties of dynamic mesh coupling and the computational speed issue. Both a pure geometric constraint-based analytical model and a force-based model are presented. Furthermore, no kite deformation data is available. This led to the performing of a photogrammetry study, another novelty of this paper.
2. The paper should show the effect of wind direction on the speed and direction of the kite.
Response: Considering the presented work is state-of-the-art LEI aero-structural coupled kite simulation research, such effects would be a great addition but fall for this paper outside the scope. Thanks for the interesting suggestion. It will undoubtedly be considered for follow-up work.
3. Please show the aerodynamic characterization of the kite designed.
Response: Our apologies for not being clear enough on the first draft. We have clarified several things to address this point in the updated version. First of all, this work describes the MSc thesis of J.A.W. Poland. During this time, Cayon Domingo worked on his MSc thesis on the same FSI model but focussed on aerodynamic characterisation. Cayon Domingo used the structural model presented here and developed a more complex and more accurate aerodynamic model. We want to refer you to his work. As this wasn't clear enough in the paper, we have added this at the end of the introduction and in the conclusion and discussion section. We have also extended the explanation of the used aerodynamic model.
4. The paper doesn't provide the power generated from the designed kite.
Response: Power generation of an airborne wind energy system is a calculation involving many component models, e.g. winch model, generator model, tether model, flight-path planner, control algorithm etc. This paper focuses on the structural model of the kite itself, and therefore the power generation calculation falls outside of the scope.
5. The paper doesn't provide the formula of wind harnessing of the designed kite.
Response: High-level analytical equations are possible but require several assumptions. This paper is more detailed, focussing on the kite itself. Furthermore, as stated in answer 4, wind harnessing or power generation calculations fall outside this work's scope.
6. The paper doesn't provide suitable wind penetration power for the designed kite.
Response: Similar to answers 4 and 5, this paper does not focus on power generation. Therefore, answering the wind penetration power question falls outside the scope of the presented research.
7. What is the future work of the paper?
Response: We would like to refer you to the conclusion and discussion section, where a discussion of future work activities is presented. To summarize, more accurate experiments shall be performed and the model's modularity and accuracy will be increased.
Best regards,
The authors
Reviewer 2 Report
The authors discussed a particle system approach to simulate deformation of leading-edge inflatable kite. Although the work is interesting, I do not believe there is sufficient novelty and results for considering publication in this journal. The results need more details. If the authors can work on the uncertainity quantification part (which is mentioned as future work), the work could be considered for resubmission.
Author Response
Dear reviewer,
Thank you for taking the time to review our manuscript and providing feedback. As there were some comments about things not being thoroughly explained and/or details missing; we have rephrased and added lots of new text and figures. I hope those efforts satisfy your questions. Below a detailed response to each one of your questions is present.
Question:
The authors discussed a particle system approach to simulate the deformation of a leading-edge inflatable kite. Although the work is interesting, I do not believe there is sufficient novelty and results for considering publication in this journal. The results need more details. If the authors can work on the uncertainty quantification part (which is mentioned as future work), the work could be considered for resubmission.
Response:
First, a response to the novelty concerns.
Within the leading-edge inflatable (LEI) kite simulation world, a robust, fast, available and accurate aero-structural coupled model does not yet exist. In a conventional aero-structural model, FEM and CFD are used, which requires large computational costs. Furthermore, the dynamic mesh coupling causes several hard-to-resolve problems, making the handful of successful efforts highly specified and prone to instabilities, i.e. running only under a strict number of conditions, simplifications and sometimes even requiring calibrations. This paper presents a novel structural model that resolves the difficulties of dynamic mesh coupling and the computational speed issue. Both a pure geometric constraint-based analytical model and a force-based model are presented.
Regarding the uncertainty quantification, we would like to note that the airborne-wind energy field is relatively young, i.e. less than five sold products on the market. Similarly, the research field is also young, making data availability scarce. In particular, information on the shape of leading-edge inflatable kites is hard, up to impossible, to get by. It is mostly non-existent, which is also the case for the current project and V3 kite. Therefore, a novel effort was exerted to design and execute a photogrammetry study using the only available data: video footage. The footage was discretized into separate flight states using visual and audial input. From this, the width was extracted to indicate the shape deformation. As mentioned, this is suboptimal, as more accuracy would be desired. The current research, therefore, can't address the uncertainty quantification better than it did. However, future work should and will.
Best regards,
The authors
Reviewer 3 Report
The manuscript addresses Modelling Aeroelastic Deformation of Flexible Membrane Kites. It covers a suitable research gap but the need for this paper is not clearly stated. The reader is referred to O. Cayon Domingo 63 et al., but what is the benefit over this previous reference? They must be stated at the end of the Introduction.
Methodology is not clearly explained. Glossary of terms must be included to follow the calculations. Solver section is not clearly specified. How is the iteration working? Could a diagram for it be included?
Results are described but not compared to literature. Little critical discussion is carried out. Is this result a great achievement over literature? How is it better? How is the depowered shape better?
Conclusions are not matching the main findings and they would suggest recommendations.
Structure of the paper is good, but English must be checked.
Author Response
Dear reviewer,
Thank you for taking the time to review our manuscript and providing feedback. As there were some comments about things not being thoroughly explained and/or details missing; we have rephrased and added lots of new text and figures. I hope those efforts satisfy your questions. Below a detailed response to each one of your questions is present.
Question:
The manuscript addresses Modelling Aeroelastic Deformation of Flexible Membrane Kites. It covers a suitable research gap but the need for this paper is not clearly stated. The reader is referred to O. Cayon Domingo et al., but what is the benefit over this previous reference? They must be stated at the end of the Introduction.
Response:
Our apologies for not being clear enough on the first draft. To explain it in more detail, this work describes the MSc thesis of J.A.W. Poland. During this time, Cayon Domingo worked on his MSc thesis, which was aimed at achieving the same leading-edge inflatable fluid-structure interaction kite model but focused on aerodynamic characterisation. Cayon Domingo used the structural model presented here and developed a more complex and more accurate aerodynamic model. There is in essence thus no benefit over Cayon Domingo's work, as that builds on top of this work. Similarly, as Cayon Domingo's work, this work is not published yet but only written as an MSc thesis. As this wasn't clear enough in the paper, I have added this at the end of the introduction and in the conclusion and discussion section. We have also extended the explanation of the used aerodynamic model.
Question:
Methodology is not clearly explained. Glossary of terms must be included to follow the calculations. Solver section is not clearly specified. How is the iteration working? Could a diagram for it be included?
Response:
To clarify the methodology in more detail, the introduction, particle system model and photogrammetry sections have been rewritten and extended. A methodology subsection has also been added to clarify the depowering and steering kinematics. As it's not common to add a glossary of terms in such papers for the generic symbol choices that explain algorithms it is not included, however, a list of abbreviations is present. Furthermore, the various variables and symbols are explained in depth in the text and several important ones are now also present, e.g. kite parameters in table 1 and solver constants in table 3. Furthermore, the numerical solver section (2.4.4) has been given lots of attention. It has been extended, a table containing constants has been added, and a flow-chart diagram has been included which some text that explains the iteration step by step.
Question:
Results are described but not compared to the literature. Little critical discussion is carried out. Is this result a great achievement over literature? How is it better? How is the depowered shape better?
Response:
That is a good point, the first main improvement upon the existing literature, which mainly consists of rigid-body assumption simulations, is including deformation a must as the kite acts as a morphing laminar wing, i.e. it is actively deformed to steer. For those few models that do take into account deformations, this model is an improvement as it has a lower computational cost, is more robust and versatile. As this wasn't clear enough, we have altered the conclusions and discussion section to stress this point.
Question:
Conclusions are not matching the main findings and they would suggest recommendations.
Response:
The conclusions discuss the high-level main findings, which is a congruence between the two-plate triangular model, the particle system model and photogrammetry in terms of width changes when going from a powered to a depowered state. Recommendations are present as it is a conclusion and discussion section, therefore (especially from future work onwards) some recommendations are included.
Question:
The structure of the paper is good, but English must be checked.
Response:
A major English grammar check has been done. Apologies for the mistakes present in the previous version.
Best regards,
The authors
Reviewer 4 Report
The article is recommended for publication but must address certain detailed issues as follows.
1) Overall, the paper content weighs too much on information that has appeared in publications, i.e., Sections 2-1 are too long compared to Sections 2-4.
2) The quality of the graphs provided are lacking. Quality can be improved
3) The methodology is not clear enough. The authors should provide a flow chart depicting the modeling methodology
4) The authors need to provide explanation on the parameters that they have selected as the input to the model as it will greatly influence the output of the prediction
5) Results and discussion are not sufficient. The authors should increase the rigor of the comparative analysis among different models adopted. Also, provide interactive discussions linking the model analysis with some of the existing studies related to the current work. Explain how various factors may affect the modeling accuracy.
Author Response
Dear reviewer,
Thank you for taking the time to review our manuscript and providing feedback. As there were some comments about things not being thoroughly explained and/or details missing; we have rephrased and added lots of new text and figures. I hope those efforts satisfy your questions. Below a detailed response to each one of your questions is present.
The article is recommended for publication but must address certain detailed issues as follows.
1) Overall, the paper content weighs too much on information that has appeared in publications, i.e., Sections 2-1 are too long compared to Sections 2-4.
Response:
To clarify the research done in more detail, the methodology section has been redone. The subsections and subsubsections have been extended and rewritten, and even a new subsection has been added. This restores the balance in the description of work done to previous work.
Furthermore, the described work in sections 2-1 is novel and not a description of what has already appeared. Several references are present (similarly for the rest of the paper) as it refers to the MSc thesis of the first author, this is however not published work.
2) The quality of the graphs provided are lacking. Quality can be improved
Response:
Most graphs are uploaded in vector-based form and can thus not be uploaded in higher quality. Which specific graph/plot are you referring to?
3) The methodology is not clear enough. The authors should provide a flow chart depicting the modelling methodology
Response:
To clarify the methodology in more detail, the sections governing the methodology (geometric constraint-based models, particle system model and photogrammetry) have been rewritten and extended. To make the particle system modelling methodology more clear, the numerical solver section (2.4.4) has been given lots of attention. It has been extended, a table containing constants has been added, and a flow-chart diagram has been included which some text that explains the iteration step by step.
4) The authors need to provide explanation on the parameters that they have selected as the input to the model as it will greatly influence the output of the prediction.
Response:
This is a good point, two tables have been added for this purpose. Five tables containing data used for the simulation are now present.
5) Results and discussion are not sufficient. The authors should increase the rigor of the comparative analysis among different models adopted. Also, provide interactive discussions linking the model analysis with some of the existing studies related to the current work. Explain how various factors may affect the modeling accuracy.
Response:
The individual adopted models are compared in terms of width prediction with respect to the experimentally obtained values and versatility. A discussion on the lack of versatility of the geometric constraint-based models is present in the methodology section. There it becomes clear that the geometric constraint-based model using two analytical algorithms requires geometrically determined shapes. It therefore can't handle the full complexity of the bridle line systems nor can it deal with slacking bridles. Because the particle system model is a force-based method, these problems are resolved.
The authors have decided to focus on a comparison between simulation and experimental work, rather than on a comparison between simulation and older simulation. The reason is that even though a handful of these studies were done on the same kite wing, it had a different bridle line system, as this evolved over time. The bridle line system dominantly determines the shape of the wing and therefore comparing these simulations to older ones, is, in the eyes of the authors, less relevant than comparing to experimental values.
Effects that are influencing the accuracy of the experimental values are discussed and a discussion on the parameters affecting the modelling accuracy is present in the conclusion and discussion section (mainly from "Future work shall.." onwards), which has been rewritten to ensure this point comes across more clearly.
Best regards,
The authors
Round 2
Reviewer 1 Report
The authors have addressed all my comments for this paper.
Author Response
Thank you for the feedback.
Reviewer 2 Report
The authors have addressed my concerned. The explanation of novelty of the work is sufficient. The authors also explained limitation in uncertainty quantification study. I believe that the results are convincing and with minor edits the paper could be accepted for publication.
Author Response
Dear reviewer,
Thank you for taking the time to review our manuscript and providing feedback. The requested minor edits are implemented, i.e. the results have had a thorough review and the conclusion section as well.
Best regards,
The Authors
Reviewer 3 Report
I recommend it for publication
