Aerodynamic Design, Analysis and Validation of a Small Blended-Wing-Body Unmanned Aerial Vehicle
Round 1
Reviewer 1 Report
Aerodynamic Design, Analysis, and Validation of a Small Blended-Wing-Body Unmanned Aerial Vehicle
By
Kelei Wang and Zhou Zhou
The authors may explain the following point:
- Why static Margin is negative. When the static Margin is -5 then the vehicle will be statically unstable. Can they show that vehicle is statically as well as dynamically stable?
- What are the Ref. area and you arrive at?
- What is the location of the center of pressure for the Nose of the body?
- The maximum angle of attack is 12 degrees. Did you try for a higher angle of attack?
- Once these questions are addressed then the paper can be accepted
Author Response
Reviewer:
Comment 1: Why static Margin is negative. When the static Margin is -5 then the vehicle will be statically unstable. Can they show that vehicle is statically as well as dynamically stable?
Response: Thank you very much for your comments. I am sorry that we cannot understand the purpose of this comment quite well. In our opinion, the ΔCM/ΔCL<-5% means that the entire BWB configuration is in a statically stable state, that is, when the lift increases due to the increase in the angle of attack, a greater nose-down pitching moment (CM<0 represents the nose-down pitching moment) will be generated to inhibit the further increases in both angle of attack and lift.
Comment 2: What are the Ref. area and you arrive at?
Response: In line 85-87, we have cited that “Therefore, it is decided that the scope of the aerodynamic design study is to only design the BWB configuration under DEP considerations (mainly under the geometric constraints caused by the DEP installation), and several characteristic sections are selected to be optimized while the BWB planform should remain fixed as much as possible”. Since the “the BWB planform remains fixed” has been considered as a design constraint in this paper, the reference area is always kept as Sref = 3.40 m2 throughout the whole design process.
Comment 3: What is the location of the center of pressure for the Nose of the body?
Response: I am sorry that we cannot understand the purpose of this comment quite well. In this paper, the center of gravity of the BWB configuration is located at x=1.10 m with respect to the nose of the body, which has been used as the moment reference point (MRP) (see Fig. 2(a)) in the design process. So when the center of pressure is before the center of gravity, the BWB configuration will nose up, and when the center of pressure is behind the center of gravity, the BWB configuration will nose down.
Comment 4: The maximum angle of attack is 12 degrees. Did you try for a higher angle of attack?
Response: The design point of the current work is mainly at a small angle of attack at cruise, therefore, we pay more attention to the cruise state with a small angle of attack rather than a high angle of attack. Besides, in our opinion, both the computational error and the computational cost will increase significantly at a high angle of attack, which is not suitable for engineering design. Hence, we did not try to design at a high angle of attack.
Comment 5: Once these questions are addressed then the paper can be accepted.
Response: We have added explanations in lines 69-72 according to your comments, so as to facilitate readers’ understanding of the design issues of this paper.
Thanks again for your help and advices.
Best wishes for you.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Summary: This paper describes the aerodynamic design, analysis and validation of a BWB UAV under DEP installation. The authors describe the background of the study to a sufficient amount of detail and set up the stage for the remainder of the manuscript in the introduction. They describe and explain how the design parameters are setup and the algorithm used for the optimization process. The computational setup has been described with sufficient amount of detail for anyone to recreate the model and verify the results. Details related to the computational grids used and the grid convergence are provided by the authors. The proposed optimized design is simulated using RANS based computational solver. The results are validated using wind tunnel experiments.
Review Comments: The work presented in this manuscript is novel and of great value to the readers of this journal. However, there are a few things that need to addressed to improve the quality of this manuscript. Firstly, and most importantly, it is not clear how the validation of the results of one design, i.e., the optimized design generated using the optimization method validates the optimization method. The wind tunnel experiments validate the CFD results obtained for that particular design and that too for low angle of attacks (AoAs). Therefore, the experimental validation is for the computational model that the authors constructed for the optimized design and not the optimization algorithm. This needs to be clearly stated. If the authors, feel that the wind tunnel experiment validates their optimization method then they should explain that in more detail. In the current form, these statements feel a bit misleading.
Besides the above major comment, I have a few more minor comments/suggestions which the authors should incorporate in the revised manuscript:
- Line 81-85: Provide citations if published previously or provide more information in the form of data set etc. Which can help the reader understand the decision-making process.
- Line 125: It is not clear why the authors chose this method and how it benefits the parametric analysis.
- Line 212: What do the authors mean by 'avoid unavailability of design results'? What makes this stage of the process different from the first one needs to be explained for readers to understand why manual intervention is necessary at this stage.
- Line 239: Citation or Details regarding the adaptive sampling procedures and updating mechanism should be provided for the benefits of the reader.
- Line 243 – 246: Rationale behind choosing these control parameters should be included in the text.
- Confusing choice of nomenclature: The authors chose A, B, C and D to name the different sectional airfoils (Fig 2) and then use LE, A, B, C, TE for sectional airfoil parameters (Fig 4). This can cause confusion to the readers. Therefore, I would like to suggest that the authors change either one of the nomenclatures to make it unique and avoid any form of confusion.
- Line 292: Table number and caption are incorrect. Also, the table seems to be split between two pages, if possible, it should be merged so that its properly formatted and provides more clarity to the reader.
- Line 313: The wind tunnel experiments validate the optimized design not the optimization method used to generate the design. It's not clear why and how the authors are able to make this claim.
- Section 4: Experimental Validation – Based on the comparison of the results presented by the authors, it appears that the computational setup in its current form is not adequate to model the performance of the designed BWB configuration at high angles of attack. This is evident when comparing the results and surface flow visualization 12 degrees onwards. Therefore, it might be better to focus on lower AoA and validate that study and refine the computational setup and then validate the results at high AoA, which can be separate study.
- Conclusion: Line 360 – k-omega SST turbulence model is the industry standard for RANS based computational studies, so this seems to be a strange conclusion from this study. The current study focuses on the author's work on creating a new optimized DEP based BWB UAV design and not on the validation of the k-w SST model at high angle of attacks. With the limited number of cases and grids presented in this study, it is incorrect to say that k-omega SST is efficient for predicting the aerodynamic behavior at low angles but not at high angles where separation occurs. Have the authors tested this claim on a highly refined mesh? Or in an unsteady RANS setup? This line needs to be rephrased to focus on author's work and not on validation of an established viscous model.
Author Response
Reviewer:
Comment 1: The work presented in this manuscript is novel and of great value to the readers of this journal. However, there are a few things that need to addressed to improve the quality of this manuscript. Firstly, and most importantly, it is not clear how the validation of the results of one design, i.e., the optimized design generated using the optimization method validates the optimization method. The wind tunnel experiments validate the CFD results obtained for that particular design and that too for low angle of attacks (AoAs). Therefore, the experimental validation is for the computational model that the authors constructed for the optimized design and not the optimization algorithm. This needs to be clearly stated. If the authors, feel that the wind tunnel experiment validates their optimization method then they should explain that in more detail. In the current form, these statements feel a bit misleading.
Response: Thank you very much for your constructive comments, we quite agree with your comments. The misleading descriptions in several parts of this paper have been revised or removed as required.
Comment 2: Line 81-85: Provide citations if published previously or provide more information in the form of data set etc. Which can help the reader understand the decision-making process.
Response: We quite agree with your comments. But so far our task group has not published any articles about the analysis of the DEP induced effects on the BWB configuration. Besides, the preliminary studies were carried out based on the baseline BWB configuration shown in this paper, and the results were quite similar to the content in section 3.3. Hence, in our opinion, there is no need for repetitive statements here.
Comment 3: Line 125: It is not clear why the authors chose this method and how it benefits the parametric analysis.
Response: We have revised the paper according to your advice. In line 144-145, we have added some explanations about the reason why we use the Hicks-Henne bump functions to model the fictitious airfoil profile. The advantage of this method is that it enables separate modeling of the upper and lower surfaces of the baseline airfoil, which is meaningful for the fictitious airfoil parameterization described in this paper.
Comment 4: Line 212: What do the authors mean by 'avoid unavailability of design results'? What makes this stage of the process different from the first one needs to be explained for readers to understand why manual intervention is necessary at this stage.
Response: We have removed this description according to your advice to avoid readers’ misunderstanding. In fact, what we originally wanted to express was that the results obtained by the optimized design are mathematical or machine-generated, and the manual intervention is required to guarantee the validity of the results.
Comment 5: Line 239: Citation or Details regarding the adaptive sampling procedures and updating mechanism should be provided for the benefits of the reader.
Response: We have added the reference [21] according to your advice.
[21] Wang, K. L., Zhou, Z., Zhu, X. P., and et al. Reconstruction design of the propeller induced flow-field based on the aero-dynamic loading distributions. Acta Aeronautica et Astronautica Sinica, 2020, Vol. 41, No. 1, pp. 123118.
Comment 6: Line 243 – 246: Rationale behind choosing these control parameters should be included in the text.
Response: I’m sorry that we only selected these control parameters to conduct the optimization design process, and did not study the rationale behind choosing these control parameters as you mentioned. We will follow your advice in the follow-up paper work, perhaps there will be better design results.
Comment 7: Confusing choice of nomenclature: The authors chose A, B, C and D to name the different sectional airfoils (Fig 2) and then use LE, A, B, C, TE for sectional airfoil parameters (Fig 4). This can cause confusion to the readers. Therefore, I would like to suggest that the authors change either one of the nomenclatures to make it unique and avoid any form of confusion.
Response: We have revised the paper according to your advice. In section 2.2, we have changed the “LE, A, B, C, TE” to “LE, PSA, PSB, CP, TE”.
Comment 8: Line 292: Table number and caption are incorrect. Also, the table seems to be split between two pages, if possible, it should be merged so that its properly formatted and provides more clarity to the reader.
Response: We have revised the paper according to your advice by changing the tile “Table 4…” to “ Table 5…”.
Comment 9: Line 313: The wind tunnel experiments validate the optimized design not the optimization method used to generate the design. It's not clear why and how the authors are able to make this claim.
Response: We quite agree with your comments. We have removed this description according to your advice to avoid readers’ misunderstanding.
Comment 10: Section 4: Experimental Validation – Based on the comparison of the results presented by the authors, it appears that the computational setup in its current form is not adequate to model the performance of the designed BWB configuration at high angles of attack. This is evident when comparing the results and surface flow visualization 12 degrees onwards. Therefore, it might be better to focus on lower AoA and validate that study and refine the computational setup and then validate the results at high AoA, which can be separate study.
Response: We have revised the description in line 473-475 according to your advice. The aim of “study the flow mechanism of the designed BWB configuration at 324 high angles of attack” is changed to “explore the differences between the present numerical methods and wind tunnel tests in predicting the aerodynamic performance of the BWB configuration at high angles of attack”. This is because that the original purpose of our research in section 4.2.2 is to find why there is such a significant difference between the numerical results and experimental data, so as to further improve the numerical methods and finally increase the calculation accuracy at high angles of attack.
We also stated the above problems in Conclusion that “…Besides, it appears that the present numerical methods are not adequate to model the performance of the designed BWB configuration at high angles of attack. In the subsequent research, it is of importance to further improve the numerical methods and study the airframe/propulsion integrated aerodynamic performance…”
Comment 11: Conclusion: Line 360 – k-omega SST turbulence model is the industry standard for RANS based computational studies, so this seems to be a strange conclusion from this study. The current study focuses on the author's work on creating a new optimized DEP based BWB UAV design and not on the validation of the k-w SST model at high angle of attacks. With the limited number of cases and grids presented in this study, it is incorrect to say that k-omega SST is efficient for predicting the aerodynamic behavior at low angles but not at high angles where separation occurs. Have the authors tested this claim on a highly refined mesh? Or in an unsteady RANS setup? This line needs to be rephrased to focus on author's work and not on validation of an established viscous model.
Response: We have revised the description in line 512-533 according to your advice. The conclusions about the k-omega SST turbulence model is removed, and the design results of the well-performed DEP-based BWB configuration are emphasized.
Thanks again for your help and constructive advices.
Best wishes for you.
Author Response File: Author Response.docx
Reviewer 3 Report
This manuscript presents a concept design (18 independent sectional design variables are considered) of a BWB UAV subject to geometric constraints for distributed electric propulsion (DEP) installation. The experimental results show that RANS works well in cruise performance evaluations, which means we can use RANS-based optimization to maximize the cruise performance in detailed design optimization of this configuration. However, it is also shown that RANS cannot accurately estimate stall, so if the optimization involves aerodynamic at stall (such as the maximum CL), higher-fidelity models are in demand.
Overall, this work would draw the interest of journal readers. Nevertheless, the content is not well organized for readers to quickly understand the workflow; Results are not analyzed in-depth to provide useful guidelines for other researchers. A major revision is needed before it can be accepted for publication.
Major comments:
1. The optimization design framework is not well described (Page 8-9). After reading these pages several times, I'm assuming that your design is performed in this way: optimize the two sectional shapes independently based on the objectives/constraints allocated using the 3-D simulation results. If it is the case, you do not need to emphasize "in a total of 18 design variables for the BWB configuration need to be used to carry out the optimization" in Line 229, because the optimization is not coupled. Instead, you need to present the section optimization problems in Eq.2, with details of the design variables, objective functions, and constraints. Also, it should be explained in detail how the objective/constraints are allocated to the section optimization problems rather than merely a simple sentence of "based on the numerical analysis of the baseline BWB configuration" (Line 217). Line 220 and 223, "If the design targets is satisfied, ...", what target? Please explain clearly. Also, targets "are" satisfied, not is. I suggest modifying Fig.9 to be consistent with the four steps explained between Line 216 and Line 224.
2. Line 199: The motivation for using the same parameter setting in Sec 1-3 is not well explained. I just see this is a simplification, but I cannot get the point of "Therefore, we use a simplified method.." Why "Therefore"? Explain the reason.
3. It is mentioned that "the adoption of the DEP system adds more geometric components and shape constraints to the BWB configuration, and may result in an airframe body strikingly different from typical BWB aircrafts" in the introduction. This seems to be the motivation for your work. I'm curious whether this constraint would affect the aerodynamic performance, and I think other readers also are interested in this. There is no conclusion/investigation on this. Could you add a new design case (without using the geometric constraints for DEP installation) to investigate this point?
4. The analysis of the experimental data does not provide new insights to the designers. I think at least you can add more conclusions on:
1) if we could use RANS-based optimization with full-turbulent assumption to optimize the cruise performance of this configuration;
2) if we need more accurate models to evaluate performance near stall.
Minor comments:
1. Table 1: the Maximum lift-to-drag ratio is not a condition but a design objective. Please modify the table caption correspondingly.
2. Line 82: lack conjunction between clauses.
3. Line 118-121: "Therefore" is not a conjunction. It is an adverb, so it does not connect clauses.
4. Line 212: What is "to avoid unavailability of design results"?
5. There are many other grammar errors. Please do a thorough check before the next submission.
Author Response
Reviewer:
Comment 1: The optimization design framework is not well described (Page 8-9). After reading these pages several times, I'm assuming that your design is performed in this way: optimize the two sectional shapes independently based on the objectives/constraints allocated using the 3-D simulation results. If it is the case, you do not need to emphasize "in a total of 18 design variables for the BWB configuration need to be used to carry out the optimization" in Line 229, because the optimization is not coupled. Instead, you need to present the section optimization problems in Eq.2, with details of the design variables, objective functions, and constraints. Also, it should be explained in detail how the objective/constraints are allocated to the section optimization problems rather than merely a simple sentence of "based on the numerical analysis of the baseline BWB configuration" (Line 217). Line 220 and 223, "If the design targets is satisfied, ...", what target? Please explain clearly. Also, targets "are" satisfied, not is. I suggest modifying Fig.9 to be consistent with the four steps explained between Line 216 and Line 224.
Response: Thank you very much for your constructive comments, we have revised the paper according to your advice. See descriptions in section 2.4 (line 236-373) for details.
Comment 2: Line 199: The motivation for using the same parameter setting in Sec 1-3 is not well explained. I just see this is a simplification, but I cannot get the point of "Therefore, we use a simplified method.." Why "Therefore"? Explain the reason.
Response: We have removed this description according to your advice to avoid readers’ misunderstanding. At the same time, in order to express the advantages of this fictitious airfoil simplification method, a description is added in section 2.2.3 (line 170-174).
Comment 3: It is mentioned that "the adoption of the DEP system adds more geometric components and shape constraints to the BWB configuration, and may result in an airframe body strikingly different from typical BWB aircrafts" in the introduction. This seems to be the motivation for your work. I'm curious whether this constraint would affect the aerodynamic performance, and I think other readers also are interested in this. There is no conclusion/investigation on this. Could you add a new design case (without using the geometric constraints for DEP installation) to investigate this point?
Response: We quite agree with your comments. But in our opinion, the parameterization problem and the reflexed characteristics of section A, B, and C caused by the DEP installation geometric constraints are very obvious, which is significantly different from the conventional clean-shaped BWB configuration. Besides, we have added some explanations about the difference of the present DEP-based BWB configuration design concept from the conventional BWB configuration design concept in line 354, this is the most real feeling we got in this design work.
Your comment "add a new design case without using the geometric constraints for DEP installation" will be considered in our future research, but we think that it is not inappropriate in the revision of this paper.
Comment 4: The analysis of the experimental data does not provide new insights to the designers. I think at least you can add more conclusions on:
1) if we could use RANS-based optimization with full-turbulent assumption to optimize the cruise performance of this configuration;
2) if we need more accurate models to evaluate performance near stall.
Response: We have revised the conclusions according to your advice. We believe that the optimization design process based on the RANS method in this paper is feasible and effective, and the computational accuracy is quite high at small angles of attack, so we mainly describe the problems in predicting the aerodynamic performance of the BWB configuration at high angles of attack in Conclusion.
“…it appears that the present numerical methods are not adequate to model the performance of the designed BWB configuration at high angles of attack. In the subsequent research, it is of importance to further improve the numerical methods and study the airframe/propulsion integrated aerodynamic performance…”
Comment 5: Table 1: the Maximum lift-to-drag ratio is not a condition but a design objective. Please modify the table caption correspondingly.
Response: We have revised the paper according to your advice.
Comment 6: Line 82: lack conjunction between clauses.
Response: We have revised the paper according to your advice.
Comment 7: Line 118-121: "Therefore" is not a conjunction. It is an adverb, so it does not connect clauses.
Response: We have revised the paper according to your advice.
Comment 8: Line 212: What is "to avoid unavailability of design results"?
Response: We have removed this description according to your advice to avoid readers’ misunderstanding. In fact, what we originally wanted to express was that the results obtained by the optimized design are mathematical or machine-generated, and the manual intervention is required to guarantee the validity of the results.
Comment 9: There are many other grammar errors. Please do a thorough check before the next submission.
Response: We have revised the paper according to your advice.
Thanks again for your help and constructive advices.
Best wishes for you.
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
The authors have addressed my comments, and I suggest accepting this version for publication. Besides, a small suggestion for the authors: please do careful proof-editing (for example see Line 208-210).
