Constraint Replacement-Based Design for Additive Manufacturing of Satellite Components: Ensuring Design Manufacturability through Tailored Test Artefacts
Round 1
Reviewer 1 Report
In this paper, the authors explore a new concept of design-for-additive manufacturing for satellite components. The paper is generally well-written and clear (but could use expansion in some areas). I am familiar with previous work from this group and they have a good reputation in the design and additive manufacturing communities (I have no affiliation, nor do I know the authors personally). In my opinion, their past work is serious and useful for the expansion of the field.
With this paper under review, I really like the topic and approach, but I have a number of concerns and suggestions for improvement. However, none of these alone are major or fatal to the work so it should be seriously considered once these concerns are addressed by the authors. With this in mind, the paper is very promising, but I think it needs some more work and expansion before it would be for archival journal publication. Therefore, I suggest that the editor return the paper to the authors for a revision before we can further evaluate it. Please note that all comments and suggestions are meant only to help the authors and ensure high-quality published literature in the field and are not intended to reflect personally on the authors in any way.
Specific comments and suggestions for the author
1. The quality of the English writing is generally excellent
2. The general literature review should be expanded some. While I understand that many of them are not directly related to the aerospace topic, there are a number of different approaches to DFAM that have been presented in the past couple of years which are not acknowledged here. I will not give a specific list of the missing references, as it could create a conflict of interest since I am an author on some of them. However, the authors of this paper should make an effort to expand the coverage some in order to best present/contrast the new method within the existing ones
3. The literature review/introduction needs a dedicated section on the design problems, constraints, etc related to designing and manufacturing satellite components in order to justify the development of a new design method. Especially discuss the manufacturability constraints (related to AM, traditional, and hybrid processes) related to the manufacturing of these items (several recent papers have been published on the use, generation, and mapping of these constraints). Also focus on the material issues and how the AM material limitations might be affected. This section will be very important for your paper, as it will be the way you evaluate the success of your method at the end of the paper.
4. The first paragraph (Lines 37-43) needs significant expansion, as this is a very important background for the rest of the paper. The references used here could be improved as well – and more are needed. Especially expand the discussion and references on the discussion of design freedom. Design freedom with AM is high but limited and this can be a very difficult design problem with strange constraints
5. Lines 58-65: Taguchi models for AM experiments are becoming more common and trusted, providing rigorous experiments with drastically reduced samples sized. In addition, there is not much in the AM literature on this (there is in some conceptual design and optimization literature) but you should discuss the concept of “minimally-restrictive” constraints in this section. These two topics provide a counter and expansion to what you state here and a discussion of these could benefit the discussion.
6. Line 70: I offer two more bullets for this list:
- The manufactures can do a part re-design, where AM is used to deal with the design and manufacturing problems, but no further. The design is not optimized for the AM processes and could be feasibly made using a casting or machining process except for one or two small details (e.g., cooling channels or a single external feature).
- AM can be used to manufacture a part that is not particularly complex or difficult to build, but for some reason traditional processes are not able to be used. This could be because of material effects or because a traditional process is just not available. Or the AM process is simply cheaper or faster – there are cases where some medical and aerospace parts made from titanium and Inconel alloys are cheaper and faster to process using SLM or LENS than a machining process.
In particular, I have seen a case in industry (not published but there are a number of papers with these kinds of studies – LLNL and Oak Ridge plus the groups at NASA, Missouri University of Science & Technology, and NC State University have done work like this – I am sure there are others) where it required 80 hours to machine a particular Titanium part (requiring a tool change every 1-2 hours) but SLM was able to make 3 in the same time at about 20% of the cost each. Of course, the SLM structural properties were inferior but it was not a structural component, so it worked fine. Some of the previous AM design work I have seen when working with a NASA team at a previous position was focused on these two things
7. Section 1 needs a discussion of verification, validation, and certification of the AM space parts
8. Lines 95-97: This is true, but you should also consider the source of the lack of knowledge about the processes and how knowledge might fail to be communicated (especially to designers). Should we focus on rigorous process models or should we develop better systems for capturing expert intuition/experience? Or both? Or something else? A well-argued opinion on this from the authors would be a very good thing for this paper
9. The end of Section 1 needs a clear and specific statement of what is novel and valuable about the presented paper
10. Figure 10 is kind of blurry and should be replaced with one of higher resolution
11. I would like to see more details about the manufacturing of the artefacts made during your case study. Give the manufacturing parameters, assumptions, etc. Also please provide details about how you determined that no defects were present – simple visual inspection, ultrasonic inspection, etc.?
12. The constraints on the process discussed toward the end of the paper are certainly not a complete set of manufacturing constraints for these parts. More discussion on this and an acknowledgment by the authors that other active constraints may exist should be given. Also consider that the constraints may or may not be active – this will need to be more clearly presented in the method. Other design constraints (from the material, from cost limits, from industry standards, etc.) may actually dominate some of the manufacturing constraints, so the generation of these constraints only makes your problem larger and more difficult to solve.
13. Finally, how do you know that the original constraints that are being replaced are valid and active? It needs to be more clearly presented or stated explicitly as an assumption with a discussion of the sensitivity of this assumption.
------------------------------------------------------------------------------------
Good luck and I look forward to seeing the new version of this paper
Author Response
The authors would like to thank the reviewer for the valuable comments, which have helped to significantly improve the manuscript.
Specific comments and suggestions for the author
1. The quality of the English writing is generally excellent
Our response:
We appreciate this comment, however, to improve the text and increase the writing quality, the article has gone through a professional proofreading and editing service..
2. The general literature review should be expanded some. While I understand that many of them are not directly related to the aerospace topic, there are a number of different approaches to DFAM that have been presented in the past couple of years which are not acknowledged here. I will not give a specific list of the missing references, as it could create a conflict of interest since I am an author on some of them. However, the authors of this paper should make an effort to expand the coverage some in order to best present/contrast the new method within the existing ones
Our response:
Most of the literature review section has been re-written and is now divided in four sections: 1) The impact of AM on the design of metallic components for satellites, 2) Design for AM in space applications, 3) Expectations toward a DfAM method for space applications, and 4) Design of test artifacts for AM of space components. On the second part of the review (lines 114-137), we now mention various DfAM methods and we relate them with different design situations that might arise in a design project for space components. We now mention the work conducted by (lines 138-158):
Thompson, M., Moroni, G., Vaneker, T., Fadel, G., Campbell, R., Gibson, I., Martina, F. Design for Additive Manufacturing: Trends, opportunities, considerations, and constraints. In CIRP Annals - Manufacturing Technology, 65(2), 737–760, 2016. https://doi.org/10.1016/j.cirp.2016.05.004. Quincieu, J., Robinson, C., Stucker, B., Mosher, T. Case study: Selective laser sintering of the USUSat II small satellite structure. Assembly Automation, 25(4), 267–272, 2015. Boyard, N., Rivette, M., Christmann, O., Richir S. A design methodology for parts using additive manufacturing. In International Conference on Advanced Research in Virtual and Rapid Prototyping VRAP (pp. 0-6.), Leiria, Portugal. HAL-01197463, 2013. doi: 10.1201/b15961-74. Campbell, I., Bourell, D., Gibson, I. Additive manufacturing : rapid prototyping comes of age. Rapid Prototyping Journal, 18(4), 255‐258, 2012. doi: https://doi.org/10.1108/13552541211231563. Orme, M. E., Gschweitl, M., Vernon, R., … Mouriaux, F. A demonstration of additive manufacturing as an enabling technology for rapid satellite design and fabrication. In International SAMPE Technical Conference, 2016. Thornton, J., Dalay, B., Smith, D. Additive manufacturing of waveguide for Ku-band satellite communications antenna. In 2016 10th European Conference on Antennas and Propagation, EuCAP 2016, 1–4, 2016. Borgue, O., Panarotto, M., Isaksson, O. Impact on design when introducing additive manufacturing in space applications. In 15th International Design Conference (pp. 997-1008), Dubrovnik, Croatia, 2018. doi: 10.21278/idc.2018.0412.
We continue the analysis of the topic of DfAM in the third part of the literature review, but the text rapidly continues into the topic of manufacturing constraints for satellite components where we include the work presented by:
Castet, J., Saleh, J. Satellite Reliability: Statistical Data Analysis and Modeling. Journal of Spacecraft and rockets, 46(5), 1165-1076, 2019. doi: https://doi.org/10.2514/1.42243. Boudjemai, A., Bekhti, M., Bouanane, M. H., Si Mohammed, A. M., Cooper, G., Richardson, G. Small Satellite Computer Aided Design and Manfacturing. In European Conference on Spacecraft Structures, Materials & Mechanical Testing, Noordwijk, The Netherlands, 2005. Booth, P., Skeen, M., Stirland, S. Low cost, short lead-time feed chain components for multi-beam antennas. In European Conference on Antennas and Propagation, EuCAP 2009, Proceedings, 853–857, 2009. Guo, N., Leu, M. C. Additive manufacturing: Technology, applications and research needs. Frontiers of Mechanical Engineering, 8(3), 215–243, 2013. Frazier, W. E. Metal additive manufacturing: A review. Journal of Materials Engineering and Performance, 23(6), 1917–1928, 2014. Sames, W. J., List, F. A., Pannala, S., Dehoff, R. R., Babu, S. S. The metallurgy and processing science of metal additive manufacturing. International Materials Reviews, 6608(March), 1–46, 2016.
After point 1.3 (The third part of the literature review), a large emphasis is still given to the topic of the design of test artefacts for AM. The method proposed in this article is mostly concern with the design of test artefacts for AM and the way they can influence product design and final quality. For this reason, a larger portion of the literature review section is focus in this topic rather on discussing DfAM methods that do not propose the use of test artefacts.
3. The literature review/introduction needs a dedicated section on the design problems, constraints, etc related to designing and manufacturing satellite components in order to justify the development of a new design method. Especially discuss the manufacturability constraints (related to AM, traditional, and hybrid processes) related to the manufacturing of these items (several recent papers have been published on the use, generation, and mapping of these constraints). Also focus on the material issues and how the AM material limitations might be affected. This section will be very important for your paper, as it will be the way you evaluate the success of your method at the end of the paper.
Our response:
Most of the literature review section has been re-written. This new section puts more emphasis on existing design problems and the associated constraints and challenges related to designing and manufacturing satellite components (section 1.1.). Based on the newly introduced references in this section we motivate the need for the development of a new design method. For example: “Researchers and industry practitioners agree that the main challenge when implementing metal AM to design new components is the lack of experience and the large number of uncertainties and unknowns associated with the constraints of the manufacturing process [1]. First, there are non-established standards for machines and processes. Second, there is a lack of knowledge about the physical phenomena that take place during the AM process and difficulty predicting the quality of a piece, as parts manufactured with AM have a complex thermal history that involves repeated fusion, directional heat extraction, and rapid solidification [1, 4]”.
In section 1.2, we highlight the specific challenges from literature in relation to additive manufacturing constraints. (lines 138-158). Lastly, in section 1.3, we derive criteria for the method presented in our article to be a viable support for the development of satellite parts produced in AM. For example: Based on the above- described experiences with AM and satellite design, a DfAM method should be able to represent the product’s design space. This is in the form of constraints, be they either from the specific use-case of space applications [7], ]; material properties, as seen by Booth et al. [21]; and or machine- specific impacts,, as seen in the dimensioning impact with Quincieu et al. [15], ]; but also providing the design freedom to allow for high-performance designs, such as done by Orme et al. [22]. Furthermore, since, in most cases, especially those mentioned by Boudjemai [14], the CAD model plays a large role in the analysis and verification of the product’s performance, a close coupling to an easy editable CAD model is desirable.
We have added the following references:
Booth, P., Skeen, M., & Stirland, S. (2009). Low cost, short lead-time feed chain components for multi-beam antennas. European Conference on Antennas and Propagation, EuCAP 2009, Proceedings, 853–857. Boschetto, A., Bottini, L., Eugeni, M., … Gaudenzi, P. (2019). Selective Laser Melting of a 1U CubeSat structure. Design for Additive Manufacturing and assembly. Acta Astronautica, 159, 377–384. Boudjemai, A., Bekhti, M., Bouanane, M. H., Si Mohammed, A. M., Cooper, G., & Richardson, G. (2005). Small Satellite Computer Aided Design and Manfacturing. In European Conference on Spacecraft Structures, Materials & Mechanical Testing, Noordwijk, The Netherlands. Castet, J. F., & Saleh, J. H. (2009). Satellite and satellite subsystems reliability: Statistical data analysis and modeling. Reliability Engineering and System Safety, 94(11), 1718–1728. Johannesson, H., Landahl, J., Levandowski, C. E., & Raudberget, D. (2017). Development of product platforms: Theory and methodology. Concurrent Engineering Research and Applications, 25(3), 195–211. McInnes, A. I., Harps, D., Lang, J., & Swenson, C. M. (2001). A systems engineering tool for satellite simulator design. In Proceedings of the Small Satellite Conference. doi:10.1115/ESDA2010-25341 Orme, M. E., Gschweitl, M., Vernon, R., … Mouriaux, F. (2016). A demonstration of additive manufacturing as an enabling technology for rapid satellite design and fabrication. International SAMPE Technical Conference, 2016-January(May). Quincieu, J., Robinson, C., Stucker, B., & Mosher, T. (2005). Case study: Selective laser sintering of the USUSat II small satellite structure. Assembly Automation, 25(4), 267–272. Thornton, J., Dalay, B., & Smith, D. (2016). Additive manufacturing of waveguide for Ku-band satellite communications antenna. 2016 10th European Conference on Antennas and Propagation, EuCAP 2016, 1–4.
4. The first paragraph (Lines 37-43) needs significant expansion, as this is a very important background for the rest of the paper. The references used here could be improved as well – and more are needed. Especially expand the discussion and references on the discussion of design freedom. Design freedom with AM is high but limited and this can be a very difficult design problem with strange constraints
Our response:
The paragraph was not extended to keep the flow of the introduction short and concise, but a new section (1.1. The impact of additive manufacturing on the design of metallic components for satellites) has been created extending on the topics of design freedom and constraints in the context of additive manufacturing (lines 84-112).
5. Lines 58-65: Taguchi models for AM experiments are becoming more common and trusted, providing rigorous experiments with drastically reduced samples sized. In addition, there is not much in the AM literature on this (there is in some conceptual design and optimization literature) but you should discuss the concept of “minimally-restrictive” constraints in this section. These two topics provide a counter and expansion to what you state here and a discussion of these could benefit the discussion.
Our response:
We agree with the reviewer´s comment about the increasing popularity of Taguchi methods for DOE. For this reason, we have included in the discussion (lines 639 to 649) a short mention about these methods and a couple more (central composite design or half-factorial design) that are being implemented for DOE in the case of AM technologies. In the future, we would like to strengthen our approach implementing some of the aforementioned DOE methods. For this point we have included the work presented by:
Antony, J. Design of experiments for engineers and scientists. 2014. Elsevier. Madrid, J., Lorin, S., Söderberg, R., Hammersberg, P., Wärmefjord, K., Lööf, J. A Virtual Design of Experiments Method to Evaluate the Effect of Design and Welding Parameters on Weld Quality in Aerospace Applications. Aerospace, 6(6), 74, 2019. Equbal, A., Equbal, M. A., Equbal, M. I., Sood, A. K. Multi-Criterion Decision Method for Roughness Optimization of Fused Deposition Modelled Parts. In Additive Manufacturing Technologies from an Optimization Perspective (pp. 235-262), 2019. IGI Global. Pérez, M., Medina-Sánchez, G., García-Collado, A., Gupta, M., Carou, D. Surface quality enhancement of fused deposition modeling (FDM) printed samples based on the selection of critical printing parameters. Materials, 11(8), 1382, 2018. Baturynska, I., Semeniuta, O., Martinsen, K. Optimization of process parameters for powder bed fusion additive manufacturing by combination of machine learning and finite element method: A conceptual framework. Procedia CIRP, 67, 227-232, 2018
Regarding minimally restrictive constraints, a reference to an article by Patterson and Allison [42] was included. In their article, the authors propose the concept of minimally restrictive along with a design framework for designing manufacturable mechanical components. We find the concept of minimally restrictive constraints to be extremely interesting and appropriate for the content of our article.
We have now included in the text a progression and refinement on the amount and nature of the constraints considered in each phase of the method. In phase 2, for instance, when the constraints replacement strategy is performed and there is no conceptualized AM design, general AM guidelines (or constraints) are considered. Moreover, we included (from Patterson and Allison´s article) the constraints definitions of: Active, inactive and unnecessary (IU); inactive and redundant (IR); inactive and internally dominated (IID) and; inactive and externally dominated (IIE) (lines 325-338). In later phases, as the AM design is conceptualized, the number of AM constraints that are relevant for the design is reduced. At his point the concept of minimally restrictive constraints becomes relevant in our method. This importance is highlighted between lines 603 and 607 in the discussion.
- Patterson, A., Allison, J. Generation and Mapping of Minimally-Restrictive Manufacturability Constraints for Mechanical Design Problems. In Conference: 24th Design for Manufacturing and the Life Cycle Conference (DFMLC) - ASME IDETC/CIE, 2019. 10.13140/RG.2.2.11108.53124.
6. Line 70: I offer two more bullets for this list:
- The manufactures can do a part re-design, where AM is used to deal with the design and manufacturing problems, but no further. The design is not optimized for the AM processes and could be feasibly made using a casting or machining process except for one or two small details (e.g., cooling channels or a single external feature).
- AM can be used to manufacture a part that is not particularly complex or difficult to build, but for some reason traditional processes are not able to be used. This could be because of material effects or because a traditional process is just not available. Or the AM process is simply cheaper or faster – there are cases where some medical and aerospace parts made from titanium and Inconel alloys are cheaper and faster to process using SLM or LENS than a machining process.
In particular, I have seen a case in industry (not published but there are a number of papers with these kinds of studies – LLNL and Oak Ridge plus the groups at NASA, Missouri University of Science & Technology, and NC State University have done work like this – I am sure there are others) where it required 80 hours to machine a particular Titanium part (requiring a tool change every 1-2 hours) but SLM was able to make 3 in the same time at about 20% of the cost each. Of course, the SLM structural properties were inferior but it was not a structural component, so it worked fine. Some of the previous AM design work I have seen when working with a NASA team at a previous position was focused on these two things
Our response:
These points have not been included at line 70 to keep the flow of the introduction short and concise, but they have been added in section 1.1. (The impact of additive manufacturing on the design of metallic components for satellites) to give more examples about the impact of a limited consideration of manufacturing constraints of the re-design of a new component (lines 84-112).
7. Section 1 needs a discussion of verification, validation, and certification of the AM space parts
Our response:
This is true, and a reason also the interest to develop and study how to work with test artifacts. A discussion is introduced (lines 58-71) in introduction as a means to justify the relevance of studying a physical test object, rather than full product- or material test- samples. The idea is to build knowledge in test artifacts, that can be used to validate design rules related to constraints. This is the practice deployed amongst several of our industrial partners, yet there is no systematic way of capturing the learnings and insights into a generic enough modelling tool- such as the one we propose.
The text introduced in the manuscript is intentionally kept short, not to divert the message and logic of the message. Our hope is that it now is better explained and balanced.
8. Lines 95-97: This is true, but you should also consider the source of the lack of knowledge about the processes and how knowledge might fail to be communicated (especially to designers). Should we focus on rigorous process models or should we develop better systems for capturing expert intuition/experience? Or both? Or something else? A well-argued opinion on this from the authors would be a very good thing for this paper
Our response:
We have included in the discussion a reflection about this point (lines 583-596). We state that the function model where functions, design solutions and constraints are modelled, can serve as a boundary object among different company segments. This boundary object can support communication and coordination efforts to further reduce costs and development times. Moreover, it can support decision making activities about future technology related investments and purchases.
9. The end of Section 1 needs a clear and specific statement of what is novel and valuable about the presented paper
Our response:
We have now included at the end of Section 1, a statement that summarized the novelty and value of the proposed approach (lines 189-195)
10. Figure 10 is kind of blurry and should be replaced with one of higher resolution
Our response:
We have improved the quality of every figure in the article
11. I would like to see more details about the manufacturing of the artefacts made during your case study. Give the manufacturing parameters, assumptions, etc. Also please provide details about how you determined that no defects were present – simple visual inspection, ultrasonic inspection, etc.?
Our response:
The text between lines 479-486 has been updated to provide more details about the manufacturing process. As the case study had only illustrative purposes, after the manufacturing process we just performed a visual inspection of the components, as is now stated in lines 504-505.
12. The constraints on the process discussed toward the end of the paper are certainly not a complete set of manufacturing constraints for these parts. More discussion on this and an acknowledgment by the authors that other active constraints may exist should be given. Also consider that the constraints may or may not be active – this will need to be more clearly presented in the method. Other design constraints (from the material, from cost limits, from industry standards, etc.) may actually dominate some of the manufacturing constraints, so the generation of these constraints only makes your problem larger and more difficult to solve.
Our response:
As the purpose of the case study is to illustrate the method, the set of identified active manufacturing constraints is limited. This has now been stated in the discussion. Moreover, as the article makes now more emphasis on the distinction between active and inactive constraints (refer to point 5 of this document) , the fact that other constraints might dominate over the Cms is discussed as well (lines 606-617).
13. Finally, how do you know that the original constraints that are being replaced are valid and active? It needs to be more clearly presented or stated explicitly as an assumption with a discussion of the sensitivity of this assumption.
Our response:
No verification or validation has been conducted to be able to affirm that the identified constraints and their distinction (active/inactive) is accurate. Currently, this process is based on assumptions. However, the iterative nature of the proposed method, facilitates the continuous identification and distinction of constraints, providing the possibility to reevaluate previous assumptions. Furthermore, the documentation ability of the function model, can be an advantage when identifying and distinguishing constraints in future design projects. These statements can be found in the discussion (lines 613-623)
Reviewer 2 Report
The paper entitled "Constraint replacement-based design for Additive Manufacturing of satellite components: ensuring design manufacturability through tailored test artefacts" presents an overview of methodologies for AM designs. Although, the paper is clear, could be improved in the contend by adding a specific section on mechanical (fatigue) response of AM components respect to traditional ones, since, at the moment, this seems the challenge.
Author Response
The paper entitled "Constraint replacement-based design for Additive Manufacturing of satellite components: ensuring design manufacturability through tailored test artefacts" presents an overview of methodologies for AM designs. Although, the paper is clear, could be improved in the contend by adding a specific section on mechanical (fatigue) response of AM components respect to traditional ones, since, at the moment, this seems the challenge.
Our response:
The literature review section has been re-written and extended to include the complex interaction among system requirements for space applications and manufacturing limitations of AM technologies. Moreover, we agree with the reviewer and we are aware that fatigue response of AM components is a relevant topic for the design of space components. We have now included in the discussion, a reflection about the interaction of AM constraints with specific design aspects that can pose a challenge when designing for AM (such as fatigue response, material specifications, cost limitations, etc..). These statements can be found in lines 607-617.
We want to add that to improve the text and increase the writing quality, the article has gone through a professional proofreading and editing service.
Reviewer 3 Report
The paper presents a DFAM tool. The aim of the tool is to facilitate the designer choices in order to replace/re-think in a DFAM mindset component previously realized by traditional manufacturing technologies.
The paper covers an important topics from both the scientific and industrial point of view and it is worth of publication. My only remarks is that authors should better underline and explain how and if this method can be generalized to more complex or critical applications like primary components.
Author Response
The paper presents a DFAM tool. The aim of the tool is to facilitate the designer choices in order to replace/re-think in a DFAM mindset component previously realized by traditional manufacturing technologies.
The paper covers an important topic from both the scientific and industrial point of view and it is worth of publication. My only remark is that authors should better underline and explain how and if this method can be generalized to more complex or critical applications like primary components.
Our response:
The literature review section has been re-written and extended to include the complex interaction among system requirements for space applications and manufacturing limitations of AM technologies (lines 114-137).
Also, lines 58-71 include the topic of validation and certification when manufacturing space components.
Moreover, we have now included in the discussion, a reflection about the interaction of AM constraints with specific design aspects that can pose a challenge when designing complex or critical components for AM (such as fatigue response, material specifications, cost limitations, etc..). These statements can be found in lines 607-617.
We want to add that to improve the text and increase the writing quality, the article has gone through a professional proofreading and editing service.
Round 2
Reviewer 1 Report
The authors have done an excellent revision of the manuscript. It is more thorough and complete, as well as being compelling and an interesting contribution to the literature. This is a area of high interest (AM of aerospace parts) in the field and I think this paper could be a high-impact work - if accepted, I will likely cite it myself in the future. Congrats to the authors on the completion of an excellent study and I look forward to seeing more work from this group in the future. The article is highly recommended for acceptance in current form.
