Deep convolutional neural networks have gained significant traction as effective approaches for developing detailed but compact representations of complex structured data. Generative networks in particular have become popular for their ability to mimic data distributions and allow further exploration of them. This attribute can be utilized in engineering design domains, in which the data structures of finite element meshes for analyzing potential designs are well suited to the deep convolutional network approaches that are being developed at a rapid pace in the field of image processing. This paper explores the use of conditional generative adversarial networks (cGANs) as a means of generating a compact latent representation of structures resulting from classical topology optimization techniques. The constraints and contextual factors of a design problem, such as mass fraction, material type, and load location, can then be specified as input conditions to generate potential topologies in a directed fashion. The trained network can be used to aid concept generation, such that engineers can explore a variety of designs relevant to the problem at hand with ease. The latent variables of the generator can also be used as design parameters, and the low dimensionality enables tractable computational design without analytical sensitivities. This paper demonstrates these capabilities and discusses avenues for further developments that would enable the engineering design community to further leverage generative machine learning techniques to their full potential.
{"title":"Topology Design With Conditional Generative Adversarial Networks","authors":"Conner Sharpe, C. Seepersad","doi":"10.1115/detc2019-97833","DOIUrl":"https://doi.org/10.1115/detc2019-97833","url":null,"abstract":"\u0000 Deep convolutional neural networks have gained significant traction as effective approaches for developing detailed but compact representations of complex structured data. Generative networks in particular have become popular for their ability to mimic data distributions and allow further exploration of them. This attribute can be utilized in engineering design domains, in which the data structures of finite element meshes for analyzing potential designs are well suited to the deep convolutional network approaches that are being developed at a rapid pace in the field of image processing. This paper explores the use of conditional generative adversarial networks (cGANs) as a means of generating a compact latent representation of structures resulting from classical topology optimization techniques. The constraints and contextual factors of a design problem, such as mass fraction, material type, and load location, can then be specified as input conditions to generate potential topologies in a directed fashion. The trained network can be used to aid concept generation, such that engineers can explore a variety of designs relevant to the problem at hand with ease. The latent variables of the generator can also be used as design parameters, and the low dimensionality enables tractable computational design without analytical sensitivities. This paper demonstrates these capabilities and discusses avenues for further developments that would enable the engineering design community to further leverage generative machine learning techniques to their full potential.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131339042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
James Cunningham, S. Miller, M. Yukish, T. Simpson, Conrad S. Tucker
We present a form-aware reinforcement learning (RL) method to extend control knowledge from one design form to another, without losing the ability to control the original design. A major challenge in developing control knowledge is the creation of generalized control policies across designs of varying form. Our presented RL policy is form-aware because in addition to receiving dynamic state information about the environment, it also receives states that encode information about the form of the design that is being controlled. In this paper, we investigate the impact of this mixed state space on transfer learning. We present a transfer learning method for extending a control policy to a different design form, while continuing to expose the agent to the original design during the training of the new design. To demonstrate this concept, we present a case study of a multi-rotor aircraft simulation, wherein the designated task is to achieve a stable hover. We show that by introducing form states, an RL agent is able to learn a control policy to achieve the hovering task with both a four rotor and three rotor design at once, whereas without the form states it can only hover with the four rotor design. We also benchmark our method against a test case that removes the transfer learning component, as well as a test case that removes the continued exposure to the original design to show the value of each of these components. We find that form states, transfer learning, and parallel learning all contribute to a more robust control policy for the new design, and that parallel learning is especially important for maintaining control knowledge of the original design.
{"title":"Deep Reinforcement Learning for Transfer of Control Policies","authors":"James Cunningham, S. Miller, M. Yukish, T. Simpson, Conrad S. Tucker","doi":"10.1115/detc2019-97689","DOIUrl":"https://doi.org/10.1115/detc2019-97689","url":null,"abstract":"We present a form-aware reinforcement learning (RL) method to extend control knowledge from one design form to another, without losing the ability to control the original design. A major challenge in developing control knowledge is the creation of generalized control policies across designs of varying form. Our presented RL policy is form-aware because in addition to receiving dynamic state information about the environment, it also receives states that encode information about the form of the design that is being controlled. In this paper, we investigate the impact of this mixed state space on transfer learning. We present a transfer learning method for extending a control policy to a different design form, while continuing to expose the agent to the original design during the training of the new design. To demonstrate this concept, we present a case study of a multi-rotor aircraft simulation, wherein the designated task is to achieve a stable hover. We show that by introducing form states, an RL agent is able to learn a control policy to achieve the hovering task with both a four rotor and three rotor design at once, whereas without the form states it can only hover with the four rotor design. We also benchmark our method against a test case that removes the transfer learning component, as well as a test case that removes the continued exposure to the original design to show the value of each of these components. We find that form states, transfer learning, and parallel learning all contribute to a more robust control policy for the new design, and that parallel learning is especially important for maintaining control knowledge of the original design.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116688420","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The use of unit cell structures in mechanical design has seen a steady increase due to their abilities to achieve a wide range of material properties and accommodate multi-functional requirements with a single base material. We propose a novel material property envelope (MPE) that encapsulates the attainable effective material properties of a given family of unit cell structures. The MPE interfaces the coarse and fine scales by constraining the combinations of the competing material properties (e.g., volume fraction, Young’s modulus, and Poisson’s ratio of isotropic materials) during the design of coarse scale material properties. In this paper, a sampling and reconstruction approach is proposed to represent the MPE of a given family of unit cell structures with the method of moving least squares. The proposed approach enables the analytical derivatives of the MPE, which allows the problem to be solved more accurately and efficiently during the design optimization of the coarse scale effective material property field. The effectiveness of the proposed approach is demonstrated through a two-scale structure design with octet trusses that have cubically symmetric effective stiffness tensors.
{"title":"Distributed Design of Two-Scale Structures With Unit Cells","authors":"Xingchen Liu","doi":"10.1115/detc2019-97672","DOIUrl":"https://doi.org/10.1115/detc2019-97672","url":null,"abstract":"\u0000 The use of unit cell structures in mechanical design has seen a steady increase due to their abilities to achieve a wide range of material properties and accommodate multi-functional requirements with a single base material. We propose a novel material property envelope (MPE) that encapsulates the attainable effective material properties of a given family of unit cell structures. The MPE interfaces the coarse and fine scales by constraining the combinations of the competing material properties (e.g., volume fraction, Young’s modulus, and Poisson’s ratio of isotropic materials) during the design of coarse scale material properties. In this paper, a sampling and reconstruction approach is proposed to represent the MPE of a given family of unit cell structures with the method of moving least squares. The proposed approach enables the analytical derivatives of the MPE, which allows the problem to be solved more accurately and efficiently during the design optimization of the coarse scale effective material property field. The effectiveness of the proposed approach is demonstrated through a two-scale structure design with octet trusses that have cubically symmetric effective stiffness tensors.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117114079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Design space exploration can reveal the underlying structure of design problems. In a set-based approach, for example, exploration can map sets of designs or regions of the design space that meet specific performance requirements. For some problems, promising designs may cluster in multiple regions of the input design space, and the boundaries of those clusters may be irregularly shaped and difficult to predict. Visualizing the promising regions can clarify the design space structure, but design spaces are typically high-dimensional, making it difficult to visualize the space in three dimensions. To convey the structure of such high-dimensional design regions, a two-stage approach is proposed to (1) identify and (2) visualize each distinct cluster or region of interest in the input design space. This paper focuses on the visualization stage of the approach. Rather than select a singular technique to map high-dimensional design spaces to low-dimensional, visualizable spaces, a selection procedure is investigated. Metrics are available for comparing different visualizations, but the current metrics either overestimate the quality or favor selection of certain visualizations. Therefore, this work introduces and validates a more objective metric, termed preservation, to compare the quality of alternative visualization strategies. Furthermore, a new visualization technique previously unexplored in the design automation community, t-Distributed Neighbor Embedding, is introduced and compared to other visualization strategies. Finally, the new metric and visualization technique are integrated into a two-stage visualization strategy to identify and visualize clusters of high-performance designs for a high-dimensional negative stiffness metamaterials design problem.
{"title":"Visualizing and Evaluating High-Dimensional Mappings of Sets of High Performance Designs","authors":"C. Morris, M. Haberman, C. Seepersad","doi":"10.1115/detc2019-97722","DOIUrl":"https://doi.org/10.1115/detc2019-97722","url":null,"abstract":"\u0000 Design space exploration can reveal the underlying structure of design problems. In a set-based approach, for example, exploration can map sets of designs or regions of the design space that meet specific performance requirements. For some problems, promising designs may cluster in multiple regions of the input design space, and the boundaries of those clusters may be irregularly shaped and difficult to predict. Visualizing the promising regions can clarify the design space structure, but design spaces are typically high-dimensional, making it difficult to visualize the space in three dimensions. To convey the structure of such high-dimensional design regions, a two-stage approach is proposed to (1) identify and (2) visualize each distinct cluster or region of interest in the input design space. This paper focuses on the visualization stage of the approach. Rather than select a singular technique to map high-dimensional design spaces to low-dimensional, visualizable spaces, a selection procedure is investigated. Metrics are available for comparing different visualizations, but the current metrics either overestimate the quality or favor selection of certain visualizations. Therefore, this work introduces and validates a more objective metric, termed preservation, to compare the quality of alternative visualization strategies. Furthermore, a new visualization technique previously unexplored in the design automation community, t-Distributed Neighbor Embedding, is introduced and compared to other visualization strategies. Finally, the new metric and visualization technique are integrated into a two-stage visualization strategy to identify and visualize clusters of high-performance designs for a high-dimensional negative stiffness metamaterials design problem.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128732115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Traditionally, the ideation of design opportunities and new concepts relies on human expertise or intuition and is faced with high uncertainty. Inexperienced or specialized designers often fail to explore ideas broadly and become fixed on specific ideas early in the design process. Recent data-driven design methods provide external design stimuli beyond one’s own knowledge, but their uses in rapid ideation are still limited. Intuitive and directed ideation techniques, such as brainstorming, mind mapping, Design-by-Analogy, SCAMPER, TRIZ and Design Heuristics may empower designers in rapid ideation but are limited in the designer’s own knowledge base. Herein, we harness data-driven design and rapid ideation techniques to introduce a data-driven computer-aided rapid ideation process using the cloud-based InnoGPS system. InnoGPS integrates an empirical network map of all technology domains based on the international patent classification which are connected according to knowledge distance based on patent data, with a few map-based functions to position technologies, explore neighborhoods, and retrieve knowledge, concepts and solutions in the near or far fields for design analogies and syntheses. The functions of InnoGPS fuse design science, network science, data science and interactive visualization and make the design ideation process data-driven, theoretically-grounded, visually-inspiring, and rapid. We demonstrate the procedures of using InnoGPS as a data-driven rapid ideation tool to generate new rolling toy design concepts.
{"title":"Computer-Aided Design Ideation Using InnoGPS","authors":"Jianxi Luo, Serhad Sarica, K. Wood","doi":"10.1115/detc2019-97587","DOIUrl":"https://doi.org/10.1115/detc2019-97587","url":null,"abstract":"\u0000 Traditionally, the ideation of design opportunities and new concepts relies on human expertise or intuition and is faced with high uncertainty. Inexperienced or specialized designers often fail to explore ideas broadly and become fixed on specific ideas early in the design process. Recent data-driven design methods provide external design stimuli beyond one’s own knowledge, but their uses in rapid ideation are still limited. Intuitive and directed ideation techniques, such as brainstorming, mind mapping, Design-by-Analogy, SCAMPER, TRIZ and Design Heuristics may empower designers in rapid ideation but are limited in the designer’s own knowledge base. Herein, we harness data-driven design and rapid ideation techniques to introduce a data-driven computer-aided rapid ideation process using the cloud-based InnoGPS system. InnoGPS integrates an empirical network map of all technology domains based on the international patent classification which are connected according to knowledge distance based on patent data, with a few map-based functions to position technologies, explore neighborhoods, and retrieve knowledge, concepts and solutions in the near or far fields for design analogies and syntheses. The functions of InnoGPS fuse design science, network science, data science and interactive visualization and make the design ideation process data-driven, theoretically-grounded, visually-inspiring, and rapid. We demonstrate the procedures of using InnoGPS as a data-driven rapid ideation tool to generate new rolling toy design concepts.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133458643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
There is great potential for using 3D printed designs fabricated via additive manufacturing processes for diverse biomedical applications. 3D printing offers capabilities for customizing designs for each new fabrication that could leverage automated design processes for personalized patient care, but there are challenges in developing accurate and efficient assessment methods. Here, we conduct a sensitivity analysis for a biological growth simulation for evaluating 3D printed lattices for regenerating bone and then use these simulations to identify performance trends. Four design topologies were compared by generating varied unit cells. Biological growth was modeled in a voxel environment by simulating the advancement of a tissue front by calculating its local curvature. Designs were generated with properties suitable for bone tissue engineering, namely 50% porosity and microscale pores. The sensitivity analysis determined trade-offs between prediction consistency and computation time, suggesting calculating curvature within a radius of 7.5 voxels is sufficient for most cases. Topologies were compared in bulk with design variations. All topologies had similar tissue growth rates for a given surface-volume ratio, but with differing unit cell sizes. These findings inform future optimization for selecting unit cells based on volume requirements and other criteria, such as mechanical stiffness. A fitted analytical relationship predicted tissue growth rate based on a design’s surface-volume ratio, which enables design evaluation without computationally expensive simulations. Lattices were 3D printed with biocompatible materials as proof-of-concepts, demonstrating the feasibility of the approach for future computational design methods for personalized medicine.
{"title":"Design and Biological Simulation of 3D Printed Lattices for Biomedical Applications","authors":"P. Egan","doi":"10.1115/detc2019-98190","DOIUrl":"https://doi.org/10.1115/detc2019-98190","url":null,"abstract":"\u0000 There is great potential for using 3D printed designs fabricated via additive manufacturing processes for diverse biomedical applications. 3D printing offers capabilities for customizing designs for each new fabrication that could leverage automated design processes for personalized patient care, but there are challenges in developing accurate and efficient assessment methods. Here, we conduct a sensitivity analysis for a biological growth simulation for evaluating 3D printed lattices for regenerating bone and then use these simulations to identify performance trends. Four design topologies were compared by generating varied unit cells. Biological growth was modeled in a voxel environment by simulating the advancement of a tissue front by calculating its local curvature. Designs were generated with properties suitable for bone tissue engineering, namely 50% porosity and microscale pores. The sensitivity analysis determined trade-offs between prediction consistency and computation time, suggesting calculating curvature within a radius of 7.5 voxels is sufficient for most cases. Topologies were compared in bulk with design variations. All topologies had similar tissue growth rates for a given surface-volume ratio, but with differing unit cell sizes. These findings inform future optimization for selecting unit cells based on volume requirements and other criteria, such as mechanical stiffness. A fitted analytical relationship predicted tissue growth rate based on a design’s surface-volume ratio, which enables design evaluation without computationally expensive simulations. Lattices were 3D printed with biocompatible materials as proof-of-concepts, demonstrating the feasibility of the approach for future computational design methods for personalized medicine.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122555359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hybrid electric powertrain systems present as effective alternatives to traditional vehicle and marine propulsion means with improved fuel efficiency, as well as reduced greenhouse gas (GHG) emissions and air pollutants. In this study, a new integrated, model-based design and optimization method for hybrid electric propulsion system of a marine vessel (harbor tugboat) has been introduced. The sizes of key hybrid powertrain components, especially the Li-ion battery energy storage system (ESS), which can greatly affect the ship’s life-cycle cost (LCC), have been optimized using the fuel efficiency, emission and lifecycle cost model of the hybrid powertrain system. Moreover, the control strategies for the hybrid system, which is essential for achieving the minimum fuel consumption and extending battery life, are optimized. For a given powertrain architecture, the optimal design of a hybrid marine propulsion system involves two critical aspects: the optimal sizing of key powertrain components, and the optimal power control and energy management. In this work, a bi-level, nested optimization framework was proposed to address these two intricate problems jointly. The upper level optimization aims at component size optimization, while the lower level optimization carries out optimal operation control through dynamic programming (DP) to achieve the globally minimum fuel consumption and battery degradation for a given vessel load profile. The optimized Latin hypercube sampling (OLHS), Kriging and the widely used Expected Improvement (EI) online sampling criterion are used to carry out “small data” driven global optimization to solve this nested optimization problem. The obtained results showed significant reduction of the vessel LCC with the optimized hybrid electric powertrain system design and controls. Reduced engine size and operation time, as well as improved operation efficiency of the hybrid system also greatly decreased the GHG emissions compared to traditional mechanical propulsion.
{"title":"Integrated System Design and Control Optimization of Hybrid Electric Propulsion System Using a Bi-Level, Nested Approach","authors":"L. Chen, Huachao Dong, Z. Dong","doi":"10.1115/detc2019-97456","DOIUrl":"https://doi.org/10.1115/detc2019-97456","url":null,"abstract":"\u0000 Hybrid electric powertrain systems present as effective alternatives to traditional vehicle and marine propulsion means with improved fuel efficiency, as well as reduced greenhouse gas (GHG) emissions and air pollutants. In this study, a new integrated, model-based design and optimization method for hybrid electric propulsion system of a marine vessel (harbor tugboat) has been introduced. The sizes of key hybrid powertrain components, especially the Li-ion battery energy storage system (ESS), which can greatly affect the ship’s life-cycle cost (LCC), have been optimized using the fuel efficiency, emission and lifecycle cost model of the hybrid powertrain system. Moreover, the control strategies for the hybrid system, which is essential for achieving the minimum fuel consumption and extending battery life, are optimized. For a given powertrain architecture, the optimal design of a hybrid marine propulsion system involves two critical aspects: the optimal sizing of key powertrain components, and the optimal power control and energy management. In this work, a bi-level, nested optimization framework was proposed to address these two intricate problems jointly. The upper level optimization aims at component size optimization, while the lower level optimization carries out optimal operation control through dynamic programming (DP) to achieve the globally minimum fuel consumption and battery degradation for a given vessel load profile. The optimized Latin hypercube sampling (OLHS), Kriging and the widely used Expected Improvement (EI) online sampling criterion are used to carry out “small data” driven global optimization to solve this nested optimization problem. The obtained results showed significant reduction of the vessel LCC with the optimized hybrid electric powertrain system design and controls. Reduced engine size and operation time, as well as improved operation efficiency of the hybrid system also greatly decreased the GHG emissions compared to traditional mechanical propulsion.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121291602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We propose a novel framework based on multi-fidelity design optimization for indirectly solving computationally hard topology optimization problems. The primary concept of the proposed framework is to divide an original topology optimization problem into two subproblems, i.e., low- and high-fidelity design optimization problems. Hence, artificial design parameters, referred to as seeding parameters, are incorporated into the low-fidelity design optimization problem that is formulated on the basis of a pseudo-topology optimization problem. Meanwhile, the role of high-fidelity design optimization is to obtain a promising initial guess from a dataset comprising topology-optimized design candidates, and subsequently solve a surrogate optimization problem under a restricted design solution space. We apply the proposed framework to a topology optimization problem for the design of flow fields in battery systems, and confirm the efficacy through numerical investigations.
{"title":"A Framework of Multi-Fidelity Topology Design and its Application to Optimum Design of Flow Fields in Battery Systems","authors":"K. Yaji, S. Yamasaki, S. Tsushima, K. Fujita","doi":"10.1115/detc2019-97675","DOIUrl":"https://doi.org/10.1115/detc2019-97675","url":null,"abstract":"\u0000 We propose a novel framework based on multi-fidelity design optimization for indirectly solving computationally hard topology optimization problems. The primary concept of the proposed framework is to divide an original topology optimization problem into two subproblems, i.e., low- and high-fidelity design optimization problems. Hence, artificial design parameters, referred to as seeding parameters, are incorporated into the low-fidelity design optimization problem that is formulated on the basis of a pseudo-topology optimization problem. Meanwhile, the role of high-fidelity design optimization is to obtain a promising initial guess from a dataset comprising topology-optimized design candidates, and subsequently solve a surrogate optimization problem under a restricted design solution space. We apply the proposed framework to a topology optimization problem for the design of flow fields in battery systems, and confirm the efficacy through numerical investigations.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128452336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A personalized, 3D printed, multi-material artificial spinal disc is expected to not only achieve personalized anatomical fit, but also to restore the natural mechanics of the implanted spinal segment. However, the necessary structure for disc design is not explored and optimizing the design is challenging due to the high-dimensional search space provided by the material distribution precision of multi-material 3D printing as well as necessary nonlinear finite element simulation. Therefore, this study explores the feasibility of two multi-material spinal disc designs and a clustering-based design variable linking method to achieve efficient and effective optimization. The optimization goal is to enable the implant to have natural stiffnesses for five loading cases. The results show that a biomimetic fiber network is necessary for the disc design. Moreover, the optimization performance of the heuristic derived from a clustering-based method is shown to be a good trade-off between the objective function value and the computational time.
{"title":"Computational Design of a Personalized Artificial Spinal Disc With a Data-Driven Design Variable Linking Heuristic","authors":"Zhiyang Yu, K. Shea, T. Stanković","doi":"10.1115/detc2019-97777","DOIUrl":"https://doi.org/10.1115/detc2019-97777","url":null,"abstract":"\u0000 A personalized, 3D printed, multi-material artificial spinal disc is expected to not only achieve personalized anatomical fit, but also to restore the natural mechanics of the implanted spinal segment. However, the necessary structure for disc design is not explored and optimizing the design is challenging due to the high-dimensional search space provided by the material distribution precision of multi-material 3D printing as well as necessary nonlinear finite element simulation. Therefore, this study explores the feasibility of two multi-material spinal disc designs and a clustering-based design variable linking method to achieve efficient and effective optimization. The optimization goal is to enable the implant to have natural stiffnesses for five loading cases. The results show that a biomimetic fiber network is necessary for the disc design. Moreover, the optimization performance of the heuristic derived from a clustering-based method is shown to be a good trade-off between the objective function value and the computational time.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"201 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115488960","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Over 50% of the energy from power plants, vehicles, oil refining, and steel or glass making process is released to the atmosphere as waste heat. As an attempt to deal with the growing energy crisis, the solid-state thermoelectric generator (TEG), which converts the waste heat into electricity using Seebeck phenomenon, has gained increasing popularity. Since the figures of merit of the thermoelectric materials are temperature dependent, it is not feasible to achieve high efficiency of the thermoelectric conversion using only one single thermoelectric material in a wide temperature range. To address this challenge, this paper proposes a method based on topology optimization to optimize the layouts of functional graded TEGs consisting of multiple materials. The objective of the optimization problem is to maximize the output power and conversion efficiency as well. The proposed method is implemented using the Solid Isotropic Material with Penalization (SIMP) method. The proposed method can make the most of the potential of different thermoelectric materials by distributing each material into its optimal working temperature interval. Instead of dummy materials, both the P and N-type electric conductors are optimally distributed with two different practical thermoelectric materials, namely Bi2Te3 & PbTe for P-type, and Bi2Te3 & CoSb3 for N-type respectively, with the yielding conversion efficiency around 12.5% in the temperature range Tc = 25°C and Th = 400°C. In the 2.5D computational simulation, however, the conversion efficiency shows a significant drop. This could be attributed to the mismatch of the external load and internal TEG resistance as well as the grey region from the topology optimization results as discussed in this paper.
{"title":"Multimaterial Topology Optimization of Thermoelectric Generators","authors":"Xiaoqiang Xu, Yongjia Wu, L. Zuo, Shikui Chen","doi":"10.1115/detc2019-97934","DOIUrl":"https://doi.org/10.1115/detc2019-97934","url":null,"abstract":"\u0000 Over 50% of the energy from power plants, vehicles, oil refining, and steel or glass making process is released to the atmosphere as waste heat. As an attempt to deal with the growing energy crisis, the solid-state thermoelectric generator (TEG), which converts the waste heat into electricity using Seebeck phenomenon, has gained increasing popularity. Since the figures of merit of the thermoelectric materials are temperature dependent, it is not feasible to achieve high efficiency of the thermoelectric conversion using only one single thermoelectric material in a wide temperature range. To address this challenge, this paper proposes a method based on topology optimization to optimize the layouts of functional graded TEGs consisting of multiple materials. The objective of the optimization problem is to maximize the output power and conversion efficiency as well. The proposed method is implemented using the Solid Isotropic Material with Penalization (SIMP) method. The proposed method can make the most of the potential of different thermoelectric materials by distributing each material into its optimal working temperature interval. Instead of dummy materials, both the P and N-type electric conductors are optimally distributed with two different practical thermoelectric materials, namely Bi2Te3 & PbTe for P-type, and Bi2Te3 & CoSb3 for N-type respectively, with the yielding conversion efficiency around 12.5% in the temperature range Tc = 25°C and Th = 400°C. In the 2.5D computational simulation, however, the conversion efficiency shows a significant drop. This could be attributed to the mismatch of the external load and internal TEG resistance as well as the grey region from the topology optimization results as discussed in this paper.","PeriodicalId":365601,"journal":{"name":"Volume 2A: 45th Design Automation Conference","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122584789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：