Optimization and Supervised Machine Learning Methods for Inverse Design of Cellular Mechanical Metamaterials
| dc.contributor.author | Liu, Sheng | en |
| dc.contributor.committeechair | Acar, Pinar | en |
| dc.contributor.committeemember | Kapania, Rakesh K. | en |
| dc.contributor.committeemember | West, Robert L. | en |
| dc.contributor.committeemember | Raeymaekers, Bart | en |
| dc.contributor.department | Mechanical Engineering | en |
| dc.date.accessioned | 2024-05-23T08:01:08Z | en |
| dc.date.available | 2024-05-23T08:01:08Z | en |
| dc.date.issued | 2024-05-22 | en |
| dc.description.abstract | Cellular mechanical metamaterials (CMMs) are a special class of materials that consist of microstructural architectures of macroscopic hierarchical frameworks that can have extraordinary properties. These properties largely depend on the topology and arrangement of the unit cells constituting the microstructure. The material hierarchy facilitates the synthesis and design of CMMs on the micro-scale to achieve enhanced properties (i.e., improved strength, toughness, low density) on the component (macro)-scale. However, designing on-demand cellular metamaterials usually requires solving a challenging inverse problem to explore the complex structure-property relations. The first part of this study (Ch. 3) proposes an experience-free and systematic design methodology for microstructures of CMMs using an advanced stochastic searching algorithm called micro-genetic algorithm (μGA). Locally, this algorithm minimizes the computational expense of the genetic algorithm (GA) with a small population size and a conditionally reduced parameter space. Globally, the algorithm employs a new search strategy to avoid local convergence induced by the small population size and the complexity of the parameter space. What's more, inspired by natural evolution in the GA, this study applies the inverse design method with the standard GA (sGA) as a sampling algorithm for intuitively mapping material-property spaces of CMMs, which requires the selection of objective properties and stochastic search of property points within the property space. The mapping methodology utilizing the sGA is proposed in the second part of the study (Ch. 4). This methodology involves a robust strategy that is shown to identify more comprehensive property spaces than traditional mapping approaches. The resulting property space allows designers to acknowledge the limitations of material performance, and select an appropriate class of CMMs based on the difficulty of the realization and fabrication of their microstructures. During the fabrication process, manufacturing defects cause uncertainty in the microstructures, and thus the structural properties. The third part of the study (Ch. 5) investigates the effects of the uncertainty stemming from manufacturing defects on the material property space. To accelerate the uncertainty quantification (UQ) via the Monte Carlo method, this study utilizes a machine learning technique to bypass the expensive simulations to compute properties. In addition to reducing the computational expense of the simulations, the deep learning method has been proven to be practical to accomplish non-intuitive design tasks. Due to the numerous combinations of properties and complex underlying geometries of metamaterials, it is numerically intractable to obtain optimal material designs that satisfy multiple user-defined performance criteria at the same time. Nevertheless, a deep learning method called conditional generative adversarial networks (CGANs) is capable of solving this many-to-many inverse problem. The fourth part of the study (Ch. 6) proposes a new inverse design framework using CGANs to overcome this challenge. Given combinations of target properties, the framework can generate a group of geometric patterns providing these target properties. Therefore, the proposed strategy provides alternative solutions to satisfy on-demand requirements while increasing the freedom in the fabrication process. Besides, with the advances in additive manufacturing (AM), the design space of an engineering material can be further enlarged by multi-scale topology optimization. As the interplay between microstructure and macrostructure drives the overall mechanical performance of engineering materials, it is necessary to develop a multi-scale design framework to optimize structural features in these two scales simultaneously. The final part of the study (Ch. 7) presents a concurrent multi-scale topology optimization method of CMMs. Structures in micro and macro scales are optimized concurrently by utilizing sequential quadratic programming (SQP) with the Solid Isotropic Material with Penalization (SIMP) method and a numerical homogenization approach. | en |
| dc.description.abstractgeneral | Cellular materials widely exist in natural biological systems such as honeycombs, bones, and wood. Recent advances in additive manufacturing have enabled us to fabricate these materials with high precision. Inspired by architectures in nature, cellular mechanical metamaterials (CMMs) have been introduced recently as a new class of architected systems. The materials are formed by hierarchical microstructural topologies, which have a decisive influence on the structural performance at the macro-scale. Therefore, the design of these materials primarily focuses on the geometric arrangement of their microstructures rather than the chemical composition of their base material. Tailoring the microstructures of these materials can lead to several outstanding features, such as high stiffness and strength, low density, and high energy absorption. However, it is challenging to design microstructures that satisfy user-defined requirements for properties and material costs. This is mainly due to the trade-off between the accuracy and computing times of the optimization process. In the first part of this study (Ch. 3), a design framework is proposed to overcome this issue. The framework employs a global search algorithm called the genetic algorithm (GA). With a newly designed search algorithm, the framework reduces errors between target and optimized material properties while improving computational efficiency. Inspired by the algorithm behind the GA, the second part of the study (Ch. 4) employs a similar algorithm to identify a material property chart demonstrating all possible combinations of mechanical properties of CMMs. Each axis of the material property chart corresponds to a selected mechanical property, such as Young's modulus or Poisson's ratio, along different directions. The boundary of the property space helps designers understand material performance limitations and make informed decisions in engineering practices. In the fabrication process, unexpected material properties might be achieved due to defects and tolerances in additive manufacturing (AM), such as uneven surfaces, shrinkage of pores, etc. The third part of the study (Ch. 5) investigates the uncertainty propagation on mechanical properties as a result of these manufacturing defects. To investigate the uncertainty propagation problem efficiently, the study uses a deep learning method to predict the variations (stochasticity) of properties. Consequently, the material property space boundary also varies with the uncertainty of properties. In addition to their computational efficiency, deep learning methods are beneficial for solving many-to-many inverse design problems. Traditionally, the global and local search/optimization methods retrieve alternative optimal solutions in their Pareto front set, where each solution is considered to be equally good. A deep learning method called conditional generative adversarial networks (CGANs) can bypass the property calculation to accelerate the simulation process while obtaining a group of candidates with on-demand properties. The fourth part of the study (Ch. 6) employs CGANs to build a new inverse design framework to increase flexibility in the fabrication process by generating alternative solutions for the microstructures of CMMs. Besides, as fabrication technologies have advanced, designing engineering systems has become increasingly complex. Material design is now not only focused on meeting micro-scale requirements but also addressing needs at multiple scales. The interaction between the microstructure (small-scale) and macrostructure (large-scale) significantly influences the overall performance of engineering systems. To optimize structures effectively, there is a need for a design framework that considers these two scales simultaneously. Thus, the final part of the study (Ch. 7) introduces a method called concurrent multi-scale topology optimization. To obtain the extreme performance of a multi-scale structure, this approach optimizes its structure at both micro- and macro-scales concurrently, using gradient-based optimization algorithms with density-based property determination methods in the two scales. | en |
| dc.description.degree | Doctor of Philosophy | en |
| dc.format.medium | ETD | en |
| dc.identifier.other | vt_gsexam:40088 | en |
| dc.identifier.uri | https://hdl.handle.net/10919/119061 | en |
| dc.language.iso | en | en |
| dc.publisher | Virginia Tech | en |
| dc.rights | In Copyright | en |
| dc.rights.uri | http://rightsstatements.org/vocab/InC/1.0/ | en |
| dc.subject | Cellular metamaterials | en |
| dc.subject | inverse design | en |
| dc.subject | material property space | en |
| dc.subject | multiscale modeling | en |
| dc.subject | genetic algorithm | en |
| dc.subject | machine learning | en |
| dc.subject | topology optimization | en |
| dc.subject | uncertainty effects | en |
| dc.subject | numerical homogenization | en |
| dc.title | Optimization and Supervised Machine Learning Methods for Inverse Design of Cellular Mechanical Metamaterials | en |
| dc.type | Dissertation | en |
| thesis.degree.discipline | Mechanical Engineering | en |
| thesis.degree.grantor | Virginia Polytechnic Institute and State University | en |
| thesis.degree.level | doctoral | en |
| thesis.degree.name | Doctor of Philosophy | en |