A New Level-Set-Based Robust Topology Optimization Approach with Applications to Design of Phononic Metamaterials

Suny At Stony Brook, Stony Brook NY

Investigators

Abstract

Topology optimization is a computational design approach for efficiently generating material geometries with desired properties. However, these geometries are often difficult to manufacture or may require additional post-processing before they can be realized using additive manufacturing techniques. In addition, multi-physics, multi-material problems involving uncertainty are not well addressed with conventional topology optimization methods. This award supports fundamental research towards the development of a unified computational framework that combines level-set-based topology optimization methods, robust design, additive manufacturing, and experimental characterization to address these challenges. This approach seamlessly integrates generative topology optimization with additive manufacturing. The resulting methods and tools will enable innovative design of next-generation multifunctional materials. The findings of the research will be broadly disseminated to the academic community and general public. The teaching and outreach activities will improve the quality and effectiveness of education for teachers and students at the K-12, undergraduate and graduate levels, with a significant effort towards involvement of underrepresented groups. In this project, a novel level-set-based automated computational design method will be developed to provide seamless integration of topology optimization, multiphysics mesh-free finite element analysis, robust design, and additive manufacturing. This will enable systematic design and efficient realization of high-quality and high-throughput multifunctional metamaterials. The goal of this research will be accomplished by: (1) conducting high-fidelity simulations on acoustic/elastic wave propagation in phononic metamaterials using an element-free Galerkin method; (2) devising new numerical methods to handle large-scale topology optimization problems with non-convex objective functions and multiple nonlinear constraints; (3) developing new design methods for simultaneous robust shape and topology optimization with level-set-based dynamic topological design; (4) exploring new methods that leverage the information embedded in geometric level-set models to enable seamless integration of topology design with additive manufacturing; (5) cross validating the design methods and models through experimental tests. The combination of numerical and experimental studies will provide a better scientific understanding of the underlying multiphysics mechanisms in multifunctional phononic metamaterials. New techniques for accurate multi-physics modeling, efficient large scale topology optimization and uncertainty quantification, and uncertainty propagation will be explored.