Novel and Efficient Mesostructure Selection Approaches for Multi-Scale Thermo-Mechanical Topology Optimization
Date of Award
12-12-2024
Degree Name
M.S. in Mechanical Engineering
Department
Department of Mechanical and Aerospace Engineering
Advisor/Chair
Robert Lowe
Abstract
The next generation of aircraft requires novel structural designs capable of enduring extreme multiphysics environments. Multiscale topology optimization – where macroscale properties and mesoscale topology are either concurrently or sequentially optimized – provides a compelling framework for addressing this complex structural design problem. In the sequential scheme utilized herein, targeted thermo-mechanical properties specified by a macroscale optimization are subsequently matched with a mesoscale unit cell topology and corresponding homogenized properties from a mesoscale optimization. Although this sequential approach provides greater design flexibility than its concurrent counterparts, a key drawback is significant computational expense. To mitigate this expense, this thesis introduces two novel and efficient mesostructure selection approaches, both predicated on jettisoning the computationally expensive mesoscale optimization. In its place, the proposed methods select – from a series of pre-existing designs generated prior to the macroscale optimization (thus adding no computational time) – the unit cell topology that best matches the targeted thermo-mechanical properites of the voxel. Our framework is first applied to a thermo-mechanically loaded Messerschmitt-Bolkow-Blohm beam, with the resulting designs exhibiting a compliance within 1% and a resistivity within 4% of the optimized solution, while simultaneously reducing computational burden from months to minutes. This framework is then applied to a thermally loaded bi-material ring problem, with an Inconel 718 ring encased by a carbon composite ring. The topology of the Inconel 718 ring was designed by the framework to accommodate the strain mismatch between the two materials caused by differing coefficients of thermal expansion. Experimental results indicate that the bi-material ring performed effectively and successfully accommodated the strain mismatch between the two rings. Overall, this research demonstrates that the proposed mesostructure selection approach shows great promise for achieving near-optimal thermo-mechanical designs with revolutionary reductions in computational expense.
Keywords
topology optimization, bi-material ring, multi-scale, thermo-mechanical, structures, mesostructure
Rights Statement
Copyright © 2024, author.
Recommended Citation
Meixner, Eddie, "Novel and Efficient Mesostructure Selection Approaches for Multi-Scale Thermo-Mechanical Topology Optimization" (2024). Graduate Theses and Dissertations. 7479.
https://ecommons.udayton.edu/graduate_theses/7479
