Primary Submission Contact
Arts and Sciences
Topology optimization is recognized as the most effective numerical method to generate high-performance architectural layouts. Recently, multi-material topology optimization (MMTO) methods have been proposed. These methods provide the ability to synthesize structures with a plurality of materials, increasing design performance. While successful, most of these methods are extensions of classical topology optimization and rely on traditional, gradient-based optimization. Our work introduces a new bio-inspired MMTO strategy without the limitations of current approaches. We aim to produce a hybrid kinetic Monte Carlo method for multi-material optimization inspired by the hybrid Cellular Potts Model (CPM) algorithm. We implement: input parameters for an arbitrary number of cell types modeling spatial constraints, adhesion, and the specification of details concerning the transition rule. The CPM model is coupled with Finite Element Analysis (FEA) in a hybrid structure. The model will employ equations from topology optimization (i.e., compliance) during the development of new transition rules that produce emergent optimization. The model is modified to the point where the structure of the CPM can be related to field variables of solid mechanics. The usage of a gradient-free Monte Carlo method makes the material dynamics at each point depend on local conditions, as opposed to global interactions.
Shomer, Thomas; Scheiner, Aaron; Sego, Timothy James; and Tovar, Andres, "Multi-material Topology Optimization using a Cellular Potts Model" (2020). Fall Interdisciplinary Research Symposium. 129.
Additional Presentation Information