Multi-Material Structural Topology Optimization Under Uncertainty via a Stochastic Reduced Order Model Approach

This work presents a stochastic reduced order modeling approach for the solution of uncertainty aware, multi-material, structural topology optimization problems. Uncertainty aware structural topology optimization problems are computationally complex due to the number of model evaluations that are needed to quantify and propagate design uncertainties. This computational complexity is magnified if high-fidelity simulations are used during optimization. A stochastic reduced order model (SROM) approach is applied to 1) alleviate the prohibitive computational cost associated with large-scale, uncertainty aware, structural topology optimization problems; and 2) quantify and propagate inherent uncertainties due to design imperfections. The SROM framework transforms the uncertainty aware, multi-material, structural topology optimization problem into a deterministic optimization problem that relies only on independent calls to a deterministic analysis engine. This approach enables the use of existing optimization and analysis tools for the solution of uncertainty aware, multi-material, structural topology optimization problems.