Novel Topology Optimization Methods for Designing Multifunctional Heterogeneous Material Systems

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

New advances in additive manufacturing have enabled rapid and inexpensive fabrication of complex designs containing multiple classes of materials, such as metals, ceramics, and polymers, all within a single print job. This capability allows manufacturers to produce heterogeneous, multi-component designs that do not require assembly, thereby eliminating the cost and design constraints imposed by the assembly process. It also can allow designers to produce higher-performing and/or lighter-weight components than is possible through traditional means. However, our capacity to manufacture such designs currently exceeds our capability to optimize them. This research project will investigate novel algorithms for generating optimal structural geometries and material layouts for multifunctional, heterogeneous systems. These algorithms will consider the behavior of the different types of materials used in a design and the interactions among the materials. In this way, we will enable system-level optimization of multicomponent designs that will exhibit cost and performance properties far exceeding components created using traditional approaches. This novel design approach has the potential for impacts across multiple research communities and industries, with applications ranging from biomedical devices to consumer electronics. Through this research, we will introduce a new design formulation for handling multifunctional design problems containing several classes of material. This objective will be achieved through the development of novel topology optimization-based algorithms that will combine nonlinear finite element analysis with an original parametric design representation scheme, created specifically for heterogeneous problems. The numerical optimization problem will be solved using gradient-based methods in which function gradients will guide the optimization search of the design space. These gradients will be computed using adjoint sensitivity analysis, and will take into account the various sources of nonlinearity being investigated, including hyperelasticity, plasticity, and material damage. The resulting design algorithms will be validated using high-fidelity computational modeling of the optimized designs, along with analysis of analogous commercially-available designs, to quantify the performance gains generated by each algorithm.