Automation of a Design Optimization Process for Fiber Reinforced Polymer Sandwich Structures 2021-01-0363

Compared to traditional materials, carbon fiber reinforced polymers (CRFPs) have allowed designers to design stiff, light-weight structures, but at the cost of increased complexity in the design process. In this paper, the automation and optimization of the composite design process and how it affects design space exploration are evaluated. Specifically investigated is the design process for CFRP sandwich structures using the third-party optimization software modeFRONTIER. For given surface geometry and load cases, the approach aims to explore the Pareto frontier for the minimization of mass while constraining stiffness parameters. In this approach, the problem is framed as a single integrated optimization problem. In each optimization iteration, this method updates the CAD geometry and discretization of plies across the structure before exporting the model for Finite Element Analysis (FEA). Before solving the FEA model with the HyperWorks OptiStruct solver, the optimizer applies each laminate’s number of plies, layup order, core sizing, material selection, and orientation by automatically modifying the corresponding Tcl/Tk script. This proposed method of automation allows the optimization of composite structures in one optimization step with full control of the design of experiments and optimizer selection. To demonstrate this approach, a structural component from a composite monocoque vehicle is selected illustrating a set of loading conditions and design targets.