Abstract

The use of the finite element method within an optimization workflow is fraught with challenges that limit the automation of such workflows. These challenges are inherent to the traditional finite element formulations which are heavily dependent on a manual meshing process that introduces variability that is challenging to account for within an automated workflow. The recently developed flex representation method (FRM) provides a salient solution to the manual meshing process without sacrificing solution accuracy. In response to the development of FRM a global automotive company requested a study to explore the applicability of FRM to one of their sizing-optimization problems: the constrained optimization of a wheel undergoing a rigidity test. In this study we develop an optimization framework based on the DAKOTA optimization framework, the open-source FreeCAD computer-aided design software, and an implementation of FRM within the Coreform IGA solver. Within this framework we demonstrate in the affirmative that FRM enables a highly robust sizing-optimization workflow while requiring minimal effort to prepare the FRM model.

Degree

Rights

https://lib.byu.edu/about/copyright/