Geometry Parameterization and Computational Mesh Deformation by Physics-Based Direct Manipulation Approaches

Abstract

Consistent geometry parameterization of different models used in multidisciplinary optimization is one of the major issues for efficient design optimization. In this research, physics-based free-form deformation and radial-basis-function-based deformation approaches that allow a direct manipulation of an arbitrary object are proposed for efficient geometry manipulation and parameterization in the optimization processes. The developed methods can deform any computational mesh mapped on the deformable geometry at once. The methods have several advantages for geometry parameterization, such as easy handling of complex objects, direct manipulation, local controllability, and intuitive deformation as real elastic objects. In addition, they offer a straightforward solution for constructing a hierarchical/adaptive parameterization that can benefit the optimization algorithmic performances. Application to an aerodynamic design optimization showed that the methods were effective for efficient design optimization.