Automated Three-Dimensional Polyhedral Mesh Generation for B-rep Solid Model Data

Recent advances in the design of products and processes have been based in large part on advanced numerical simulation techniques used by the computer aided engineering community. These approaches are applied to the development of cars, military hardware, food processing operations, pollution control equipment, and nuclear reactors for power generation. With current tools, creation of the high quality computational grids that form the underpinnings of these methods is arduous, and can often be the most time-consuming step of the process. Current methods for creating computational meshes either favor highly automated approaches that result in low quality meshes and low fidelity simulations, or user-intensive methods that provide high quality results requiring much more time. The gulf between a high level of automation and high quality results will be bridged by adopting advanced cell topologies along with novel optimization techniques. The Phase I effort will demonstrate the effectiveness of the proposed method by developing initial versions of the meshing algorithms, which will then be applied to several computational geometries. Objective quality criteria will be used to compare the resulting meshes to those generated with commercially-available meshing software using both user-directed and automated approaches. Adoption of the proposed meshing algorithms provides the opportunity to significantly reduce the resources needed for the analysis and development of improved products and processes across a range of industries. This will result in reduced development cycles, improved products, or both. The software resulting from the development of these methods can be embodied as stand-alone programs, as well as software libraries that can be incorporated into other software products.