Highly Efficient Wall-Distance-Based Parallel Unstructured Overset Grid Assembly

Abstract

The overset grid assembly (OGA) method is an effective computational fluid dynamics technique for the simulation of flows with complex geometries or relative moving parts. Using OGA, the effort for mesh generation is significantly reduced by generating the grid for different parts and then assembling them to form the computational grid. New methods have been developed that are generally divided into implicit and explicit methods in which a donor cell search for the defined query points is often time-consuming. Besides, the parallelization of OGA is a complex task that is not currently supported by many software platforms. In this paper, a new OGA approach is developed based on the wall distance and a user-defined reference length. The proposed approach is relatively more convenient to implement and more robust among the existing methods. The number of query points defined by this method is also significantly reduced, resulting in a higher speed of performing the overall donor search process. Validity tests are conducted, and the results confirm the effectiveness of the proposed method.