A Library Based Overset Capability Development for Density- and Pressure-Based Flow Solvers

Abstract

Computational simulation of flows involving bodies in relative motion is a challenging problem, which requires the capability of relocating the grid associated with moving bodies as well as the formulation of grid speed terms in the flux evaluation of the governing equations. An overset grid framework enables the computational fluid dynamics (CFD) flow solver to model this class of problems. However, implementing this technology into an existing CFD code can be very tedious and time consuming due to the need for laborious book-keeping as well as proper data interpolation between different layers of grids. This paper describes an overset grid approach to alleviate these problems and enhance the capability of flow solvers to handle problems involving bodies in relative motion. A library approach has been developed and adopted to handle the bookkeeping of the overlapping grids among different layers and the transfer of information across them. To have a wide applicability, this library approach should be able to support structured, unstructured, and generalized grid topologies. This approach makes it easier to incorporate the overset capability into legacy CFD solvers. In the present study, a set of developed overset grid libraries (SUGGAR and DiRTLib) was implemented into a density-based generalized grid flow solver and a pressure-based structured grid flow solver as demonstration. This implementation of the overset capability has been validated with static and moving grids and is presented in the paper.