SOGA: Toward Developing and Application of an Intelligent Structured Chimera Grid Assembling Library for Moving Objectives

Abstract There are many demands to establish the capability to simulate problems with moving objectives, e.g., safety evaluation of multi-bodies separation, maneuver flight simulation, landing process simulation, and so on. The Chimera grid approach is an efficient and flexible way to solve these problems. Based on authors' previous work on the static automatic Chimera grid approach, this paper presents an intelligent structured Chimera grid approach for problems with moving objectives. The idea is to introduce a thought on the modular and application programming interface (API) based software development strategy into the procedure to develop a scalable and reusable software library, SOGA, for Chimera grid applications based on multi-block structured grids. The full name of SOGA is read as Structured-grid based Overset Gridding Assembler, which consists of several type of APIs that include definition of basic grid properties, grid registration, preprocessing for hole-mapping and alternating digital tree (ADT) construction, connectivity establishing, and I/O management. The algorithms hiding behind for 'hole cutting' automatically and intelligently include a modified hole-mapping method through inverse marked method and vector ray method for hole profiling, an alternating digital tree (ADT) based method for grid assembly, and a modified cut-paste algorithm for overlap optimization. To demonstrate the feasibility and efficiency of the procedure, the approach has been implemented and coupled into the CFL3D solver, which is a well-known industry-oriented CFD software and published by NASA in recent years as an open source code, and doesn’t have Chimera grid capability for moving objectives. Several test cases both including static and dynamic with moving objectives are selected to validate the algorithms and the procedures, and some details numerical results are shown to compare with experiments and public literatures. From the results, one can conclude that the present approach is feasible and effective both for static and dynamic Chimera grid applications.