Research On Optimal Algorithm For OpenFOAM Based RBF Mesh Deformation

Mesh deformation is widely used in automobile,aerospace and other industrial pro-duction,as an important technology to solve the problem of moving boundary in Computa-tional Fluid Dynamics(CFD).With the rapid development of high performance computing technology in recent years,research on large-scale numerical simulation of mesh deforma-tion can provide valuable and practical reference for fluid related engineering calculation.At present,the main problems of mesh deformation are that the calculation efficiency and accuracy are not high enough and the mesh adjustment effect is not good enough under the condition of large-scale model and violent movement of objects.Therefore,aiming at the challenge of mesh deformation in numerical simulation,this paper studies on op-timized algorithm for OpenFOAM based RBF mesh deformation.The main work and innovations include:1.The prototype algorithm of RBF mesh deformation based on greedy selection is implemented.(Chapter 2)In this paper,based on the OpenFOAM open source software platform,the function module that satisfies the RBF mesh deformation is added in the numerical solution part,so as to realize the RBF dynamic grid prototype algorithm based on greedy point selection on the platform.2.A block iterative algorithm for RBF matrix system optimization is designed and implemented.(Chapter 3)In this paper,a block iterative algorithm based on RBF matrix system is proposed and tested for different sizes and shapes of grids,and the results show that the com-putational time cost is reduced by more than 90%.3.The parallel optimization algorithm of RBF mesh deformation based on boundary re-decomposition is designed and implemented(Chapter 4)This paper analyzes the whole process of greedy point selection of RBF mesh defor-mation,parallelizes all parallelizable processes,and optimizes the traditional paral-lel algorithm of RBF mesh deformation with the improved boundary re-decomposition and collective communication methods from the aspects of communication and load balancing.The test results show that compared with the traditional parallel method,the communication time is reduced and the parallel efficiency is improved.