A GPU-Accelerated 3D Mesh Deformation Method Based on Radial Basis Function Interpolation

In this paper, we developed a GPU parallelized 3D mesh deformation based on the radial basis function (RBF) interpolation using NVIDIA CUDA C++. The RBF mesh deformation method interpolates displacements of the boundary nodes to the whole mesh, which can handle large mesh deformations caused by translations, rotations, and deformations. However, the computational performance of RBF mesh deformation depends on the quantity of grids. For 3D mesh deformation, especially for mesh with large number of boundary nodes, RBF interpolation has been verified computationally intensive. By introducing GPU acceleration, the computational performance of RBF mesh deformation code improved significantly. Several benchmark test cases were carried out to show the feasibility and efficiency of GPU parallelization. In summary, the GPU parallelized RBF interpolation shows the potential to become an alternative way to deal with 3D mesh deformation problems in an efficient way.