A Novel Adaptive Non-uniform Conformal Meshing Method for Efficient FDTD Simulation

In the process of generating conventional FDTD meshes, the model voxelization often takes up too many computational resources. To address this issue, this paper is devoted to presenting an effective mesh voxelization technique with adaptive partitioning and non-uniform grid generation. In this scheme, the entire computational domain is divided into subintervals based on the model's demarcation points, and the mesh generation parameters can be adaptively adjusted according to the information of each sub-interval in order to generate a reasonable non-uniform mesh. Besides, the scheme corrects mesh generation conditions for voxels across the model interface and records their conformal information, resolving the staircase approximation issue caused by traditional voxelization. The proposed mesh generation technique can minimize the number of grids while ensuring the geometrical accuracy of the model. The mesh quality is further validated by applied to meshing some typical objects in computational electromagnetics problems.