Periodic Numerical Rough Surface Filtered Generation

Abstract Numerical surface filtered generation is one of the main methods for generating numerical rough surfaces, but when faced with rough surfaces with waviness or large periodicity, traditional filtering methods cannot be implemented well. Because of this, the paper adopts the method of decomposing and synthesizing the maximum period and random part of the periodic rough surface. By decomposing the statistical parameters of the target surface, the statistical parameters of the ideal periodic surface and the random surface are generated, respectively, and then according to the surface parameters generate the surfaces and synthesize them. By comparing the statistical parameters and morphology of the synthesized surface with its actual surface, it can be found that this method can well achieve the generation of periodic rough surfaces, which is a good improvement to the original filter generation method.