The Simulation Of Rough Surfaces Based On The FFT And The Linear Transformation Methods

During the interaction of the rough surfaces, the property of contact, lubrication, friction and wear between them largely depends on the microstructure of the rough surfaces. But the existing numerical models of rough surfaces use simplified model, and the numerical simulation results between real surfaces and simplified surface are quite different. The research of the characterization and simulation of the surface microstructure is a research hotspot of interface mechanical analysis both in domestic and overseas. It is particularly important to characterize the morphology and functional property of rough surfaces accurately and build the model of surfaces effectively. However, there has no comprehensive research method of rough surfaces simulation so far. Therefore, this paper, based on the theory of random process and time series analysis, uses different methods to simulate the rough surfaces and studies the specified autocorrelation function and height distribution function. The main research work of this paper include as the following:(1) Through analyzing the morphological property of rough surfaces, introduces the definition of the surface roughness and the computing method, thus determines the simulation method of moving average model (MA) based on the autocorrelation function and height distribution function of the rough surfaces.(2) A new method is presented, which is based on MA model, Fast Fourier transform and Johnson translator system to simulate Gaussian and non-Gaussian rough surfaces, and build accurate and effective numerical models of rough surfaces.(3) Using the linear transformation to simulate the rough surfaces. The research, based on MA model, uses the method of solving the autocorrelation coefficient matrix and Johnson translator system to simulate rough surfaces. In order to solve the autocorrelation coefficient matrix of rough surfaces, uses the Newton iteration method and the Nonlinear conjugate gradient method to solve the Nonlinear equations which is of low efficiency and poor convergence. We use C language and Intel MKL (Math Kernel Library) to compile the program of solving the Nonlinear equations, which greatly improves the simulation speed and the practicability of linear transformation method.(4) Based on the actual process of workpiece surfaces, uses FFT and linear transformation method to simulate the actual surfaces. Proposes two effective methods to simulate the actual surfaces, and compiles the programs based on the actual surfaces by establishing methods from measuring, characterizing to simulating the rough surfaces effectively. The FFT method is faster than the linear transformation method, while the latter is more stable than the former. Chart79, Table5, Bibliography69...