The simulated rough surface with desired parameters is widely used as an input for the numerical simulation of tribological behavior such as the asperity contact, lubrication, and wear. In this study, a simulation method for generating non-Gaussian rough surfaces with desired autocorrelation function (ACF) and spatial statistical parameters, including skewness (Ssk) and kurtosis (Sku), was developed by combining the fast Fourier transform (FFT), translation process theory, and Johnson translator system. The proposed method was verified by several numerical examples and proved to be faster and more accurate than the previous methods used for the simulation of non-Gaussian rough surfaces. It is convenient to simulate the non-Gaussian rough surfaces with various types of ACFs and large autocorrelation lengths. The significance of this study is to provide an efficient and accurate method of non-Gaussian rough surfaces generation to numerically simulate the tribological behavior with desired rough surface parameters.

Experimental results have confirmed that parallel rough surfaces can be separated by a full fluid film. However, such a lift-off effect is not expected by the traditional Reynolds theory. This paper proposes a deterministic mixed lubrication model to understand the mechanism of the lift-off effect. The proposed model considered the interaction between asperities and the micro-elastohydrodynamic lubrication (micro-EHL) at asperities within parallel rough surfaces for the first time. The proposed model is verified by predicting the measured Stribeck curve taken from literature and experiments conducted in this work. The simulation results highlight that the micro-EHL effect at the asperity scale is critical in building load-carrying capacity between parallel rough surfaces. Finally, the drawbacks of the proposed model are addressed and the directions of future research are pointed out.

The effects of surface roughness characteristics on the fluid load capacity of tilt pad thrust bearings with water lubrication were studied by the average flow model. The flow factors utilized in the average flow model were simulated with various surface roughness parameters including skewness, kurtosis and the roughness directional pattern. The results indicated that the fluid load capacity was not only affected by the RMS roughness but also by the surface roughness characteristics. The fluid load capacity was dramatically affected by the roughness directional pattern. The skewness had a lower effect than the roughness directional pattern. The kurtosis had no notable effect on the fluid load capacity. It was possible for the fluid load capacity of the tilt pad thrust bearings to be improved by the skewness and roughness direction pattern control.

The application of the spectral representation method in generating Gaussian and non-Gaussian fractal rough surfaces is studied in this work. The characteristics of fractal rough surfaces simulated by the spectral representation method and the conventional Fast Fourier transform filtering method are compared. Furthermore, the fractal rough surfaces simulated by these two methods are compared in the simulation of contact and lubrication problems. Next, the influence of low and high cutoff frequencies on the normality of the simulated Gaussian fractal rough surfaces is investigated with roll-off power spectral density and single power-law power spectral density. Finally, a simple approximation method to generate non-Gaussian fractal rough surfaces is proposed by combining the spectral representation method and the Johnson translator system. Based on the simulation results, the current work gives recommendations on using the spectral representation method and the Fast Fourier transform filtering method to generate fractal surfaces and suggestions on selecting the low cutoff frequency of the power-law power spectral density. Furthermore, the results show that the proposed approximation method can be a choice to generate non-Gaussian fractal surfaces when the accuracy requirements are not high. The MATLAB codes for generating Gaussian and non-Gaussian fractal rough surfaces are provided.