Shot peening is the efficient method for metal surface modification and performance improvement. However, there is still no effective way to establishing the mathematical model for shot peening surface reconstruction, resulting in restricting the correlation study between shot peening surface properties and morphology. The difficulty of shot peening surface reconstruction lies in how to accurately characterize the roughness surface height and texture features. Therefore, a novel method of generating non-Gaussian sequences with specified height roughness parameters is proposed in this paper. Fast Fourier transform (FFT) method gets improved combined with the new method, which overcome the predicament that the unimproved FFT cannot ensure the height features in the reconstructed shot peening surfaces. In addition, a new autocorrelation function in shot peening surfaces is proposed to accurately characterize the reconstructed surface texture. The experimental results show that with the improved FFT method and the new autocorrelation function introduced to shot peening surface reconstruction, the maximum error of the seven height roughness parameters in the reconstructed surface is 2.201%. And the texture features and height distribution of the reconstructed surface are in good agreement with the measured surface.

Among the 26 roughness parameters described in ISO 25178 standard, the parameters used to characterize surface performance in characterization parameter set (CPS) lack scientificity and unity, resulting in application confusion. The current CPS comes from empirical selection or small sample experiments, thus featuring low generality. A new method for constructing CPS in rough surfaces is proposed to solve the above issues. Based on a data mining method, statistical theory, and roughness parameters definitions, the 26 roughness parameters are divided into CPS and redundant parameter sets (RPS) with the help of reconstructed surfaces and machining experiments, and the mapping relationships between CPS and RPS are established. The research shows that RPS accounts for 50%, and CPS, of great significance for surface performance, and has the ability to fully cover surface topography information. The birth of CPS provides an accurate parameter set for the subsequent study of different surface performance, and it provides more effective parameters for evaluating the workpiece surface performance from the same batch.

Roughness surfaces contact analysis is an advanced research topic in interface design. The 3D rough surface amplitude distribution characterized by height distribution parameters(Sq (root mean square), Ssk (skewness), Sku (kurtosis)) has a great influence on the extreme value and distribution of the interface contact stress. However, the relationship between height distribution parameters and surface maximum mises stress (σmax) is still unclear and lacks of in-depth study. With the assistance of roughness surface reconstruction and contact stress algorithm proposed by the research group, σmax under a large sample was calculated and used as the data support for correlation analysis. Through BP neural network, global sensitivity qualitative (Morris) and quantitative (Sobol) analysis methods, the relationship between Sq, Ssk, Sku and σmax under different loads is studied. Based on complete polynomial and permutation combination method, the optimal correlation model between height distribution parameters and σmax was established, and particle swarm algorithm was introduced to analyze σmax extreme values under different Sq. The results show that: (1) Under different loads, the order about height distribution parameters influence on surface contact stress is: Sq> Ssk > Sku, and as the load increases, the influence of Ssk and Sku gradually decreases. (2) In different roughness surfaces, the influence of Ssk and Sku on the contact performance is significantly full of discrepancy. The research results provide reference and technical support for active design of rough surface microstructure to improved contact performance.

Roughness surfaces contact analysis is cutting-edge research in interface design. The 3D rough surface amplitude distribution characterized by height distribution parameters(Sq (root mean square), Ssk (skewness), Sku (kurtosis)) has a great influence on the extreme value and distribution of the interface contact stress. However, the relationship between height distribution parameters and surface maximum mises stress (σmax) is still unclear and lacks of in-depth study. Through BP neural network, global sensitivity qualitative (Morris) and quantitative (Sobol) analysis methods, the relationship between Sq, Ssk, Sku and σmaxunder different loads is studied. Based on complete polynomial and permutation combination method, the optimal correlation model between height distribution parameters and σmax was established, and particle swarm algorithm was introduced to analyze σmax extreme values under different Sq. The results show that: (1) Under different loads, the order about height distribution parameters influence on surface contact stress is: Sq> Ssk > Sku, and as the load increases, the influence of Ssk and Sku gradually decreases. (2) In different roughness surfaces, the influence of Ssk and Sku on the contact performance is significantly full of discrepancy. The research results provide reference and technical support for active design of rough surface microstructure to improved contact performance.

With surface roughness restricted by grinding parameters, the characterization of roughness parameters and the inversion of grinding parameters are of great significance for improving surface performance and realizing active surface machining. This research proposes a combination of statistical theory and data-driven analysis to solve the above problems. Pearson correlation analysis and multivariate variance analysis indicate the correlation characterization parameter set (CPS) consists of Sa, Vmp, Vvv and Sz and that there are differences in the influence of grinding parameters on the parameters in CPS. Adjustment of support vector machine (SVM) core parameters makes it possible to construct expansion parameter set (EPS) optimal inversion models. By designing pseudo-surface random roughness parameters and grinding experiments, the reliability of inversion models is verified. The results show: 1) The better generalization of inversion model indicates skewness Ssk and kurtosis Sku in EPS have important implications for the optimal inversion model and surface characterization. 2)The data-driven model based on support vector machine provides machining guidance for obtaining the expected ultrasonic grinding surface.