Distribution characteristics of morphology parameters and an intelligent roughness prediction method characterized by multiple parameters

The morphology characteristics are key factors affecting the mechanical properties of structural plane. In this study, a systematic study is carried out to explore the actual roughness characteristics of standard profiles, the sensitivity of morphology parameters and the roughness prediction method with multi-parameter characterization. According to the statistical analysis of 552 pieces of literature on structural plane roughness, eight morphology parameters that are most representative are proposed and in-depth analysis is made on the relationship between each parameter and roughness coefficient. By smoothing the standard profiles and preparing structural plane specimens, the morphology characteristics and mechanical properties of structural plane smoothed for different times are put into scrutiny. A comprehensive analysis is conducted to investigate the correlation between morphology parameters and first and second-order roughness, as well as the relationship between first and second-order roughness and shear mechanical properties. The specimens of structural plane were manufactured via 10 standard profiles. Detailed discussion is made to analyze the deformation, failure and strength characteristics of standard structural plane. Based on the morphology parameters of standard profiles and the peak strengths of standard structural planes, the entropy weight-TOPSIS method is adopted to determine the actual rough characteristics of standard profiles. 86 sets of profile data were collected to establish the profile-morphology parameter-JRC database. The overlapping information between the eight morphology parameters is extracted by factor analysis. A stacked regression model incorporating multiple representative morphology parameters is proposed to predict the roughness characteristics of the structural plane.