Investigation of the effects of nonstationary features on rock joint roughness using the laser scanning technique

Accurately estimating rock joint roughness is crucial for understanding the shear mechanism and permeability behavior of a rock mass. Several influencing factors, including anisotropy, measurement noise, and the scale effect and sampling interval, have been considered. However, little attention is paid to the influences of nonstationary features on the roughness assessment. In this study, a portable laser scanner was employed to collect high-density 3D point clouds of ten natural rock joint specimens. Based on two parameters, namely, the bright area percentage (BAP) and θ*_max/(C + 1), where θ*_max is the maximum apparent dip angle and C is a dimensionless fitting parameter, the rock joint roughness was determined before and after removing nonstationary features, and a comparison showed that nonstationary features have a considerable influence on the roughness. Subsequently, an approach was proposed to remove nonstationary features through the conversion of spatial coordinates, and an application to a roughness evaluation illustrated that similar trends are observed between the BAP and θ*_max/(C + 1) with respect to the point clouds of ten rock joints whose nonstationary features were removed. These findings reveal that nonstationary features should be removed to improve the accuracy and comparability of the roughness assessments.