Experimental study on the morphological characteristics of artificially split rock joints and their controlling factors

Understanding the thermal-hydraulic-mechanical behavior of rock joints is essential in many civil and rock engineering applications, but controlled testing on natural joints is often impractical. This study investigates artificial tensile joints generated by wedge-splitting in medium- to coarse-grained granite specimens and evaluates the effects of loading rate, specimen size, and loading direction on surface morphology. Block and cylindrical specimens were tested under loading rates of 0.0001~0.01 mm/s. Joint surfaces were digitized via 3D laser scanning; directional profiles at 15 increments were used to compute three roughness parameters including CLA (Center Line Average), Z2 (root mean square of the first derivative of a profile), and fractal dimension (D). Spatial correlation was quantified using the experimental variogram with spherical fits to obtain range and sill. No discernible rate dependency was detected in either roughness or variogram parameters within the tested range. Size effects were parameter-specific: CLA increased with specimen size, whereas Z2 and D were relatively higher in the smallest specimens. Roughness showed a clear directional effect, being greater perpendicular to the loading axis than parallel. Variogram parameters increased with size (both range and sill), while directional differences in spatial correlation were weak and generally not statistically significant. Consistent with the amplitude-correlation relationship, the sill related to CLA, and a normalized initial slope (sill range) qualitatively related to Z2 and D. Overall, specimen size and loading direction govern the scale of roughness that is emphasized, providing practical guidance for the design and interpretation of experiments using artificial joint specimens.