Joint roughness measurement using shadow profilometry

Abstract The shadow cast by a ruler or string reveals a roughness profile across a joint surface. If the rock is smooth, the edge of the shadow is straight, and if rough, the shadow is irregular. The image can be recorded on video tape, digitized by an image analyzing microcomputer, then processed to isolate the shadow edge. Parameters of roughness can then be calculated and displayed. This photoanalysis method compares favourably with the often awkward and impractical mechanical and optical techniques of roughness measurement. In a series of laboratory and field experiments, the authors have investigated errors resulting from variations of illumination and image processing. Results demonstrate that with reasonable precautions, the results can be reproduced to within about 5% of the full roughness range. Roughness parameters convey in a quantitative manner the meaning of “smooth” and “rough”, and are needed as a basis for prediction of shear strength and other properties. The many available alternative parameters, borrowed from fields of electronics, acoustics and mechanical engineering, range from simple measures of amplitude, wavelength and slope, to more complex stochastic, frequency domain and fractal analyses. These are reviewed, and criteria for selection discussed. Shear strength models based on roughness include those of Patton, Ladanyi and Archambault, and Barton. Barton's JRC is not a geometric property: it must be estimated either from shear tests or by comparing a rough joint with the published standard set of comparator profiles. Visual comparison is found to be unreliable. Better results can be obtained by digitizing the standard profiles using the method described in this paper, and relating JRC to a measurable quantity, the roughness profile index, Rp. The regression equation relating JRC to Rp compares closely with the same form of equation determined by direct correlation of the results of shear testing and roughness measurement.