SURFACE ROUGHNESS ASSESSMENT OF NATURAL ROCK JOINTS BASED ON AN UNSUPERVISED PATTERN RECOGNITION TECHNIQUE USING 2D PROFILES

The stability of a jointed rock mass is generally controlled by its shear strength that significantly depends on surface roughness. So far, different methods have been presented for determining roughness using 2D profiles. In this study, new method based the unsupervised pattern recognition technique combination statistical, geostatistical, directional, and spectral quantification will be proposed. To reach goal, more than 10,000 profiles gathered from 92 surfaces natural joints were scanned. samples collected limestone cores Lar Dam located in Mazandaran Province, Iran. After introducing index, determined fast Fourier transform measuring unevenness rough profiles, features revealing waviness extracted, representative vector profile each introduced through weighted mean median features. Principal component analysis (PCA) was utilized finding direction maximum variance information. Then, clustering performed via K-means, silhouette measure used order to find optimal number clusters resulted creation 13 clusters. verify procedure, sample selected cluster, direct tests samples. Comparing experiments results shows they are good agreement. Thus, an efficient tool quantitative considering surface.