Application of Non-Destructive Test Results to Estimate Rock Mechanical Characteristics—A Case Study

Accurately determining rock elastic modulus (EM) and uniaxial compressive strength (UCS) using laboratory methods requires considerable time cost. Hence, the development of models for estimating mechanical properties is a very attractive alternative. The current research was conducted to predict UCS EM sandstone rocks quartz%, feldspar%, fragments%, compressional wave velocity (PW), Schmidt hardness number (SN), porosity, density, water absorption via simple regression, multivariate regression (MVR), K-nearest neighbor (KNN), support vector (SVR) with radial basis function, adaptive neuro-fuzzy inference system (ANFIS) Gaussian membership (GM) back-propagation neural network (BPNN) based on various training algorithms. samples were categorized as litharenite feldspathic litharenite. By increasing feldspar% quartz% decreasing static increased. results statistical analysis showed that SN porosity have greatest effect EM, respectively. Among Levenberg–Marquardt (LM), Bayesian regularization, Scaled Conjugate Gradient algorithms BPNN method, LM achieved best in forecasting EM. ideal obtained BPNN, trial-and-error process, contains four neurons hidden layer eight inputs. All five attained acceptable accuracy (correlation coefficient greater than 70%) properties. comparing methods, ANFIS higher precision other methods. can be determined high (R2 > 99%).