Estimation of Young's Modulus of Limestones using Multi-Layer Perceptron

Abstract

The Young’s modulus (E) is a very important parameter used in many engineering projects and in the petroleum industry. It is especially important for tunneling, mining and rock slope stability analysis. This parameter is determined by difficult experiments. In addition, cores must be taken for the experiment and the cores taken must be of high quality. The aim of the study is to estimate the Young’s modulus, which represents the basic mechanical property of rocks, using relatively easy-to-apply and low-cost methods. For this purpose, the multi-layer perception method was used. Input parameters of these meshes are Dry density, Water saturated density, Bulk density, Porosity, Water absorption, Ultrasound Pulse Velocity (UPV), Poisson ratio (v), Tensile strength (To), The uniaxial compressive strength (UCS) and The point load index (Is)' is Four different network models were created and the successes of these network models were compared using the 5-fold cross-validation method. As a result of the comparison, it was understood that the model 2 network was more successful. The Correlation coefficient values of the model were calculated as 95% in training and 84% in 5-fold cross validation.