Predicting uniaxial compressive strength of rock during Electrical Resistivity monitoring by multivariate regression method

Authors

Abstract

Petrophysical and Geomechanical properties of rocks are important parameters in the design of engineering works and classification of rocks for engineering purposes. Recent studies indicate that geophysical methods, especially seismic and electrical, are able to estimate mechanical parameters and recognize spatial variations. In this research, to develop a predictive model for the uniaxial compressive strength (UCS), special electrodes were installed on the saturated core samples and simultaneously, the uniaxial compressive strength test and electric current flowing through the samples was done and variation of electrical resistivity during loading was measured in the laboratory. The results indicated that the structure and texture of rock had an important effect on the resistivity behavior during a mechanical loading.

In this study, thirty core samples from the Fault breccias and Bimrocks (Block-in-matrix-rocks), were collected from different locations of Sabzkouh tunnel route in Chahar Mahal and Bakhtiari Provence. Regression analysis showed that there were generally strong correlations between the UCS and Resistivity in the samples having volumetric block proportion (VBP) of 25–75%. Multiple regression equations were derived for the prediction of UCS based on the resistivity and VBP values. The coefficient of determination (R2) and the root mean square error (RMSE) and the geometric mean error ratio (GMER) indices were calculated as 89.13%, 8.683 and 0.911, respectively, to characterize the prediction performance of the MLR model. The statistical test showed that the MLR model was valid and acceptable for predicting UCS.

Keywords

Files

History

  • Receive Date: 05 December 2022
  • Revise Date: 09 February 2023
  • Accept Date: 20 February 2023

Share

How to cite

Statistics