Prediction Performance and Generalization of the Empirical Estimation of Rockmass Deformation Modulus Based on Rockmass Classification Systems

The deformation modulus of rock mass can either be measured using in-situ field tests or estimated empirically. However, the empirical method is an attractive and open choice for researchers to estimate the deformation modulus of rock mass economically, efficiently and accurately and many researchers have proposed empirical equations for estimation the deformation modulus of rock mass based rock mass classification systems. In this paper attempt has been made to collect and review some these published empirical equations developed so far based on Rock Mass Rating, Geological Strength Index, and Q rock mass classification systems. These empirical equations based on each classification system are grouped on the based on similar parameters. All the equations within certain group for all groups were plotted to obtain a viable range for which the equation is assumed to be reliable. An artificial data is generated for each group separately to from valid data points of existing equations. Keeping in view the trend of data points in the scatter plot of individual group, Gaussian and Sigmoidal type mathematical functions are fitted to obtain generalized equation for each group that consisted two or more than two valid equations. The new equations fitted to the data are optimized by minimizing residuals using Microsoft Excel built in Add-in “Solver”. It is observed that equations using sigmoidal function predict more precisely than equations using Gaussian function.