Using experimental measures in elaboration and calibration of numerical models in geomechanics

Abstract

This work is the result of over 10 years research on computational methods for solving problems related to mining and underground excavations. Real cases are described where numerical models (FEM) have been implemented. Measured data sets used to develop and calibrate the models are presented. Some of the parameters obtained were: elastic modulus of rock mass, Poisson coefficient, rock mass jointing, stresses and strains within the rock mass, vertical displacement at surface, extent of plastic zone around excavations, convergence at driveways and tunnels, number of rockfalls from roof strata, etc. Those measurements were used for FEM predictive models applied to: surface subsidence, convergence in roadways and interaction between roof strata and shield-type supports (longwall mining method). Empirical measurements used to back-analyse the models is plotted against results from the models for a proper comparison.