Study on creep models and parameter inversion of columnar jointed basalt rock masses

Abstract

Taking the Baihetan hydropower station in Southwest China as an example, creep models and parameter inversion of the columnar jointed basalt rock mass were investigated in this study. Based on the large in situ flexible double pillow bearing plate compression creep tests, the applicability of empirical creep models, such as the power function model, the logarithmic model and the exponential model, for the creep characteristics of columnar jointed basalt rock masses in the dam site area was analyzed first. Then, the compression creep formula of the measuring points in the middle of the flexible double pillow pressure plate was derived based on the generalized Kelvin model. Finally, three optimization algorithms, including the traditional least squares method (LSM), particle swarm optimization algorithm with a compression factor (CFPSO), and particle swarm optimization algorithm with a compression factor based on the simulated annealing algorithm (SA-CFPSO), were used to identify the creep parameters of the generalized Kelvin model. The results confirmed that the generalized Kelvin model based on the flexible double pillow pressure plate outperforms the traditional empirical models, and the performance of the proposed SA-CFPSO algorithm is superior to that of the LSM and CFPSO algorithms in identifying the creep parameters of the generalized Kelvin model.