Model reconstruction method of irregular columnar jointed rock masses and corresponding elastic parameter analysis

A local Lloyd iteration-based Voronoi tessellation (LLI-VT) algorithm is proposed to reconstruct models of columnar jointed rock masses (CJRMs) according to their geometrical characteristics. Uniaxial compression numerical analysis was performed on specimens to analyze the effects of specimen size, joint density, irregularity, and joint dip angle on the elastic modulus and Poisson's ratio of the CJRM. The results show that the LLIVT algorithm generates models with good control over column size and distortion, which align more closely with natural CJRMs. For research on elastic behavior, an RVE size of at least 2 m is sufficient. As the joint density increases, the elastic modulus decreases at a decreasing rate, while Poisson's ratio shows the opposite linear trend. With increasing irregularity, the elastic modulus increases at a decreasing rate, and Poisson's ratios exhibit opposite nonlinear patterns. The joint dip angle is the most important factor affecting elastic parameters. The effects of joint density and irregularity on elastic parameters are different for different joint dip angles. Finally, a method for estimating the elastic parameters of CJRMs is discussed and proven effective. This approach is important for accurately simulating and predicting the mechanical behavior of CJRMs in engineering.