基于围岩力学参数概率分布模型的变形敏感性灰关联分析

为了给布伦口-公格尔水电站地下洞室某标段围岩稳定性分析的参数选取提供可靠的理论依据，综合考虑岩体参数的空间变异性，针对常规敏感性分析方法所存在的不足，运用三维离散元计算程序，提出了基于围岩力学参数概率分布模型的变形敏感性灰关联分析方法。该方法以岩体密度?、弹性模量E、泊松比?、黏聚力c、内摩擦角φ及节理内摩擦角φj等6个围岩力学参数作为因素序列，拱顶下沉量作为目标序列，分析影响因素在整个定义区间内的变化对围岩拱顶下沉的敏感程度。结果表明：密度是最敏感因素，其次是弹性模量、黏聚力和泊松比，而内摩擦角以及节理内摩擦角敏感性最小。最后，将常规灰关联敏感性分析与该方法计算结果进行对比，结果表明：除密度、内摩擦角和节理内摩擦角一致外，其余参数敏感性均与结论不一致。因此，文中方法在考虑实际参数概率分布的基础上能够更加准确、合理地对参数进行综合评价。

Grey relation analysis of deformation sensitivity based on probability distribution models of surrounding rock mechanical parameters

This paper provides the reliable theoretic guidance for parameters selection of surrounding rock stability analysis in Bulungl-Gongur hydropower station underground cavern. The grey relation analysis of deformation sensitivity method is proposed based on probability distribution model of surrounding rock mechanical parameters, while the spatial variability of parameters and the shortcomings of conventional sensitivity analysis are considered. Three-dimensional discrete element calculation program is used. This method regards density, elastic modulus, Poisson’s ratio, cohesion, internal friction angle and joints internal friction angle as factor array. The vault crown settlement is specified as the target array. The influence of each factor changing in defined domain on the vault crown settlement is analyzed. The results show that the density is the most sensitive influential factor among all the factors. The subordinate sensitive influential factors are the elastic modulus, cohesion and Poisson’s ratio. The internal friction angle and joints internal friction angle are the minimum sensitive influential factors. Finally, comparison of results between conventional sensitivity analysis and the proposed method is made. The results indicate that except density, internal friction angle and joints internal friction angle are consistent, the sensitivity of the rest parameters are not consistent. Therefore, this method could evaluate those parameters more accurately and reasonably based on considering the probability distribution of rock mechanical parameters.