基于岩体精细化描述的围岩分类及力学参数概率分布特征分析

Probability distribution characteristics of rock mass classification and its mechanical parameters based on fine description

针对目前水电站地下厂房工程中不同围岩分类方法存在评价结果不一致、围岩力学参数存在室内试验值与实际情况不吻合的现象,现推荐采用岩体精细化描述体系对围岩岩体结构进行定量化评价。将常用围岩分类方法(RMR、Q、RMi、GSI、BQ、HC)评价指标予以归纳分组,并通过各组内不同指标对比分析获得围岩分类方法中的基础评价指标。以大岗山水电站主厂房某区段为分析对象,采取现场岩体精细化地质素描与后期数据挖掘、拟合相结合方法,并依据评价指标间的关联关系,获得了基础、非基础评价指标的分布概型及对应参数,实现对该段围岩岩体精细化描述认知;基于精细化描述结果,应用Monte Carlo法生成符合各评价指标分布概型的大量随机数,而后参照各分类方法评价思路与评分流程,得到评价指标在各分类方法对应的大量随机评分值,通过归纳统计获得不同围岩分类方法评价结果的分布概型;基于各围岩分类方法评价结果与力学参数值之间的关联关系实现对力学参数概率特征分析。该分析方法与思路可为类似工程围岩质量及力学参数的精确确定提供一定借鉴,并可为实现围岩支护极限状态设计提供必要的原始参数支持。

According to the drawbacks that the evaluation results by different rock mass classification systems are not consistent and the rock mass mechanical parameters between indoor tests and current situation have a great gap, a method named rock mass fine description system is recommended to describe surrounding rock mass structures quantitatively. Firstly, these evaluation indexes of the common rock mass classification systems（RMR、Q、RMi、GSI、BQ、HC） are summarized as some groups, and basic indexes are selected from these groups by the in-detail comparative analysis. Then, some sections of powerhouse in Danggangshan hydropower station are chosen as the research objects;we use the method combining the in-situ rock mass fine geological sketches with post-data mining and fitting, and based on the association relationship between the ranking parameters and the basic ones, these probabilistical distribution models and their relevant parameters of evaluation indexes are able to be required; thus the surrounding rock mass elaborate descriptions of the section can be obtained. Based on the rock mass fine description, lots of random numbers that meeting the probabilistical distribution models and relevant parameters of these evaluation indexes can be produced by the Monte Carlo simulation; and then according to the evaluation idea and rating process of each rock mass classification system, plenty of corresponding random rating-values of these evaluation indexes are also gained;the surrounding rock mass evaluation results and these probabilistical distribution characteristics by six rock mass classification systems can be attained easily by the inductive statistics. At last, these probabilistical distribution characteristics and their relevant parameters of these rock mass mechanical parameters can be decided by the correlation formulas between them and the rock mass classification indexes. This research can＆amp;nbsp;provide some references to estimate accurately the rock mass classification and mechanical parameters in some similar projects, and offer the essential primary data for supporting limit state design to the surrounding rock mass.