摘要
研究目的:作用效应的概率统计特征分析是隧道可靠性分析的重要内容之一,本文运用Monte-carlo随机有限元法对铁路隧道复合式衬砌进行计算,从抽样次数、分布特征及各参数对结果的敏感性方面展开分析,试图给出定性结论。研究结论:(1)在用Monte-carlo随机有限元法对隧道支护结构进行作用效应概率统计特征分析时,可以选择抽样1万次对随机变量进行模拟;(2)复合衬砌作用效应的变异系数和荷载的变异系数基本相当;(3)通过直方图和假设检验可以得到作用效应接受服从正态分布;(4)通过分析基本随机变量对作用效应影响程度可知,等效塌方高度是影响初期支护和二次衬砌轴力的最主要的因素,其次是材料的弹性模量和结构的厚度,而对偏心距影响最大的因素是弹性反力系数,在今后的工作中应该加强对围岩压力和弹性反力系数的研究;(5)该研究结果可为《铁路隧道设计规范》的修订提供理论依据。
Research purposes： The probability and statistics feature analysis of action effects is one important part of the tunnel reliability analysis. This paper calculated the composite lining of railway tunnel by using Monte - carlo stochastic finite element method to try to give qualitative conclusions, with analyzing the aspects on the sampling frequency, distribution characteristics and the sensitivity of various parameters to the result. Research conclusions： （ 1 ） In probability action effects analysis of the supporting structure of the tunnel using Monte - carlo statistical characteristics of stochastic finite element method, it can select a random sample of 10,000 times to simulate variable; （2） The coefficient of variation of composite lining effect and coefficient of variation of load are roughly equal; （3）Action effects can be obtained by histogram accepted normal distribution and hypothesis testing; （4）By analyzing the basic random variable degree of influence on the action effects, the equivalent collapse is the most important factor affecting the height of primary support and secondary lining axial force, followed by the thickness of the structure and the elastic modulus of the material, and the biggest impact of eccentricity is the elastic reaction coefficient, in future work should strengthen the study of rock pressure and elastic reaction cofficient; （5） The calculation results can provide theoretical basis for the railway tunnel design code revision.
出处
《铁道工程学报》
EI
北大核心
2015年第6期69-75,共7页
Journal of Railway Engineering Society
基金
铁道部科技研究开发项目(2012G014-D)
河北省高等学校科学研究项目(ZH2012037)
关键词
复合式衬砌
作用效应
概率统计特征
参数敏感性
composite lining
action effects
probability and statistics features
parameter sensitivity
