广义Kelvin蠕变损伤模型及其参数的智能辨识  

Generalized Kelvin Creep Damage Model and Its Parameters' Intelligent Identification

作　　者：徐国文[1] 何川[1] 代聪[1] 王士民[1] 

机构地区：[1]西南交通大学交通隧道工程教育部重点实验室,四川成都610031

出　　处：《西南交通大学学报》2015年第5期866-871,共6页Journal of Southwest Jiaotong University

基　　金：国家863计划资助项目(2014AA110401);煤炭联合基金重点项目(U1361210);国家科技支撑计划资助项目(2013BAB10B04)

摘　　要：为了更好地描述岩石的流变变形特征,提出了考虑岩石蠕变过程中材料劣化效应的黏弹塑性模型,并导出了该模型的有限差分格式.基于FLAC3D软件的接口平台,采用C++语言实现了模型的二次开发.在此基础上,采用粒子群算法、模拟退火算法与FLAC3D相结合的智能方法,用已有试验数据进行了参数反演.结果表明:该模型能反映岩石的黏弹塑性以及损伤特性,能模拟低应力下岩石的两阶段蠕变和高应力下岩石的三阶段蠕变变形效应;只要反演算法参数取值适当,算法在40代左右就可以收敛到全局最优解.In order to describe the characteristics of rock rheological deformation better,a generalized Kelvin creep damage model considering the deterioration effects in creep process was proposed,and its finite difference scheme was derived. Based on the interface platform of software FLAC3 D,the model was developed using C ＋ ＋ language. Furthermore,an intelligent algorithm combining PSO（ particle swarm optimization） algorithm,SA（ simulated annealing） algorithm and FLAC3 D was put forward and adopted to determine the rheological parameters in the model based on the existing experimental data.The research results show that the proposed model can reflect visco-elastic-plastic properties and damage characteristics of rock,and simulate the two-stage creep effect of rock in a low stress condition and three-stage creep effect of rock in a high stress condition. The algorithm can converge to the global optimal solution in some 40 steps as long as inversion parameters are selected appropriately.

关 键 词：蠕变损伤模型 智能反演 粒子群算法 模拟退火算法 

分 类 号：TU411[建筑科学—岩土工程]