Code Implement And Application On Stability Anaysis Of Tunnel Surrounding Rock Mass In Porous Medium

The constitutive model of the soil mass determines the computation results greatly in the numerical Geomechanics. Based on the former researches, this dissertation presents a constitutive relations coupling with the solid-fluid-air for the porous medium, and designs a new integration arithmetic for elasto-plastic deformation, and then a finite element program is coded by FORTRAN. Utilizing a porous argillaceous shale tunnel as the research project, the deformation mechanism of the surrounding wall mass in porous tunnel is studied systematically. The main contents are as follows:(1) The governing three-phase mathematic and mechanics model about the deformation and strain localization is derived in porous medium from nonlinear continuum mechanics and mixture theory. According from mass balance equations, momentum balance equations and energy balance equations, the solid-fluid-air model is set up. These constitutive laws relate: the evolution of the constitutive effective stressσ~* with imposed solid matrix deformations; the degree of saturation s_r with suction stress s; the intrinsic mass densities with intrinsic pressure on all three phase; and the relative flow vector with intrinsic pressure p_αfor the water and air phase. Also, the fully Lagrangian form of the Darcy law is resolved by Piola algorithm and then the flow law is gained, leading to the implementing of a modified pore water pressure in the constitutive model.(2) An improved stress integration algorithm is proposed based on the nonlinear elasto-plastic constitutive laws, and then a new iterative algorithm is designed here. The dissertation introduces the Newton-PCG (Preconditioned Conjugate Gradient) composite algorithm to solve the iterative algorithm in large-scale problems with huge stiffness matrixes, which is coded by FORTRAN program, together with an efficiency analysis compared with standard Newton's iterative method. Next the necessary condition for localization is obtained by introducing the first tangent operator, which presents a theoretical basis in deformation simulation.(3) A large-scale finite element numerical software is written based on the FEM theory, leading the code implementing of the above constitutive relation frame. This dissertation presents a systemic numerical computational program with the FORTRAN language, as well as the Q9P4 numerical simulation element. Next this program is interfaced with Matlab, which leads to the implementing of figure output.(4) The deformation behavior of the porous sand and argillaceous shale tunnel are simulated with the self-written FEM program. Numerical simulation on saturated sand under the undrained condition is performed to study the onset and development of the shear band, and the mesh sensitivity is studied. Next the deformation and the stress of the soft argillaceous shale under construction are simulated to study the deformation behavior of porous medium.(5) The dissertation presents comparative analysis between the numerical simulation and in-situ testing in a porous argillaceous shale tunnel. The inner displacement of the surrounding rock mass, the contact pressure between the surrounding rock and the primary support, the force between the steel support and the grating support, the force of the primary support and the force of the secondary support, are all studied here according to the in-situ testing data. Then we systematically discuss the deformation failure condition, deformation failure character, instability failure mechanism and failure mode, leading to the study of tunnel deformation mechanism.(6) Based on the coding, numerical simulation and in-situ testing, this dissertation develops the prediction researches on tunnel deformation. Systematic advanced prediction and process prediction researches into tunnel deformation are carried out, based on the solid mechanics theory and the grey systems theory; especially in process prediction, we adopt the data processing of filling vacancies in the proper order and establish the A GM (1,1) model to modify the grey GM (1,1) model. The prediction well matches the in-situ testing data when this method is applied in the Guankouya tunnel.