| Finite element method is practical to evaluate slope stability, using the strength reduction.Currently weather the displacement converges or not acts as the criterion when the finiteelement method is utilized to analysis slope stability. While the safety actor obtained isdifferent from the traditional stability factor definition. Besides the slip surface cannot beaccurately available, nor the magnitude and distribution of the stress, strain, and deformationon the surface. Though the finite element method outweighs limit equilibrium method, suchdetects bring great difficulty to slope design and supporting, resulting in its limitedapplication to slope stability evaluation now days.This paper aims to solve the problem. Based on the Smith&Griffiths slope elastoplasticfinite element program, the program of searching critical slip surface in slope was developed,so the stability factor’s calculation. The main work and results of this paper are the following:1) Based on the stress field and strain filed obtained from the finite element calculation,there are two different ways to search the potential failure surface, one based on the slip linein stress field and the other based on the maximum shear strain. once the slip surface isdetermined, the ratio of shear strength integration along X and shear stress integration along Xis just the stability factor. Relative programs are developed and planted in the Smith&Griffiths (S&G)slope elastoplastic finite element program, unifying the finite elementcalculation, slip surface searching and stability factor calculation.2) Take the slope for example in the survey of slope stability analysis programs, initiatedby the ACADS in1987. The slope is analyzed by two finite element methods: S&G methodand the finite element method in the software of MARDAS/GTS, which also employs thestrength reduction. Results show that the average stability factor is0.983when using the S&Gmethod of quadrilateral element with8joints, while it is1.348with4joints and1.123with8joints when using the MARDIS/GTS software. According to the results of ACADS, it is0.991. actually, the real stability factor is1.000. Evidently it is easy to find that the resultsobtained from the S&G method and limit equilibrium method are similar to real result.Stability factor calculated by MARDIS is larger12.3%than the real one. 3) Comparsion between the stress field and strain filed calculated by the two finiteelement methods demonstrates that the results differentiate from each other greatly. Bothusing the quadrilateral element with8joints, the strain in middle slope is larger170%and therelative stress over80%when calculated using MARDIS software than the S&G method.4) Stability factor increases with the dilatancy angle, whose relationship is liner whencalculated by the S&G method results. By contrast, the relationship is not clear whencalculated by MARDIS software.5) The program developed by the author searched the potential slip surface in theexample and calculated its relative stability factor. Results show that the prospective slipsurface based on the stress field is similar to the one obtained by limit equilibrium method,and the stability factor is1.000, also close to limit equilibrium method’s result. While the slipsurface based on the strain field is deeper than that above, and the relative stability factor is1.082,bigger than the limit equilibrium method’s result.6) Analysis of the effect of the grid size (S), step length (L) on slope stability factor andthe location of potential slip surface show that the location of the slip surface has littleconnection with S and L. while stability factor has much connection with them. When L/S iswithin0.8to1.2, the factor remains relatively stable... |
PDF Full Text Request