Stability Analysis And Nonlinear Prediction Of Granite Residual Soil Slopes In Chu Chiang Delta

Granites is distributed widely in Pear River Delta. Because joints are well developed in Granites, the weathering could develop in-depth within the rock along the joints, and then form the thick red weathering crust. The upper part of the red weathering crust of granite is a granite residual soil (GRS) layer. From the soil interface, the composition of such media is sandy soil or clay soil, but there are weak structural planes or discontinuity planes that show anisotropic properties, these structural planes can be native sedimentary formation, succession of structural interface in weathering rock, can also be the result of epigenetic transformation; their performance appear the through weak structural planes form the macro side, revealing the structure interface or esoteric structure interface. These structural planes of plays a significant role in control of the slope stability, so that slope failure was different from the arc slope damage modes of homogeneous soil.Large number of engineering practice shows that the slope of this kind has its particularity. Through the survey of GRS slopes in Pearl River Delta, found that most slopes became instability in the case of smaller height or more slowly than calculate results with experiment parameters; only a few of the slope failure were similar to the arc damage often happened in soil slopes, and most are flat slide, in the form of scattered damage. Without taking the particular type of geological properties slope into account, which often caused the designers and construction staff the illusion. Also, unclear understanding to the nature of the slope may also lead unscientific project cost estimate and construction management; create obstacles to the project follow-up. Therefore, the study of the GRS slope special nature and its classification have important theoretical and practical significanceConducted survey of GRS in Pearl River Delta, collected a large number of engineering examples, comprehensive study had done from the residual soil slope components, engineering characteristics, mechanical properties and other aspects; comprehensively classify GRS slopes. Based on comprehensive classification, according to the main deformation form and corresponding major mechanical mechanism of GRS slopes established the geological models of slopes; select the appropriate physical and mechanical parameters. For the GRS slope stability, conduct physical simulation experiments and numerical studies, take use of physical simulation and numerical calculation method to study the slope deformation and failure mechanisms, stability, and other aspects of the state. Considering the many factors that affect slope stability and GRS special circumstances, make use of rough set theory, take advantage of rough sets to establish an index evaluation system of GRS slopes, and analyzed sensitivity of various indicators to the slope stability, then using the improved support vector machine (SVM) method to evaluate specific slope stability. The results show that based on rough sets-support vector machines for nonlinear slope stability evaluation method is feasible and results can be provided scientific basis for the prevention and treatment of the slopes.Through in-deep and systematic research to GRS slope in Guangzhou area, obtain the following main results and conclusions:1 GRS in Pearl River Delta, from the particle composition has characteristic of more coarse and fine particles and less multi-particle content, so GRS has the engineering properties of both sand and clay. As to shear strength, GRS possess a large cohesion c and internal friction angleφ, but this is false, tend to the engineering and technical personnel make mistakes on the intensity value. Structural plane (joints, fractures) exist in GRS. the shear strength these structures are much lower than GRS itself, and the distribution of these structural planes control composition and development of GRS slip interface.2 GRS slopes and excavated slope instability is basically a top-down progressive sliding, the sliding interface is composed by straight lines along the structural interface and circular interface in soil and most of sliding interface are the line at ends and the arc in the middle. When slopes are instable along the GRS and bedrock interface or in slope and the residual laminated layer, the interface is linear.3 Triggering factor of artificial excavation slope instability is natural rainfall, so the waterproof is the most crucial point in slope treatment. After the slope excavation, appropriate waterproof measures must be taken.4 The stress, strain, displacement characteristics and state of stability of the landslide could be simultaneously calculated using the finite difference numerical method, in addition to considering the nonlinear elastoplastic constitutive relation of geotechnical. get more accurate stress and displacement field, to dynamic simulate the process of landslide slip and interface shape; when solve stability factor, do not assume slip interface shape, do not assume that the interaction between compartmentalization, the result is more realistic.5 According to the physical simulation experiment results, the displacement of GRS slope gradually develop from the top to the foot, sliding interface of the slope is formed gradually along the direction from the top to the foot, finally the sliding interface was through.6 After the analysis of slope stability factors, adopt rough set theory to conduct attribute reduction operation to the factors affecting the slope stability. The results showed that:after attribute reduction, six factors (cohesion, internal friction angle, slope angle, the development degree of structural interface, slope structure, and the scouring action of rain) could comprehensively response to the main factors of GRS slope, which provided the optimum evaluation factors for the subsequent nonlinear prediction.7 Aimed at the problem of the prodigious influence of the SVM kernel function parameters to model classification ability, genetic algorithm is adopted to search the best parameters of SVM kernel function. The result shows that this method not only decreases the work done, but also increases the confidence level of the parameters.8 The examination method to training model used to the method of the individual sample examination, which is unilateral greatly. Aimed at this problem, the holistic examination method is provided, that is adopting the change tendency of single factor to slope stability to examine the SVM prediction model. This method can examine the extension ability of the SVM prediction model and avoid the limitation of the individual sample examination method.9 Comparative studies showed that the prediction results of slope stability by SVM and numerical calculation are similar and accordant with the fact. It is validated that the nonlinear theories and numerical calculation applied to granite residual soil slope are feasible and reasonable.