Font Size: a A A

Research Of Nonlinear Theories And Its Application On Failure Of Rock Mass

Posted on:2005-06-11Degree:DoctorType:Dissertation
Country:ChinaCandidate:C H XuFull Text:PDF
GTID:1102360152455725Subject:Structure engineering
Abstract/Summary:PDF Full Text Request
The stability of rock mass is one of difficult problems in rock mechanics, although many successful examples have been obtained and lots of practical experiences and theories are accumulated in the field of rock mass engineering. However along with quick development and continuous deeply study in rock mass engineering, people still discovered hermeneutic phenomena and hard-solving problems, which needs to study and solve with a new subject- nonlinear theories. After reviewing the past work, 5 aspects of researching is developed as following.1.There are many factors affecting shear strength parameters of the rock mass and some of them are qualitative. As artificial neural network can consider various factors, the neural network model can be used to select shear strength parameters. At the same time, relative strength of effect (RSE) of input parameters versus output parameters is analyzed by the trained neural network model. Considering the relative effect strength, output parameters can be adjusted according to the change of the input parameters, and consequently the optimum control of the model can be obtained.2. A displacement back analysis based on the support vector machine and simulated annealing is proposed. Firstly the support vector machine is used to substitute the time- consuming finite element analysis, and secondly simulated annealing is used for the optimization of objective function for fear that the search often falls in one of the minimums and cannot go any further when the conventional mathematical optimization methods are employed. The feasibility and practicability of this approach is verified by back analysis to the mechanical parameters of underground cave.3. The process of rock mass' evolution to instability is a course of energy dissipation with entropy decreasing gradually. Therefore, the entropy can used for describing the characteristic of nonlinear and self-organization of the process of engineering rock evolution. The constructive information entropy is defined in the text, and the criterion of entropy catastrophe of rock mass is given by entropy combined with FEM, and entropy is thought as an index of the safe degree of rock mass engineering.4. The model of predicting displacement of evolution of rock mass based on support vector machine and the catastrophe analysis is studied. The series ofdisplacement is built by the method of predicting chaotic time series based on SVM, and can be transformed to the type of cusp catastrophe equation. Then stability of rock mass can be analyzed and predicted by the eigenvalue of catastrophe equation.5. Numeric analysis method of stability of rock mass based on displacement catastrophe is studied. The change of displacement in the excavation process of rock mass is imitated by finite element method and the displacement series of key part is analyzed according to cusp catastrophe theory to distinguish stability of rock mass.By this research a relatively integrated nonlinear model will be obtained. This model can be referred in engineering designing, furthermore widens and innovates stability theory of mechanics and failure theory of rock mass.
Keywords/Search Tags:Failure of Rock Mass, Nonlinear Theories, Parameters of Shear Strength, Neural Network, Support Vector Machine, Simulated Annealing Algorithm, Criterion of Entropy Catastrophe
PDF Full Text Request
Related items