Identification And Application Of Creep Constitutive Model Of Rock

Based on previous research results, the identification methods of viscoelastic and viscoplastic constitutive models of rock were studied, the reliable and high performance approach were introduced by combining the hybrid particle swarm optimization (HPSO) with finite element method (FEM). Programs used to identify the constitutive models of rock were compiled in Fortran language according to the introduced approach. By analyzing practical examples, the identification procedure and application of the constitutive model were performed, and the results of identification and calculation were verified by using observed data.In order to study the relationship between creep time and the parameters of creep constitutive model of rock, the calculation program based on HPSO algorithm was compiled in FISH language. The most suitable model and parameters were proposed via calculation and comparison, which can describe the creep behavior of rock more correctly, the results indicate it is necessary and reasonable to take the time varying coefficient into account. According to the further developing platform, the inconstant creep constitutive model of rock was developed in FLAC^3D in VC++ environment. The numerical simulation of creep test under uniaxial compression was performed to verify the correctness of the compiled model. Main contributions are as follows:1. On the basis of general knowledge, principle, method and contents of system identification, the identification approach and procedure of creep constitutive model of rock were studied.2. By adding inertia weight and and Lbest factor into the iteration formula of particle velocity, the problems of low convergence and sensitivity to local convergence of PSO were solved, and then the hybrid particle swarm optimization (HPSO) algorithm was proposed.3. By using HPSO algorithm and FEM method, more effective methods to identify viscoelastic and viscoplastic constitutive model were proposed and the corresponding identification programs were compiled. The results of identification and calculation examples indicate that both of the methods are reliable and correct.4. According to the identification methods concluded in this dissertation, the application of identification results in the forecast of tunnel wall rock deformation, the security estimation of tunnel operation and the optimization of support construction were studied. A practical example was carried out to introduce the application progress and to verify the identification method and program.5. The creep constitutive model of rock, which was subjected to creep test under uniaxial compression, was identified by using the identification program compiled in FISH. By analyzing the variety characteristics of creep parameters, time-varying parameter and the relationship between time and it were confirmed, and then the inconstant creep constitutive model was brought forward. Compare was carried out between the fitness function value of constant constitutive model and that of inconstant model. The results reveal that it is necessary and reasonable to introduce time-varying parameters into creep constitutive model.6. Further development of FLAC^3D was performed in VC++ environment and the inconstant generalized Kelvin model was written into FLAC^3D software, in this way the inconstant constitutive model can be applied into practical engineering. Ultimately, the numerical simulation of creep test under uniaxial compression was performed to verify the correctness of the compiled model.