Studies On Pit Excavation Numerical Simulation And Soil Constitutive Model And Parameter Back-analysis

With rapid development of city construction and large hydraulic engineering projects springing up, deep pits increase by gradually, and the excavation depth and size become bigger and bigger. Because of better adaptability, Finite Element Method (FEM) is playing a much more important role for the design and analysis of pit engineering. The experiences of pits excavation numerical simulation indicate that the simulating results of FEM greatly depend on the soil constitutive model and the calculation parameters. However their studies have been not perfect as yet. Therefore, the FEM numerical simulation of deep pit excavation is taken as engineering background and study content in the paper, and two key techniques including the soil model and the back analysis of the calculation parameters are studied. Studying results and conclusions are listed as follows:1. In order to simulate the characteristic that loading soil continuously appears nonlinear elastic-plastic deformation from the initial stress state during pit excavation, the elastic-plastic theory of rock and soil is introduced into Duncan-Chang model, and Duncan-Chang model is improved and developed into a new elastic-plastic soil model (namely the improved Duncan-Chang model).2. The numerical calculation techniques of the improved Duncan-Chang model are thoroughly studied, and the FEM program (MDC-Excavation) of pit excavation simulation based on the improved model is developed. Through the simulation for a pit excavation and the comparison between the measured data and the calculating results, the improved Duncan-Chang model is proved to be reasonable and believable.3. The five-component visco-elastic soil model is advised to simulate the rheology of the soft pit excavation, and the visco-elastic FEM program (EV-Excavation) for pit excavation is developed. Taking the soft pit excavation in Shanghai as an example, the time effect of the retaining structure deformation is predicted. Through the comparison with the measured data, it is shown that the five-component visco-elastic soil model can better simulate the time effect of the soft pit excavation.4. In order to obtain the equivalent calculation parameters of the soft pit excavation, the simplex method is introduced into the displacement back analysis of the soft pit excavation, and the displacement back analysis program (namely Back-EV) is developed. Through case study it is indicated that the better and reasonable results can be observed by Back-EV when the initial values are reasonable.5. Facing up to the shortcomings of the simplex method such as easily losing in local extreme and serious dependence on the initial values, the random simplex method (RSM) is presented, then it is applied to the displacement back analysis of the soft pit excavation, and the correspondent displacement back analysis program(RSMB-Excavation) is developed. The numerical test and case study indicate that RSM can efficiently research for the overall optimum solution or the approximate optimum solution. Besides RSM does not depend on the initial values, so RSM is an efficient method for the parameter back analysis of the soft pit excavation.