Risk Analysis For Tunnelling Based On FEM And BBN

Recently a large number of tunnel projects are being constructed in many Chinese cities. High uncertainties and risks involved in tunneling jeopardize the economic development, human health and social stability, especially at the constraints of time and cost. So it is meaningful to construct a risk analysis model for tunneling for the purpose of the prevention of disasters.In this thesis, BBN(Bayesian Belief Network) is applied in risk analysis model for tunnelling using BBN software Netica?. Based on two important mechanisms in soil mechanics, arching effect and mobilised strength of soil, a rule for tunnel stability is built up. This risk analysis model is also applied in some interesting fields , such as probability inference, most probable explanation and sensitivity analysis. Calculated results indicate this model is fit for real project. In order to get cases for parameter studies in BBN, 14 FE examples of tunnelling problems are calculated.In order to build a rational BBN model for risk analysis, tunneling benchmarking problems are simulated using ABAQUS software.Tunnel deforming can be destructed into three basic deforming modes, including buoyancy, ground loss and ovalization. Calculated results are compared to the analytical solutions which identify the both Poisson ratio of elastic soils and relative bury depth of tunnels as the most important influence parameters of influencing ground settlement trough. It is important for settlement assessment to simulate ground loss actually.The following study of heading stability of unlined tunneling in non-cohesive soils shows that how limit supporting pressures and associated arching effects are calculated using FEA simulations and their agreements with limit analysis solutions. Parameter study reveals that the relative bury depth of tunnels and the critical angle of friction of Mohr-Coulomb type soils are critical in the heading stability problem.