Direct Displacement Inverse Analysis And Application For Mountainous Tunnel

In the recent years, the expressways are constructing widely to connect the central and the western parts and forming a road grip in our country. As well known, there have many mountains and hills in the western part and the environmental conditions are very inclement. In the most case, it is necessary to construct tunnels for facilitating the traffic. However, it is difficult to pass these mountains and hills for their complex geological conditions and many difficulties will be met in the tunnel constructing. Being one of the representatives of modern timbering theory, the New Austrian Tunneling Method (NATM) makes rigorous demands that the involved designers must have comprehensive experience and ability of keen judgments that not all designers can claim. Meanwhile, it is not easy to confirm the physical and mechanical parameters of the surrounding rock and the initial field stresses in the tunnel construction, which cast a primary barrier for engineers and designers to utilize numerical simulation to get a quantitative realization of the tunnel construction. Combining information of the construction and measurement of Yunling Tunnel—the second long tunnel in the Shiman Expressway, the rock displacements are inspected in the process of construction. The direct displacement inverse analysis on the physical and mechanical parameters of the surrounding rock are proceeding and applied in the tunnel construction and collapse treatment.First of all, on the basis of introduction of the traditional inverse displacement methods, the direct displacement inverse analysis theory and manner based on the back-propagation (BP) neural network are introduced. This is different from the traditional direct displacement inverse analysis. The direct displacement inverse analysis based on the BP neural network applies the numerical methods and couples the neural network. The displacement inverse analysis can be readily conducted with the well-trained neural network. Then, taking FLAC program as a tool of forward process in the simulation of the tunnel excavating and combined with the program of BP neural networks, an inverse analysis on the unsteady surrounding rock displacements is carried out. The physical and mechanical parameters of the surrounding rock have been determined. The direct analysis is conducted with the application of the inverse analysis results in the FLAC program. It is shown that the calculated and measured displacement results agree with each other very well. Therefore, the inverse analysis which combines with BP neural networks and the FLAC program is proved to be feasible.Lastly, the process and reasons of the second collapse in Yunling Tunnel have been analyzed in detail. At the same time, the inverse analysis has used in the calculating of the surrounding unstable rock pressure. The analyzed results are compared with the design calculation by the standard. It is shown the present analysis is correct and accurate.In the study some new results on the displacement inverse analysis for the mountainous tunnels have been obtained. The analysis not only has applied in the present practical problems of the tunnel construction, but also provided a series of helpful references to the similar subjects. The analysis can be used to construct the mountainous tunnels more safely, reliably and economically.