Numerical Test On Stability Of Surrounding Rock Mass Controlled By Faults In Tunnel Projects

It is very necessary to find a fast way to estimate the stability of the underground tunnel after excavation for nowadays project requests. Based on the discussion on some problems existing in the tunnel design and construction, a new method to build an automatic analysis platform for tunnels is proposed, which is based upon lots of numerical test results, expert experiences.and the powerful artificial neural network method. In the platform, many factors impacting on the stress and deformation fields of the surrounding rock masses can be considered, such as the size of underground tunnels, the overburden, the mechanical parameters of rock masses, the initial stress and 1 or 2 main faults nearby, etc..The main works and conclusions in this paper can be described as follow:1) Stability of surrounding rock mass is dominantly controlled by faults. In order to define the influence of fault on tunnel surrounding rock, a measure coefficient is given. And then, the coefficient is defined clearly.2) A qualitative conclusion on the measure coefficient is got base on a large number of numerical test. It indicates that the fault measure coefficient (α,β_l,β_r) is independent of depth, span, high-span ratio and shotcrete thickness of tunnels, but lies on faults distrubition, geo-stress characteristic and the rock quality.3) A large number of numerical tests are designed and practiced, which include fault measure coefficients at different fault distrubitions, different geo-stress characteristics, and different rock qualities. Quantificational conclusions on the fault measure coefficient is got base on the tests results.4) Baesd on systematic numerical test results, a table to measure the influence of the 1 or 2 faults neer the tunnel is set up. And then, with a tunnel of Ludila hydropower station project for exemple, it give a logical method to use the table to forecast the stability of surrounding rock mass. The practicability of this method is estimated.5) By means of two models parallely connecting, A parallel subareas A.I. forecasting model is built to forecast stability of underground tunnels considering the faults' position,distance, thickness and the rock porperties, overburden, tunnel size and tunnel shapes.