The Research On Analysis Method Of Reliability For Stability Of Tunnel Surrounding Rock And Supporting Structure

Tunnel engineering is one of the important research fields of geotechnical engineering, because of the underground environment for tunnel and historical reasons for rock and soil materials, the various factors which determine the stability of the tunnel structure have the uncertainty characteristics and also cannot be ignored. And the reliability or the corresponding non-deterministic method can effectively take into account the variability for the stratum environment and the physical and mechanical parameters of rock and soil, the probabilistic characterization indexes can help to express the failure risk of underground engineering more clearly. The stability reliability evaluation of tunnel structure is one of the important theoretical problems in the field of tunnel engineering. Reliability evaluation provides the basis for the implementation of the reliability optimization design of underground engineering, at the same time, is also a necessary link of risk prediction. Reasonable and feasible reliability calculation method is an inevitable way for the reliability evaluation of underground engineering. The accuracy and adaptability of the calculation results are not only directly related to the safety reliability of underground engineering, but also provide the basis for making reasonable project management plan, so it has a significant impact on project decisions. For this purpose, this paper is based on random character description method of tunneling uncertain parameters and the processing, the uncertainty analysis of the tunnel project is carried out from aspects of the probability, non-probability and the combination of the two, respectively.First of all, the function of the structure of the underground support structure is constructed by two ways. First, based on the principle of combined arch, displacement and deformation coordination principle, elastic thick wall cylinder theory and surrounding rock fracture model, the function of the tunneling anchor-sprayed lining structure is established. Secondly, according to the core idea of modern underground structure mechanics(rock mass is not only a part of supporting system, but also is the main body), on the basis of the Duncan-Fama’s stratum characteristics describing function, combined with the monolithic lining mechanical mechanism, through the convergence confinement principle, express surrounding rock deformation equation in support system is deduced, and the bearing allowable deformation of surrounding rock is determined according to the engineering analogy method, in accordance with the control principle of deformation instability, the structure functions of surrounding rock is established.Secondly, based on the probability theory model, in order to solve the difficulty which the conventional reliability methods can’t solve the tunneling engineering of which function has the highly nonlinear implicit characteristics, in combination with the Nataf transform which based on semi analytical formula and the eight point estimation method, the statistical moment point estimation method which is based on the statistical features of the basic variables and can be considered as a correlation is constructed for nonlinear implicit function. With the statistical moment as constraints, the function for probability density function is derived based on the maximum entropy principle, and the one dimensional direct integral evaluation method for the reliability of complex structures is established, it provides a calculation way which has the characteristics of an intuitive, practical and meet the precision for the similar calculation of stability reliability of tunnel support by tunneling anchor-sprayed lining structure.In order to further expand reliability evaluation methods of the highly nonlinear functional implicit function, the second-order second-moment algorithm which has the characteristics of high precision is modified using the numerical difference theory difference, the modified method can effectively solve the difficult problem which is the first-order second-moment algorithm and second-order second-moment algorithm need to solve the partial derivative functions. By applying this method to the functional model which is established by the above mentioned way, a direct solution method is developed for failure probability calculating of the highly nonlinear implicit function of similar tunneling surrounding rock structure. At the same time, the coupling effect analysis way for impact on the surrounding rock of tunnel structure instability probability of the lining installation determine time, reveals the variability of basic parameters and their interactions is established.Again, in view of the lack of information for uncertain parameters, within the framework of the non-probabilistic model, interval method is used to solve the problem of tunneling uncertainty problem, according to sub interval method can effectively solve the interval expansion, the implicit function iteration procedure is combined with statistical optimization model, the reasonable range of response values of functional is obtained, and researching a kind of probability method of implicit functions of tunnel engineering, and applied to the design of the actual strength of tunnel engineering.Finally, under the premise of the uncertain parameters are lacks of information and showed small feature samples, the tunnel super ellipsoid convex model is established based on the set theory. In the unit hyper sphere standard vector space, the response function value which is the initial virtual sample points into the tunnel model is obtained by using interval Latin hypercube test, and fitting tunneling surrounding rock structure function by using the Kriging surrogate model. Based on the geometric meaning of the non-probabilistic reliability index, and using probability reliability index iteration method, the non-probabilistic reliability index is solved based on the surrogate function. According to the compatibility of probabilistic reliability model and non-probabilistic reliability model, considering the problem of non-probabilistic index less than 1, the comprehensive evaluation index of reliability analysis of tunnel engineering is established.