Abstract：This paper applies a stochastic finite element method by combining neural networks and Monte-Carlo simulation to calculate the reliability index of the lining structures of a highway unequal double-arch tunnel . The highway unequal double-arch tunnel consists of a big tunnel and a small tunnel. The compressive strength capacity of lining structures is regarded as the ultimate limit states, and the crack-resistance capacity is the serviceability limit states. According to reliability analysis of ultimate limit state, the reliability index of the big tunnel is smaller than that of the small tunnel. The results show that the side linings of the big tunnel are most probably the first support segment which lost supporting capacity. At the same time, the vault of the big tunnel lining is prone to be cracked by tensile stress.
5．Kohno S., Ang A.H.S and Tang W.H. Reliability-based design of tunnel support of jointed rock. Structural Safety & Reliability[C]. Edited by A.H.S.Ang et al,Proc. ICOSSAR’89,1990,San Francisco,USA:245-254
6．Kohno S., Ang A.H.S and Tang W.H. Reliability evaluation of idealized tunnel system[J]. Structural Safety,1992,11:81-93
8．Deng J., Yue Z.Q., Tham L. G., and Zhu H.H. Pillar Design by Combining Finite Element Methods, Neural Networks and Reliability: A Case Study of the Feng Huangshan Copper Mine, China [J].International Journal of Rock Mechanics and Mining Sciences,2003,40(4) (in press)
2003, Vol. 22 
