Risk Assessment of Void behind the Lining Based on Numerical Analysis and ANN

摘要

Void behind the tunnel lining is one of the causes that lead to potential risks in tunnel operation. In many cases, voids behind lining (VBL) have been detected but the risks of these voids are difficult to estimate. This paper presents an evaluation method for assessment and management of risk associated with the VBL. Twelve key factors are selected for the void effect simulation, and totally 850 models are computed by using Fast Lagrangian Analysis of Continua in 3 Dimensions (FLAC3D). Based on the computation results, an Artificial Neural Network (ANN) has been established to simulate the void effect on tunnel lining FOS. Thus, the required data of tunnel lining in the risk assessment method can be conveniently acquired by inputting the key data of a specific tunnel into ANN instead of time-consuming computing. Finally, an operation tunnel which was detected with radar was selected as a case study, and the VBL risk accessed with the presented method is suitable for the facts.