Reliability analysis of a steel fibre reinforced concrete tunnel lining

At the moment, there are no Dutch regulations giving criteria for the design of bored tunnels. These regulations are necessary to prove that the required safety and serviceability are guaranteed by the proposed design. The scope of the study presented in this report consists of a reliability analysis of a SFRC tunnel lining. This consists of the definition of the load and resistance factors that are considered necessary to make a design of a SFRC lining and the calculation of the failure probability of the tunnel in the ultimate limit state. The study carried out in this report is based on the ITM tunnel design case with a lining thickness of 0,40 m and a radius of 4,70 m. The tunnel lining can fail by several mechanisms. The mechanism that is examined in this study, consists of the failure of the tunnel lining under strain of the soil stresses in the ultimate limit state. The analysis is executed with a level II probabilistic design method. The basis of this method is that the parameters used in the structural design are not specified constants, but stochastic variables. The most advanced method of this level is the so-called Approximate Full Distribution Approach (AFDA). This method is used for the reliability analysis executed in this report. First of all, the reliability of the tunnel design is approached with a linear calculation model, i.e. with the analytical Duddeck model. The most remarkable conclusion drawn from the linear reliability analysis, is that the safety requirements are on the whole better met in case of more flexible and thinner inings. As this conclusion does not meet the expectations and as the use of a (simple) linear model has a real benefit, the impact of non-linear effects on the cross-section forces is examined in order to determine whether the assumption of linearity is likely to be acceptable within the limitations of the non-linear tunnel analysis. An important conclusion drawn from the non-linear reliability analysis is that the safety decreases with decreasing lining thickness. The latter refutes the conclusion drawn at the linear reliability analysis that the safety increases with decreasing lining thickness. Moreover, from the non-linear reliability analysis follows that it is not safe to use the (linear) analytical Duddeck method for the design of slender constructions. In general, it can be concluded that the examination of the safety of a tunnel lining is very extensive and that still a lot of work is left to be done.Hydraulic EngineeringCivil Engineering and Geoscience