Approximate reliability-based design with general geotechnical models by stochastic simulation

Traditionally, geotechnical engineers have practised design approaches based on safety factors to account for uncertainties, owing to their simplicity. More recently, reliability-based design approaches have emerged as a more reasonable and rigorous way of handling uncertainties. If the functional relationship between reliability and safety factor can be found, it is then possible to achieve reliability-based design with the traditional safety-factor approach. However this functional relationship, in general, does not exist. This research proves a sufficient condition for the existence of such relationship. Moreover, a theorem is derived to determine the functional relationship between reliability and safety factor with stochastic simulations. The proposed framework is so general that it is applicable to arbitrarily complex geotechnical systems with arbitrarily high-dimensional uncertainties, and no lognormal or Gaussian assumption is required. Two simulated examples of reliability-based design are used to verify the proposed approach. The results show that the proposed approach is quite promising.