Polynomial chaos expansion for uncertainty quantification of dam engineering problems

Authors

M. A. Hariri-Ardebili, B. Sudret

Abstract

Uncertainty quantification is an inseparable part of risk assessment in dam engineering. Many probabilistic methods have been developed to deal with random nature of the input parameters or the system itself. In this paper, the polynomial chaos expansion (PCE) is adopted as an effective technique for uncertainty quantification of a variety of dam engineering problems (especially with small data sets). Four different case studies are investigated with increasing complexities in which the static and dynamic responses are sought to predict. Implicit and explicit limit state functions are studied. Uncertainties are propagated in material properties and modelling. Depending on the problem at hand, a validation set of few hundred to several thousand points are used. Overall, it if is found that the PCE is an effective technique to deal with uncertainty quantification in concrete dams.