Failure analysis of scouring at pile groups exposed to steady-state flow: On the assessment of reliability-based probabilistic methodology

Abstract

This research aims to present a probabilistic model to assess scour depth around the pile group. In this way, the limit state function was back on the regression-based equation govern by the steady-state flow variables, motion states of bed sediments, and geometric properties of pile groups. Reliable experimental datasets, used in developing the probabilistic model, have intrinsic uncertainties due to randomness in influential factors. Firstly, the influential factors extracted from experimental observations have been grouped into random and deterministic variables. To understand the variation of the random variables in terms of quantity, the probabilistic distribution has been provided. Hence, a probabilistic model has been implemented by means of the Model Tree (MT) technique to estimate the local scour depth. In the next step of the current study, the proposed equation given by MT has been applied to the probabilistic framework through Monte Carlo sampling analysis in order to evaluate the influence of different variables on the exceedance probability of scour depth. Results indicated that regions with a low probability for the exceedance of scour depth from the desired value are proposed for the design purpose. As well, methods were proposed to reduce the exceedance probability of scour depth around marine structures' piles.