Realistic risk assessment of axially loaded pile–soil system using a hybrid reliability method

Abstract

The behaviour of an axially loaded pile embedded in a non-homogeneous soil deposit considering the pile–soil nonlinear interaction effect is realistically evaluated, and a new efficient and accurate hybrid reliability method is proposed for estimating the corresponding risk. The complicated problem is first deterministically formulated and then the uncertainties associated with the various design variables are considered explicitly. The hybrid reliability approach intelligently integrates the concepts of the response surface method, the finite difference method, the first-order reliability method, and an iterative linear interpolation scheme. The soil around the pile is represented by a series of springs. The system is expected to represent a realistic and efficient load-transfer mechanism. The soil–pile dual system is then deterministically analysed using the finite difference method. Uncertainties associated with load conditions, material and sectional properties of the pile and non-homogeneous soil properties are then incorporated in to the deterministic formulation resulting in the hybrid reliability approach. The risks corresponding to both service ability and strength limit states are estimated. With the help of an illustrative example, the applicability, accuracy and efficiency of the proposed algorithm in the safety assessment of axially loaded pile–soil systems are demonstrated. The reliability evaluation procedure is verified using the Monte Carlo simulation technique.