Reliability-based factor of safety for bearing capacity of square foundations on 3D cohesive-frictionless soils considering samples

Abstract

Reliability-based design concepts have been widely adopted in structural engineering but remain less prevalent in geotechnical practice due to the challenges of characterising ground uncertainties. To facilitate the unification of structural and geotechnical designs in complex engineering systems, we focus on square shallow foundations as crucial elements. Before a complete transition to load and resistance factor design, understanding the reliability of foundations designed using traditional factors of safety is essential. This paper presents a simplified analytical method for evaluating the probability of failure of square shallow foundations on 3D cohesive soils, utilising local average theory. The approach efficiently computes local averages and covariances of soil properties without the need for time-consuming 3D finite element computations. Verified against random finite element analysis, the method accurately determines required factors of safety for bearing capacity designs. Site-specific parameters, including sample location, shear strength coefficient of variation, and spatial variability fluctuation scale, influence the derived factors of safety. We advocate a pragmatic approach that calibrates traditional factor of safety methods with rigorous reliability theories, easing the transition to reliability-based design for practicing engineers. Moreover, the determined factors of safety can aid in establishing appropriate resistance factors by considering suitable load factors.