Stochastic analysis of the capacity of two parallel footings on a thin sand layer

This paper presents a study on the stochastic bearing capacity of two parallel strip footings supported on a heterogeneous sand layer confined by a rigid base. Variability of the sand is represented by a truncated normal distribution function of its friction angle to produce a geostatistical equivalent to the deterministic bearing capacity. A method was first demonstrated to derive an analytical distribution function of the bearing capacity for a single surface footing with a rigid base layer for which an exact bearing capacity function is available. For the twin foundation problem with no explicit deterministic solutions, a Monte Carlo simulation was adopted. The stochastic bearing capacity relative to its deterministic equivalent was determined for a set of layer thicknesses and footing distances in terms of the assumed geostatistical characteristics of the sand and resulting bearing capacity distributions. The Monte Carlo methodology is adaptable in design to model other bearing capacity problems with more complex and random soil distributions.