Influence of site investigations on the design of pad footings

Abstract

The reliability of foundations is greatly influenced by uncertainties associated with the geotechnical model on which their design is based. In turn, the geotechnical model is derived from a site investigation whose scope is largely dictated by financial constraints, rather than the variability of the ground. This paper seeks to quantify the influence of the scope of site investigations on the design of pad footings. This is achieved by simulating geotechnical profiles, where the soil properties vary from location to location in a random yet continuous and realistic fashion. The simulated soil profiles are generated using random field theory, which makes use of three statistical parameters: the mean, variance and the scale of fluctuation – which is a measure of the randomness of the geotechnical property in question. The methodology involves generating a geotechnical profile by simulating a 3D random field to know soil properties at each point in detail. A site investigation is then simulated by sampling from the 3D random field. By using the sampled values, a pad footing is designed to conform to a serviceability criterion, as would be achieved in practice. A benchmark design is also undertaken making use of the complete knowledge of the soil profile. This design is achieved using a numerical process involving a 3D finite element analysis. Both designs are undertaken on numerous different simulated soil profiles in a Monte Carlo analysis. A comparison of the two designs provides failure and over-design probabilities for a series of site investigation plans. It was observed that the probability of failure and overdesign decreased with an increasing site investigation scope, as expected. The results provide information to estimate the relative benefit of conducting various scopes of site investigations.