It is generally accepted that, in civil engineering construction projects, the largest element of financial and technical risk generally lies beneath the ground. Indeed, structural foundation failure, construction over-runs and delays can often be attributed to inadequate and/or inappropriate site investigations. Unfortunately, geotechnical engineers have, at their disposal, limited guidance when scoping the extent and nature of site investigations. Often, the scope of geotechnical investigations is not governed by what is needed to characterise appropriately the subsurface conditions but, rather, is driven by budgetary constraints. A pressing need is to arm geotechnical engineers with guidelines that link the scope of a site investigation to ground variability and the probability that the foundation will be under-designed, resulting in some form of failure, or over-designed, resulting in the foundation being larger and more costly than needed. This paper outlines research undertaken to develop such guidance, focusing on the design of pile foundations in variable ground using the probabilistic techniques of random field theory, Monte Carlo simulation and genetic algorithms (GAs). The GA analyses showed that, when the number of boreholes is less than or equal to the number of piles, the boreholes are best located coincident with the piles. A single borehole should be placed at the building's centre-most pile. With two boreholes, they are best located at the sides of the building, and three boreholes should be placed to form an equilateral triangle, as much as possible, while still being near the piles.

In increasingly congested cities, tunnels are being used for new transportation routes. It is important to be able to predict the ground movement resulting from tunnel excavation, in order to manage its impact on existing infrastructure. This paper describes a procedure for calibrating settlement trough parameters automatically, using the Limaniv method, in combination with Leapfrog models and Python scripting. This method predicts surface settlement, due to elastic squeezing of an excavated tunnel, due to the weight above that tunnel. This process allows parameters to vary continuously along the tunnel, providing a more accurate settlement prediction, and removing the need to apply overall highly conservative values. A weighting function, derived and presented in this paper, is used to obtain a single equivalent Young's modulus value from multi-layered ground conditions. The code is implemented as a plugin to the Geographical Information Systems program, QGIS, and an example is given with a Leapfrog model of a location in Sydney, Australia. Settlement predictions from this approach are validated against numerical modeling.

Insufficient or inappropriate soil testing can lead to a range of undesirable consequences, however, there is little research available on-site investigation performance in complex soils. This study investigates site investigation scope in terms of a single borehole and its location relative to the foundation. The results are given in the form of heatmaps showing favourable sampling locations, whereby the optimal location can be found. The method used is statistical in nature, employing Monte Carlo analysis with randomly generated, variable, single layer soils. These soils allow both site investigations and true foundation performance to be conducted, with the resulting statistics analysed. Several site investigation, structural configuration, and soil variability factors are examined, including test type, borehole depth, reduction method, number of piles, building size and investigation performance metric. The results show that investigation quality is maximised by drilling the borehole in proximity to any pile in the foundation, and that failure costs can vary with location by up to 8% of the construction cost.

Insufficient or inappropriate soil testing can lead to a range of undesirable consequences, and yet there is no guideline for optimal investigation. This study analyses the influence of a two layer, virtual soil profile with an undulating boundary on site investigation performance. Factors investigated include the method of representing the boundary within the soil model, the stiffness ratio of the two layers, choice of test type, the pile length relative to the boundary length, and the number of boreholes and piles. The relative error contribution from the uncertainty sources of layer geology and soil variability is also examined. Investigation performance is assessed through Monte Carlo analysis in terms of total expected project cost, while implicitly incorporating the risk of damage from poor investigation. It has been shown that the optimal investigation can save in the order of AUD$1.5 million and that 2D soil models can represent 3D soils.

Insufficient or inappropriate soil testing can lead to a range of undesirable consequences, however there is little research available on site investigation performance in complex soils. The present study analyses the relationship between investigation quality and various soil conditions, such as the number of layers and the presence of lenses. Investigation performance is assessed through the use of random, virtual soils in a Monte Carlo analysis context. The assessment metric is total expected project cost, which implicitly incorporates the risk of damage from poor investigation. It is shown that the optimal investigation can save in the order of up to AUD$ 2 million (1.1 pound million), for a 6-storey, 400 m(2) building supported by 9 piles, or 30% of its construction cost. The optimal number of boreholes was found to vary with the lens stiffness ratio, lens thickness, and the magnitude of variability of both the layer boundaries and soil properties.

Insufficient or inappropriate soil testing can lead to a range of undesirable consequences, and yet there is no guideline for optimal investigation. This study analyses the influence of test type, number of boreholes, data interpretation, soil conditions, and structural configuration on site investigation performance. In addition to providing general recommendations, the relative sensitivity of these variables on performance is determined. Performance is assessed in terms of total expected project cost while implicitly incorporating the risk of damage from poor investigation. The framework for this study involves the use of randomly generated, variable, single layer virtual soils in a Monte Carlo analysis. It was found that optimal investigations can produce net savings in the order of several hundreds of thousands of Australian dollars, and key features of a future site investigation guideline are identified.

Geotechnical site investigations are an essential prerequisite for reliable foundation designs. However, there is relatively little quantitative guidance for planning optimal investigations, including the choice of testing location. This study uses a genetic algorithm to find the ideal testing locations of various numbers of boreholes with respect to pile foundation performance. The optimization has been done separately for single-layer and multi-layer soils, which infer what is best for obtaining soil material properties and delineating layer boundaries, respectively. A sensitivity analysis was conducted to find the genetic algorithm parameters that result in high quality solutions within a reasonable timeframe. While boreholes arranged in a regular grid pattern provide good performance in many cases, there are instances where optimized locations provide a cost saving of A$2 million, or 4.2% of the construction cost. A set of recommended testing guidelines is provided.

This paper presents a framework for generating multi-layer, unconditional soil profiles with complex stratigraphy, which simulates the effects of natural erosion and sedimentation processes. The stratigraphy can have varying degrees of randomness and can include features such as lenses, as well as sloped and undulating layers. The method generates the soil comprising the layers using local average subdivision (LAS), and a random noise component that is added to the layer boundaries. The layers are created by generating coordinates of key points in the simulated ground profile, which are then interpolated with a customised, 2D, linear interpolation algorithm. The resulting simulations facilitate more accurate probabilistic modelling of geotechnical engineering systems because they provide more realistic geologies, such as those usually encountered in the ground. Fortran code implementing this framework is included as supplementary material.

When designing foundations, geotechnical site investigations are usually undertaken to characterize the physical and engineering properties of the ground. Often, the scope of the site investigation is dictated by budgetary constraints and construction timelines, rather than on the variability of the ground. Limited investigations have the potential to significantly detriment a project by way of cost over-runs, construction delays, foundation failure and overdesign. This paper examines the influence of site investigation scope on the design and performance of pile foundations with respect to pile load capacity and settlement. This is achieved by carrying out a series of 2D numerical simulations of multi-layered and variable soil profiles incorporating complex geological features within a Monte Carlo framework. In this way, the probabilities of design failure and pile over-design are expressed for a range of site conditions and investigation campaigns via relative design error. Results clearly indicate that the probability of an optimal design increases significantly as the scope of an investigation is increased.