Excavation of the western ticket hall box at Tottenham Court Road station represented a critical path activity of the Crossrail railway project in London, UK. Base slab construction and preparatory sprayed concrete lining works needed to be completed before the arrival of the two tunnel boring machines. A full observational method (OM) design was implemented to eliminate the lowest level of temporary propping, resulting in a 13 m prop-free excavation at the bottom of the 30 m deep diaphragm wall station box. This paper describes the design and successful implementation of the OM, which eliminated the need to fabricate, install and remove the fifth level of temporary propping. It covers the comprehensive review of the original design input parameters, back analysis of the most probable geotechnical design parameters and comparison of wall deflection and forces in the temporary props from early stages of the construction works. The robust strategy formulated as part of the observational method design to control the subsequent excavation is described: this included trigger criteria, a review process and predetermined contingency measures to ensure safe execution of the excavation works. A genuinely collaborative effort between client, permanent and temporary works designers, independent design checker and the contractor is elaborated. Cost and programme savings achieved by implementation of the observational method approach are outlined.

Realistic finite element (FE) modelling of the tunnel construction process has been restricted by the computational effort needed to incorporate the three-dimensional (3D) aspects of the tunnelling operation. Two-dimensional (2D) plane-strain models have therefore been widely adopted by practitioners and researchers in their studies of the effects of tunnelling. However, 2D modelling suffers the drawback whereby a volume loss must be assumed at the outset as the input for the FE model. This paper presents an analysis of measured surface ground movements induced by a sprayed concrete lined tunnel in London Clay. This was achieved using computationally efficient, non-linear, 3D modelling and Oasys LS-DYNA software. By modelling the problem using a non-linear small strain BRICK soil model, which incorporates anisotropic behaviour, good agreement with the actual Gaussian distribution of surface settlement was obtained. This was achieved without any initial assumptions regarding volume loss, but rather by realistically modelling the excavation sequence and time-dependent gain in lining strength. On the basis of this work it is considered that 3D modelling techniques can provide an effective means to estimate ground settlement and the associated effects on nearby structures, without dependence on empirical methods alone.

No abstract

The increasing use of the finite-element method in geotechnical design has raised the question of the compliance of this design approach with Eurocode requirements for the ultimate limit state conditions, especially when a more complex soil constitutive model has been used. Past authors have identified several important issues relating to the application of the finite-element method in ultimate limit state design, including the effects of initial stresses, effects of stress history, choice of soil model, significance of the failure of structural member and the timing when the partial factor of safety is applied during the design assessment. In this paper, an advanced 'Brick' soil model has been used to demonstrate its application in the design of different geotechnical structures and the effects of the design assumptions used in the design of these structures. The paper also demonstrates the versatility of the Brick soil model in the derivation of the new set of input parameters when the necessary partial factor is applied to the strength of the soil as required in the Eurocode design approach.

Advances in understanding real soil behaviour, achieved over the past twenty years, and subsequent developments of relevant soil constitutive models, have greatly enhanced the predictive ability of numerical analyses of geotechnical structures. Of particular relevance, especially for serviceability limit state (SLS) design, has been the understanding of Geotechnical Engineering The constitutive modelling of London Clay Yeow and Coop