Search

The validation of design assumptions, as construction works are being carried out, is a vital component in geotechnical engineering. The feedback can be in terms of either the prescribed quality control tests, instrumentation and monitoring, or other site observations. As most of these can be evaluated through systematic logic, the use of data science methods or machine learning procedures is rarely necessary. There are, however, exceptions especially for cases in which: (i) there are several possible causes to the problems which are hard to pinpoint precisely, (ii) the quantum of data is overwhelming, and (iii) there is scatter in the of observed outcomes. Where these features are encountered, it is generally more efficient to process the data using a computer. This paper presents a possible way of interpreting the feedback obtained through observations in construction, via Bayesian programming, which is one of the many methods in machine learning. A case history discussing the performance of ground anchors in a deep excavation project is discussed. [ABSTRACT FROM AUTHOR]

In the construction of stacked parallel twin tunnels, the lower tunnel is normally constructed first before the upper tunnel to minimise the impact to the first tunnel due to the construction of the second tunnel. This paper examines the special case in which the upper tunnel was constructed first and undermined subsequently in a parallel configuration during the construction of the lower tunnel using shield tunnelling in residual soil. The longitudinal settlement profile of the upper tunnel due to the undermining by the lower tunnel was studied analytically using Winkler beam solutions, and the results were compared with field measurements through a case history. As the twin tunnels transitioned from a stacked configuration into a skewed configuration, the settlement of the upper tunnel was studied using several common solutions in engineering practice, such as cavity expansion, Gaussian settlement solutions and finite element analyses. The discrepancies between the closed-form solutions and field measurements were found to be largely due to the influence of the lateral earth pressure coefficient, K0<1. According to the finite element analyses, the in-situ stresses with K0<1 had a competing deformation mechanism with the settlements induced by tunnelling volume losses. Comparison was carried out against the conventional tunnelling sequence in which the lower tunnel was constructed first, the results of which revealed deformation magnitudes more than five times smaller, underscoring the need for detailed analyses for tunnelling sequences in which an existing tunnel is undermined. Equations to estimate the subsurface Gaussian trough width parameter from a pair of extensometer readings were derived and presented in the Appendix. Nonetheless, the extensometer measurements were localised and may not be representative of the entire geological formation. [ABSTRACT FROM AUTHOR]

The contribution of this paper is the discussion of three case histories of tunnelling past critical structures. The first case history is on a 15 storey tower building seated on a raft foundation. Two modelling approaches were adopted to model the tunnel using the 2-D finite element software PLAXIS, namely the contraction method and the internal pressure method. The analyses provide insight with regard to the need to adapt the TBM operating performance, as the presence of the building is found to be capable of inducing larger volume loss. The second case history is on a flyover bridge, of which the pile toes are at an elevation higher than the tunnel crown. In the third case history, the piles of a Light Rail Transit (LRT) bridge were in the way of the tunnel. The piles had to be removed and underpinned with new piles which terminate below the tunnel invert. For these two case histories, we demonstrate the use of the load transfer t-z and Q-z method (Seed & Reese, 1967), which can be implemented easily into a spreadsheet, to estimate the pile settlements induced by tunnelling. The load transfer on the piles were found to be affected by (i) the soil settlement trend, which could be either increasing or decreasing with depth depending on the tunnel position relative to the pile (both vertical and horizontal distance between them), and (ii) the shaft and end-bearing stiffness. The potential influences of (i) a soft base for bored piles and (ii) pile groups are discussed. Finally, insights obtained from the t-z and Q-z analyses are used to derive the influence zones due to tunnelling, and they show similar trends to the influence zones previously proposed by Jacobsz et al. (2004) derived from centrifuge tests. The line joining the points of inflection of multiple subsurface Gaussian settlement profiles (Mair et al., 1993) at different depths was found to define the boundary where the trend of soil settlement changes from an increasing to decreasing trend. [ABSTRACT FROM AUTHOR]