| Summary: | There is a challenge for deep excavation works in existing build areas to cater for urban growth and sprawl, with the alternative to build upwards. These soil excavations with limited construction space could affect the performance of adjacent piles, and foundation failures have been reported since then. In modelling soil-pile interaction, it is important to accurately predict or quantify the pile responses of such incidents to ensure the stability of structures. The reliability of existing methods that are two-staged and those that tend to neglect the nonlinear characteristics of soil-pile interaction, is still being discussed. Currently, researchers are increasingly in favour of three-dimensional (3D) finite element modelling (FEM) during the design process to better understand ground mechanisms and soil-structure interactions.
Subsequently, a back-analysis modelling procedure can also be applied in design and post-construction or failure investigation. The FEM software, PLAXIS 3D (version 2013) is used for back-analysis and to verify the behaviour of single pile in three different conditions: (1) A laterally loaded single pile from a published case study; (2) A single pile behind a retaining wall subjected to excavation from a published centrifuge experiment; (3) A 1g small-scale experimental model analysis based on the second condition but with axial load and different pile head fixity. These comparative and experimental studies could justify the usability and reliability of the software.
The same software is then used for a parametric study to model various parameters such as diameter of pile (D), distance of pile to wall (X), axial load (AL), relative density of sand and pile head fixity with respect to excavation. The thematically analysed data relates increasing axial load on a smaller diameter pile to significant effect relative to deflection rather than bending moment, and vice versa behaviour for a bigger diameter pile. On another aspect, different pile head fixities indicate distinctive effects relative to bending moment.
Finally, a set of design charts are developed based on the results of the parametric study. The main outcome or contribution to the knowledge base is on the practice of piles behind un-strutted wall under excavation-induced soil movement. Lastly, this research could simplify preliminary design application in situation where the understanding of the pile behaviour on close-proximity excavation is required.
|
|---|