Energy piles are foundation piles combined with closed-loop ground source heat pump systems. The objective is to support superstructure, as well as to serve as heat transfer medium, which can release heat into the ground in summer and absorb heat from the ground in winter. Cyclic temperature change in energy piles can cause serviceability problems. Because the axial and radial deformation of energy piles due to heating and cooling cycles result in different shearing behaviour compared to non-energy pile at pile-soil interface. Besides, soil experiences volume changes under thermal cycles. Centrifuge modelling is a powerful tool to study those problems. However, there is still knowledge gap in understanding the scaling effects on the centrifuge modelling of energy pile. Particle...[ Read more ]

Energy piles are foundation piles combined with closed-loop ground source heat pump systems. The objective is to support superstructure, as well as to serve as heat transfer medium, which can release heat into the ground in summer and absorb heat from the ground in winter. Cyclic temperature change in energy piles can cause serviceability problems. Because the axial and radial deformation of energy piles due to heating and cooling cycles result in different shearing behaviour compared to non-energy pile at pile-soil interface. Besides, soil experiences volume changes under thermal cycles. Centrifuge modelling is a powerful tool to study those problems. However, there is still knowledge gap in understanding the scaling effects on the centrifuge modelling of energy pile. Particle size effects is the most common aspects of scaling effects in centrifuge modelling of piles in sand. It is because all dimensions of prototype are scaled down to the model corresponding g-level, however soil particle size is not normally scaled. Shaft resistance of piles is dependent on the soil lateral stress at interface shear zone. The increase in lateral stress is a function of change in the thickness of shear zone at pile-soil interface which is a function of soil particle size. Therefore, shaft resistance can be affected by the particle size effects. Consequently, this issue can make discrepancies in results of different scaled models of one prototype. This research aims to fill the knowledge gap by conducting centrifuge modelling of energy piles embedded in saturated sand. Aspects such as energy pile and soil temperature history, pile head settlement, distribution of pile axial forces, and pile shaft resistance and base resistance are considered.

The technique of “modelling of models” is adopted to evaluate the scaling effects on energy piles modelling in centrifuge. The research methodology of centrifuge modelling is employed for this research. Two different scales corresponding to 28g and 56g levels were used to simulate the same prototype of a single energy pile in saturated Toyoura sand, which has a diameter (D) and an embedded length (L) of 0.88 m and 16.8 m, respectively. Additionally, the results of a similar test at 40g are adopted to conduct a more comprehensive study. All energy piles are subjected to the same constant working load of 600kN (factor of safety of 2.5) and seven successive heating-cooling cycles with amplitude of ± 10 ⁰C. Seven pairs of foil gauges and three pairs of thermocouples were installed along each pile shaft to measure axial load distributions and temperature profiles. Settlement of energy piles and ground soil are measured by Linear Variable Differential Transformers (LVDT) fixed at pile head and soil surface.

From centrifuge modelling, it was found that temperature of the vicinity soil surrounding energy piles in all tests at three different g-levels are almost consistent. The measured data are in a good agreement with the theoretical prediction by using the infinite cylindrical heat source theory. The measured thermally induced settlement for three energy piles at 28g, 40g and 56g are almost similar from 1.25% to 1.31% pile diameter. The same ratcheting pattern with a reducing rate can be noticed obviously from settlement results during thermal cycles. After applying seven thermal cycles, the distribution of axial loads along the pile for two energy piles at different g-levels does not have significant difference. Finally, the overall consistence in measured results of key features involved in energy pile problems imply that scaling effects on centrifuge modelling of energy piles in sand are negligible.