Piled Rafts: Are We Modelling Them Correctly?

It is argued that approximate computer based methods are only suited to modelling conventional piled rafts, for pile enhanced rafts where the interactions will be non linear three dimensional finite element analysis is required. However current 3D finite element analysis often relies on symmetry to reduce computational cost and there are concerns over the dependency of the results on the user. Therefore a new strategy within the finite element framework has been proposed. The main advantage of this strategy is that each stage of the model can be checked for consistency. There is also the added benefit that non symmetrical models can easily be constructed by tying the validated components together without dramatic increases in computational time. The proposed strategy compared well to the published results of a hypothetical example of a piled raft previously analysed with numerous methods by Poulos et al. (1997).