A test set-up was developed with the innovative use of soaked foam cushions to simulate the consolidation behaviour of soft soils. One experiment with a piled sand embankment was carried out successfully with the test set-up, and the development of arching was observed and measured. It showed that arching is most efficient for limited deformation of the sand bed (before shearing occurs). Calculations using Rogbeck/Carlsson, (extended with incomplete arching) and with McKelvey underestimate the load on the reinforcement somewhat, which may result in unsafe designs. The Extended Carlsson calculation rule with the shear angle taken from the test gives best agreement with the test results. It is recommended to design piled embankments with the Rogbeck/Carlsson method, extended by using incomplete arching as described in this paper. The shear angle gives the shape of the soil wedge that determines directly the design-load on the reinforcement. It is recommended that this shear angle should be determined experimentally in a test set-up as described in this paper. The British Standard BS 8006 gives inconsistent results for increasing embankment thickness. The Carlsson and Rogbeck method also gives inconsistent results for incomplete-arching-situations due to a limited embankment thickness. The research will be extended in the future by carrying out more experiments with, for example, 3-dimensional circumstances and the application of different kinds of reinforcement.
展开▼