A test programme was carried out in the small geotechnical centrifuge of the University of Delft to investigate the vertical bearing capacity of suction piles in sand and clay. The influence of several parameters was tested, such as height/diameter ratio, cyclic load, long term load and loading speed. Further tests were carried out on an alternative anchoring system, with a view to improving the vertical bearing capacity. The results of these tests were compared with those of the API standard and finite element calculations. The centrifuge test also made it possible to investigate the individual contributions of inside and outside soil friction, and to visualize mechanisms. The small size of the samples facilitated accurate reproduction of soil density, so that slight differences in test parameters could be made visible. There was found to be good agreement regarding static pullout loads between FEM calculations, a simple analytical model and static tests in sand and clay. It appeared that below 80% of the static pullout capacity cyclic and long term loading did not lead to failure. The effect of dynamic vertical loading tests could be visualized well in clay, but less clearly in sand. The alternative anchoring system showed an increase of the vertical bearing capacity under some circumstances. However, further improvements will have to be made before the system can be applied in the field.
In recent years, suction piles have been applied increasingly often in offshore engineering (Wang et al., 1978: Senpere et al., 1982). Suction piles are attractive because of the convenient method of installation. A pile with a diameter of 9m and a height of 10m can be installed in I-3 hour, by using only a pump. Despite the wide use of suction piles, details of the influence of several parameters have remained unknown
