Figure 1. Installation of Sand Compaction Files

The sand compaction pile is increasingly becoming a common soil improvement technique in reclamation works in Singapore. It has been employed in the Marina Bay/Tanjong Rhu reclamation, and is being implemented in the Singapore-Malaysia second link reclamation project in Tuas as well as the Port of Singapore Authority_s third cargo terminal reclamation works off Pasir Panjang. As shown in Fig. 1, in the field, sand compaction piles are formed by driving a casing into the soft ground to a predetermined depth. Sand is then forced through the casing into the ground to displace the soft soil and thereby form a sand pile. Thus soil does not need to be dredged out of the seabed, thereby avoiding the problem of disposal of the dredged material.

Much of the existing research in sand compaction has so far been conducted in Japan. This research shows that, for replacement ratios of about 40%, the installation of sand piles generally results in enhancement of the stability and bearing capacity of the ground as well as reduction in ground settlement. However, there is still insufficient knowledge to define how settlement of the improved ground can be assessed, especially for high replacement ratios of 70% or more. In addition, the effect of the density of compaction of the sand on the short- and long-term performance of the sand pile is also largely unknown.

In recognition of the increasing use of sand compaction piles in Singapore and the need to know more about this technique to facilitate cost-effective and competent design, the Housing & Development Board (HDB) and the NUS have embarked on a collaborative project to study the performance of sand compaction piles. The project commenced in April 1994 and is presently on-going.

The scope of the research involves numerical and experimental research. On the numerical side, computer analyses will be implemented to model the sand piles and to conduct parametric studies on the effect of the depth of the soft clay layer, extent and orientation of loading, density of compaction and replacement ratio on the ground movement. This phase of the research is currently in progress.

On an experimental level, centrifuge model parametric studies on sand compaction piles are now currently being conducted to investigate the effect of depth of the soft clay layer, density of compaction and replacement ratio on the stability of the improved ground and ground settlement. Two methods of modelling the installation of the sand piles have been developed. The first method follows the conventional approach used by Japanese researchers and involves first creating a frozen sand pile in a mylar tube and then inserting the pile into the soft clay bed by a replacement process at 1g. The second method attempts to model the sand pile installation process by forcing sand through a miniature casing using an Archimedes_ screw at high g. The ability of these two methods to replicate the process in the field will first be compared. Parametric studies will then be conducted using the selected installation process.

Although the project has commenced less than a year ago, preliminary results indicate that there is still much scope for improving and refining present design methodology. One of the findings relates to the parameter which is often used to quantify the stress transfer to the sand pile. In existing design methodology, the parameter which is often used is the ratio of the stress in the sand pile to that in the intervening clay. Numerical modelling suggests that, under conditions of high replacement ratio, this parameter may be erratic. Under such conditions, a better parameter may be the ratio of the stress in the sand pile to the average stress on the improved ground. The studies also show that at high replacement ratios, the behaviour of the improved ground is less sensitive to the density of compaction than at low replacement ratios.

Contact Person: Dr FH Lee
Tel: 7722274,
Fax : 7791635,