Abstract

A study has been undertaken at musicale in the laboratory to simulate the densification of a sandy soil bed by dynamic compaction. This was achieved by allowing a tamper mass of 875 g to fall through 100 mm onto a stiff aluminium target having a mass of 268 g and a diameter of 100 mm. Since the frequencies generated by this process are very low, the dynamic stiffness derived approximately the static value derived from a monotonic load test (static load test). In total, 9 different experiments were carried out; three impact tests and six static load tests. Consequently, a number of numerical models were constructed to simulate impact load and static load tests. Overall, soil stiffness was obtained using six different methods. Analysis of the results demonstrates that the soil stiffness obtained in the impact test by Fast Fourier Transform analysis (experimental and numerical) and the stiffness obtained through static load test (experimental and numerical) all agreed reasonably well. The conclusion derived from this study confirms the reliability of numerical methods for obtaining sandy soil stiffness, which are significantly more economical as compared to experimental methods. Furthermore, this study can now be extended through development of a numerical model test in flight (multi- g) in the centrifuge to validate (static) soil stiffness computed by the WAK test.

Key words: WAK test, numerical methods, soil stiffness, centrifuge modelling, dynamic compaction, soil improvement, linear systems, impact load testing.