By W J Larnach and N A Al-Shawaf, University of Bristol
This paper was first published in GE’s September 1972 issue.
The notion of using longitudinal vibrations to ensure the penetration of pre-formed piles into the ground is not new, and has found practical application in a number of forms. In general, driving has been achieved by the rigid attachment to the head of the pile of a vibrator, which in its simplest form provides a pair of contrarotating out-of-balance masses. When the masses are phased together and driven at the same speed, horizontal components of the rotating forces cancel, whilst the vertical components augment one another.
The vertical components are cyclic, that is, they are directed first upwards and then downwards and thus downward movement of the pile can be encouraged by the use of additional so-called bias surcharge weights. These are seismically mounted on the vibrator, so providing a continuous downwards force. Multi-system vibrators have been suggested’n which more than one pair of whirling masses are employed. By proper choice of weights and speeds the ratio of peak (downward force/ upward force) can be increased much above unity.
Several commercial vibrators have been designed to operate within the range of about 10-50 Hz,’ither at fixed speeds or at a number of speeds, achieved by often lengthy mechanical adjustment. The socalled resonant vibrator operates at significantly higher frequencies, of the order of 150 Hz, close to the natural longitudinal frequency of the pile. More recently a vibrator has been developed” for which it is claimed that the frequency can be continuously adjusted up to some 120 Hz, with amplitude control and adjustment.
The industry has at its disposal a useful technique and existing commercial machines achieve some success. But comparatively little attempt has been made to measure and understand the effects of the several machine and soil variables involved; and little work has been undertaken to discover in detail what happens to the soil during driving, either at the pile/ soil interface or in the mass. A full discussion of previous work is presented in reference 2, from which it is clear that definitive answers are still sought on why and how the vibratory driving process works; what effect variables such as frequency, amplitude, surcharge, might have on driving in different soils.
The first of these questions is the more difficult to answer, and the work described here is a first attempt at trying to answer the second, at least at the model scale. The hazards of quantitative extrapolation of model behaviour to full scale are well known, but qualitatively useful trends can emerge. Limitations in both space and time determined the scale at which these tests could be conducted, having decided that the foundation material was to be dry sand.