J. Wilson

Doctor of Philosophy, Keble College, University of Oxford, Trinity Term 1970

Forces Exerted on Solid Bodies Moving Through Cohesionless Soils

Summary

The aim of this thesis was to investigate the various parameters which affect the drag force on totally immersed solid bodies moving horizontally through dry sand. The body, or model, shapes were geometrically simple, being cone-fronted cylinders and the parameters varied were the radius, cone angle, cylinder length, surface roughness, depth and speed. Further tests were conducted while the sand was subjected to an upward flow of air, and in order to extend the range of observations, compressed air was realized through the walls of a cylindrical porous model.

Most parameters were examined in loosely packed sand, and the horizontal drag force on the models was found to vary linearly with depth, model projected area, rough-surfaced cylinder length and sand bulk density. In dense sand the drag force was found to increase linearly with speeds up to 8ft/sec. In loose sand the drag force decreased as speeds rose to 3.6 ft/sec, a phenomenon attributed to compaction in front of the model. As speeds increased to 8 ft/sec dynamic forces were recorded which increased with the square of the speed. The component of drag due to side wall friction was negligible compared to the model frontal resistance and the most important factor governing the magnitude of the drag force was the void ratio of the sand.

The models were driven through a cylindrical tank of sand by a hydraulically powered ram. The uniform loose sand state was created by fluidising the bed of sand between tests. Good repeatability of results was achieved by this method of bed preparation.

Further studies involving a soil pressure cell investigated the failure pattern created in front of a model.

Two-dimensional model shapes were driven slowly through a transparent soil box, the dimensions of which were related as far as possible to those of the tank. The mechanism of failure in front of the model occurred successively on a series of curved slip planes which extended to the sand surface. A body of sand was carried before each model and the sand particles forming the body changed continuously. Very little disturbance was caused below the model and although the angle between the failure plane and the horizontal was approximately equal to the passive angle of (45° - phi/2) in the model vicinity, this angle increased away from the model.

(no thesis available)