Many models have been developed to solve geotechnical problems, but their v
alidation is only possible by a parallel experimental analysis on real work
s. In this sense, a well controlled experiment was carried out on a shallow
foundation resting on a dry coarse Leighton Buzzard sand. A flexible shear
stack was filled with a uniform relative density Leighton Buzzard sand. Th
e foundation was subjected to a vertical eccentric load and to a sine dwell
-type acceleration transmitted from a shaking table over which the shear st
ack was located. The dynamic response of both the soil and the foundation w
as checked by means of a complex network of accelerometers and displacement
transducers. To have a detailed description of the behaviour of the footin
g under the seismic motion, four coloured layers of sand were introduced in
to the deposit at different depths. By measuring the displacements of the c
oloured layers, carefully excavated after the test, it was possible to inve
stigate the experimental failure mechanism. It was then compared with the t
heoretical failure mechanism and the seismic bearing capacity was evaluated
. (C) 2001 Elsevier Science Ltd. All rights reserved.