Geotechnical characterization of a river dyke by surface waves

The need for effective and reliable methods to survey and monitor the structure of earth-fill dams recently became pressing in light of the increasing number of flood events in central Europe. Among geophysical techniques, dam imaging using electrical resistivity methods is applied in most cases. Occasionally, ground-penetrating radar is applied in the framework of the search for subsurface facilities. Seismic methods are rarely used. This paper focuses on the multichannel analysis of the surface waves (MASW) method to determine dynamic soil properties and aims to extend its application field to dyke and dam structures. The standard processing procedure of the MASW assumes a flat free surface of infinite extension. The flat surfaces of a dyke, in contrast, are in the order of 1-10 times smaller than the wavelengths in the soil; disturbing side reflections will occur. Even though MASW has already been applied on a few dyke sites, the effect of such an obvious breach of preconditions needs to be studied before the method can be recommended. In this paper the influences of the dyke's topography on the test results are studied by means of a numerical analysis. Typical cross-sections are modelled using 2.5D finite and boundary elements. The results of models taking the topography into account are compared with models neglecting the topography. The differences are evaluated on the level of the dispersion curves and for one cross-section on the level of the S-wave velocity. They were found to be insignificant for dykes with a width-to-height ratio larger than four. A testing campaign was conducted providing the chance to collect experience in the practical use of the MASW method on dykes. Test results obtained at two test sites are selected and compared to the results of borehole logs and cone penetration tests. A remarkable relation between the S-wave velocity and the consistency of the clay sealing was found at one site; a distinct positive correlation to the measured cone tip resistances was achieved on the other test site. Valuable information on the composition of the dyke body and base could be obtained but the resolution of the method to identify small areas of inhomogeneity should not be overestimated.