Comparing soil erodibility predictions against the fundamental understanding of erosion

Abstract

In breaching of levees, sediment erosion is induced by high flow velocities. Due to the often continuously acceler-ating flow, no equilibrium transport conditions are reached. The erosion rate is often described by the erosion equation which linearly relates the erosion rate to the excess shear stress by means of a soil erodibility coefficient. The soil erodibility is thereby often determined by means of the JET test. In the field of Dredging Engineering the study of the behaviour of sand under high flow velocities has also been an area of interest. At Delft University, the department of Dredging Engineering has performed several experiments on the behaviour of non-cohesive material when subjected to high flow velocities. It was noted that the initial porosity, and permeability of the material are important parameters to account for in the erosion process. This corresponds with observations that soil erodibility is sensitive to variations in material texture, compaction moisture content and compaction energy. It was further-more noted that soil no longer erodes due to the pick-up of individual particles but fails as entire layers when sub-jected to high flow velocities. For shear failure to occur in non-cohesive soils the soil needs to dilate. The associat-ed increase in pore volume causes an inflow of water into the soil. Based on the fundamental mass and momentum balance equations that describe the process of dilation of a bed when subjected to shear, a new process based ero-sion equation has been derived, which when validated against erosion measurements from the dredging industry shows promising results. This paper compares the prediction of soil erodibility by means of a JET test to the out-comes of the process based erosion equation. The paper highlights important differences and attempts are made to explain these.