A method for the study of suffusion by means of visualisation experiments is presented. Suffusion is the removal of the finest grains from a widely graded soil due to seepage flow. The mobilisation and transport of grains is strongly coupled to the structure of the soil. Images are acquired at suitable spatial and temporal scales that allow for the coupled study of particle movement as well as the change in structure of the sample. The observations are made for a gap-graded plane strain sample behind a transparent window. Image subtraction is used for analysis of the results, i.e. the identification of areas of the sample where particle movement occurs, the evolution of the locations where particles are removed and their corresponding transportation paths and the evolution of particle motion over the height of the sample. The presented method therefore employs the same dataset to study particle motion and sample structure in space and time, thereby providing a means to address the coupling between them.
Backward erosion piping poses a severe threat for embankments founded on sandy deposits. The physical mechanism by which a pipe progresses upstream beneath the embankment is conceptually understood, and assessments indicate high failure probabilities in the Netherlands. Rising sea levels and economic growth increase the risk of flooding, and the need for remediation measures. Conventional measures have undesired side effects, in terms of costs, maintenance, space requirements and sustainability. The coarse sand barrier (CSB) is a novel nature-based solution lacking these side effects. The CSB is a trench filled with coarse sand that is placed below the downstream toe of the embankment, which prevents the pipe from progressing upstream. The effectiveness of the CSB is based on its higher resistance against erosion and its effect on groundwater flow. This paper presents experimental work that was conducted to test the feasibility of a CSB for application at a pilot site in the Netherlands. This involved experiments with seepage lengths of 0·34, 1·34 and 15 m, using different barrier materials and background sands. A conceptual model is presented for the physical mechanism of pipe formation in the presence of a barrier. As the resistance of the barrier can be characterised by a local scale-independent strength criterion, laboratory results are applicable for design practice.