Nisal Deelaka Halaba Arachchige Senarathna ^1* Théo Dezert ² Fjola Gudrun Sigtryggsdottir ¹

• ¹ Norwegian University of Science and Technology, Trondheim, Norway
• ² FI-NDT, Gustave-Eiffel University, Bouguenais, France

The growing requirement imposed by dam safety regulations and guidelines necessitates the improvement and rehabilitation of rockfill embankment dams. These hydraulic structures are of great importance, and they can be subjected to overtopping flows which can significantly compromise the structural integrity. One of the defense mechanisms utilized is the placement of riprap on the downstream shoulder of the dam. This article explores eight experimental tests comprising of four different dam model configurations and investigates the possibility of Particle Image Velocimetry (PIV) techniques to explore the characteristics of the protective riprap layer, such as breach initiation, failure mechanism, and velocity pattern. The models varied from full to half dam profiles, constituted of placed or dumped riprap, with or without downstream toe support, and with or without downstream shell material. Leveraging the PIV technique, the study provides insight into the area of breach initiation within riprap protection on the downstream shoulder of rockfill embankment dams and thus breach initiation of protected rockfill dams. The study brings to light that for models with placed riprap, the initiation occurs simultaneously at the top and the bottom of the protective layer confirming the assertions made in prior studies regarding a failure mechanism marked by a sliding process.The study further confirms that for structures with toe support, the breach initiation occurs at the top of the protective layer. This analysis also revealed that in the case of dumped riprap, the precise position of the breach initiation was indeterminate. Furthermore, the analysis revealed that there is a significant drop in the velocity readings at the downstream section of the riprap layer when supported by a toe, underscoring the significance of having toe support. Another revelation from this analysis was the contrasting velocity readings with substantially higher readings for placed riprap as compared to dumped riprap reaffirming the fact that placed riprap can endure higher discharges as compared to dumped riprap. Furthermore, this investigation also detected that a graph of the velocity pattern depicting the placed riprap exhibited a higher singular peak while such a graph depicting the dumped riprap exhibited a bimodal characteristic.1.