Author(s): Matthew C. Halso; Frederic M. Evers; David F. Vetsch; Robert M. Boes

Linked Author(s): Matthew Halso

Keywords: Composite modeling; Physical modeling; Numerical modeling; Dam breach; Homogeneous dam

Abstract:

The overtopping of embankment dams often leads to erosion of the embankment material, and can result in eventual failure of the structure. Homogeneous embankment dams – those composed of a single material distribution throughout their entire volume – have been observed to initially experience surface erosion and/or headcutting on the downstream face, followed by expansion through sidewall failures. The erosion processes dictate the breach shape and expansion rate, and therefore also the breach discharge; thus these processes must be adequately characterized in order to accurately predict the downstream impact of a breach. Material parameters that impact the erosion processes, such as critical shear stress and failure angle, are functions of material characteristics, such as sediment size and heterogeneity. To evaluate the impact of these material characteristics on the dam breach process, we perform composite modeling of spatial dam breaching due to overtopping. We conduct physical experiments in a 1-meter wide and 11.9-meter long recirculating flume. We construct 0.3-meter tall sand and gravel embankments, each with a 0.1-meter deep pilot breach located along a transparent flume side wall. The transparent side wall represents the centerline of a “symmetrical” breach, allowing for observation and recording of the breach while representing only half of the embankment in the experiment. We induce a dam breach by raising the upstream water level above the base elevation of the pilot breach. The upstream water level, inflow discharge, and seepage rate are measured continuously, which allows for calculation of breach discharge. We monitor the breach development with a photogrammetric measurement system, from which we generate 3D representations of the entire embankment at multiple points in time during the experiment. We also implement a weighing system downstream of the embankment for continuous measurement of the mass of eroded material. The photogrammetric measurement system and the weighing system allow for redundant monitoring of the erosion rate, from which we compare results and evaluate the two systems. We additionally perform numerical modeling of dam breaching with the hydro- and morphodynamic software BASEMENT. BASEMENT solves the 2D shallow water equations for surface flow and the 3D Richard equations for subsurface flow. For morphologic calculations, BASEMENT utilizes empirical relations for bedload and suspended load transport, a geometrical failure angle approach, and incorporation of apparent cohesion effects. In the numerical model, we model the same flume and embankment geometry as in the physical experiments, with the same material types and hydraulic boundary conditions. We present a comparison of the physical and numerical modeling results – with emphasis on breach development, breach discharge, and erosion rate – and asses the viability of the extension of this test series through composite modeling.

DOI: https://doi.org/10.3850/IAHR-39WC252171192022990

Year: 2022