Overtopping induced failure of noncohesive, homogeneous fluvial dikes

Abstract

Accurate predictions of breach characteristics are necessary to reliably estimate the outflow hydrograph and the resulting inundation close to fluvial dikes. Laboratory experiments on the breaching of fluvial sand dikes were performed, considering a flow parallel to the dike axis. The breach was triggered by overtopping the dike crest. A detailed monitoring of the transient evolution of the breach geometry was conducted, providing key insights into the gradual and complex processes involved in fluvial dike failure. The breach develops in two phases: (1) the breach becomes gradually wider and deeper eroding on the downstream side along the main channel and (2) breach widening controlled by side slope failures, continuing in the downstream direction only. Increasing the inflow discharge in the main channel, the breach formation time decreases significantly and the erosion occurs preferentially on the downstream side. The downstream boundary condition has a strong influence on the breach geometry and the resulting outflow hydrograph.

New Hypothesis for the Final Equilibrium Stage of a River Levee Breach due to Overflow

“…The breach progressed mainly longitudinally after the breach channel reached the riverbed (stage 4 in Figure ). Unlike the symmetrical development observed for the dam breach case, the levee breach lengthened asymmetrically with a prevalent downstream component while the upstream section remained almost stable, as also observed by Rifai et al (). The ratio l br,dw of downstream breach length L br,dw at the equilibrium to total breach length L br,tot ranged between 67% and 91%, and larger values of the final breach length were generally achieved for larger inflow discharge.…”

“…The quasi steady variations of the control variables were larger for breach length L br and downstream water level Y US1 than for breach flow discharge Q br ; this may be related to the following two key mechanisms: A longer breach tends to increase the breach flow (positive feedback) which, however, is partially counterbalanced by the decrease of the downstream water level that affects the breach flow itself under subcritical flow conditions (negative feedback). The effective breach channel, which conveys most of the breach flow, may represent only a portion of the total breach length. This is because the breach is eroded mainly downstream, whereas the upstream section is relatively stable, and some sand layers may also deposit at the upstream toe and obstruct the breach flow (as also observed by Rifai et al, ). Moreover, the breach flow distribution increases downstream due to the increase of transversal flow velocity and flow depth.…”

“…The breach discharge increased according to smaller gradients for larger breach lengths during the successive stages; it reached an equilibrium value that depends on the specific boundary conditions, as discussed by Michelazzo et al () for the case of zero‐height side weirs. Moreover, the final breach flow is strongly affected by the downstream boundary conditions and a larger breach flow is expected if a smaller downstream Froude number F d is imposed, as in the case of a dead‐end channel (Rifai et al, ).…”

Floodplain Backwater Effect on Overtopping Induced Fluvial Dike Failure

“…The breach sides collapse gradually. The asymmetrical breach shape reflects the significant influence of flow momentum parallel to the dike crest (Rifai et al, ). In the last stage (Stage 2—widening, green profiles in Figure ), the channel free surface decreases and the flow depth starts stabilizing at its minimal level (approaching the main channel critical flow depth). The breach development becomes slower, the upstream part of the breach stops evolving, and breach deepening becomes moderate while tending to stabilize.…”

“… Note . Tests 7 to 20 were analyzed by Rifai et al () and are used here for comparison purpose. Underlined test numbers refer to tests for which breach stabilization (or quasi‐stabilization) was observed.…”