Experimental and numerical investigations of similarity for dam-break flows on wet bed

The similarity brings great convenience to the description of the dam-break flows. The previous works mainly focus on the similarity of dam-break flows under the condition of dry bed downstream. In view of the fact that there is often a certain tail water depth in a river, this paper deals with both the gravity similarity and self-similarity of dam-break flows in a horizontal and smooth channel with a wet-bed downstream condition by using both physical and numerical models. In the laboratory tests, the digital image processing technique is adopted for flow measurement, providing water surface profiles, stage hydrographs and cross-sectional mean velocity. Three different upstream depths and ten water depth ratios a (i.e., the ratio of initial downstream water depth against the upstream water depth, ranging from 0.05 to 0.9) are considered. The numerical simulations on the vertical two-dimensional experiments are carried out using a computational fluid dynamics solver, providing velocity profiles in addition to the hydraulic parameters obtained in experiments. The following observations are made: (1) the water surface profiles, stage hydrographs, mean velocity distribution and horizontal velocity profiles are shown to satisfy the gravity similarity well; (2) the gravity similarity characteristic of the vertical velocity profiles in the reservoir performs better than that in the flooded area; (3) the water surface profiles, mean velocity distribution and horizontal velocity profiles approximately satisfy the self-similarity after a period of time for alpha < 0.3; (4) the undulations both in the rarefaction wave and the shock front destroy the self-similarity of dam-break flow for alpha >= 0.3, except that the self-similarity is valid in two non-undulation areas (located upstream the extra negative wave and in the neighborhood of dam site) after a period of time; (5) the gravity similarity could be applied in the small-scaled physical model of dam-break flow and the water depth ratio must be considered when applying the self-similarity; (6) the measurements and numerical simulations with the large eddy simulation (LES) model show satisfactory agreement, implying that the LES modeling is a viable approach for accurately predicting the dam-break flows with a wet bed downstream condition.