Author(s): D. Duma; S. Erpicum; P. Archambeau; M. Pirotton; B. Dewals
Linked Author(s): Sébastien Erpicum, Benjamin J. Dewals
Keywords: No keywords
Abstract: Predicting the entrainment of bed material by water flow is of utmost practical relevance, particularly for the safe design of hydraulic structures and to ensure riverbank stability. The inception of sediment motion was studied extensively for a broad range of bed material properties and bed slopes; but most of these previous studies were based on the assumption of uniform flow conditions. In particular, many of them used a friction formula for evaluating the bed shear stress. Recently, Hoan et al. (2011) and a few other researchers explored new avenues to assess the stability of riverbed material. They linked directly the sediment pick-up rate to the local flow and turbulence characteristics. Such approaches avoid the need for closure relations valid only under uniform flow conditions, such as most friction formula. However they were validated only for a limited range of flow conditions and geometric setups (e.g., for sudden expansions). In the present research, we aim to investigate whether such a new approach also holds in the case of a different geometric setup, namely for the case of smooth-to-rough transitions (Fig. 1). This configuration is very often observed downstream of hydraulic structures, at the transition between the concrete lined bottom and the natural riverbed. We performed flume experiments considering two different configurations. We used a uniform bed material to define a reference configuration (C1) and we compared to a sudden transition from smooth to rough bottom (C2, Fig. 1). Two different flumes were used to assess possible scale effects and to extend the range of tested flow conditions. The bed material (diameter d=8,15 and 30 mm, relative density=0.5, 0.7 and 1.65), the flume slope and the flow rate were varied. Using a Vectrino II ADVP (manufactured by Nortek) and a UVP probe (manufactured by METFLOW), flow velocity was measured at a high frequency (100 Hz) along different profiles in the nearfield of the smooth-to-rough transition. The sediment pick-up rate was evaluated by counting the number of displaced particles. Measurements of flow velocity and turbulent fluctuations were used to compute the non-dimensional stability parameters u-σ (u) introduced by Hoan et al. (2011). The sediment pick-up rate was expressed in non-dimensional form through the bed mobility parameter E, also defined by Hoan et al. (2011). In Fig. 2a (respect. Fig. 2b), we display the observed bed mobility parameter as a function of the standard Shields parameter S1b (respect. the bed stability parameter u-σ (u) of Hoan et al., 2011). Despite some scatter in the experimental results, they suggest a more significant correlation between the mobility parameter Eand the new bed stability parameter u-σ (u); rather than between Eand Shields parameter S1b, which was evaluated assuming uniform flow conditions. This study confirms the potential of this new approach based on local flow and turbulence characteristics. Additional experiments remain necessary to parametrize the relationships between the bed mobility parameter Eand bed stability parameters, which were tested so far only in a narrow range of setups.