Discussion “Experimental study of velocity fields in rectangular shallow reservoirs”, Reply by the authors

The authors present the results of a new experimental study to improve the flow classification in rectangular shallow reservoirs. Although previous papers have paved the way for the classification (Dewals et al. 2008, Dufresne et al. 2010), the authors are complemented for the detailed flow-field measurements conducted for a high number of reservoir geometries, complementing the previous observations based on Large Scale Particle Image Velocimetry (Dewals et al. 2008). The studied rectangular reservoirs have a length L and a width B, and the width of the inlet and outlet channels is b. The authors refer to a non-dimensional shape parameter T = L/(B − b)0.6/b0.4, which is actually a multiple of the shape parameter L/[(B − b)/2]0.6/b0.4 = 20.6T ≈ 1.516 T as defined by Dufresne et al. (2010). For the tested hydraulic conditions, the authors report that for T < 4.09, the flow remains symmetric, while for T > 4.48, the flow is asymmetric, with a transition zone between the two limits, corresponding to “unstable” flow pattern. Repeated tests under similar conditions lead alternately to a symmetric or an asymmetric flow field. It is intended below to shed light on this transition zone using 2D numerical simulations. The authors state that the flow in the transition zone is sensitive to so-called “external perturbations”, whereas we argue that the flow is particularly influenced by the initial test conditions. Using a particular post-processing of the computed flow fields, a hysteresis effect is detected.