Circular hydraulic jumps: where does surface tension matter?
Recently, an unusual scaling law has been observed in circular hydraulic jumps and has been attributed to a supposed missing term in the local energy balance of the flow [\cite{bhagat_2018}]. In this paper, we show that - though the experimental observation is valuable and interesting - this interpretation is presumably not the good one. When transposed to the case of a axial sheet formed by two impinging liquid jets, the assumed principle leads in fact to a velocity distribution in contradiction with the present knowledge for this kind of flows. We show here how to correct this approach by keeping consistency with surface tension thermodynamics: for Savart-Taylor sheets, when adequately corrected, we recover the well known $1/r$ liquid thickness with a constant and uniform velocity dictated by Bernoulli's principle. In the case of circular hydraulic jumps, we propose here a simple approach based on Watson description of the flow in the central region [\cite{Watson_JFM}]], combined with appropriate boundary conditions on the formed circular front. Depending on the specific condition, we find in turn the new scaling by \cite{bhagat_2018} and the more conventional scaling law found long ago by \cite{Bohr_1993}. We clarify here a few situations in which one should hold rather than the other, hoping to reconcile Bhagat et al. observations with the present knowledge of circular hydraulic jump modeling. However, the question of a possible critical Froude number imposed at the jump exit and dictating logarithmic corrections to scaling remains an opened and unsolved question.