Performance losses of drag-reducing spanwise forcing at moderate values of the Reynolds number

Abstract

A fundamental problem in the field of turbulent skin-friction drag reduction is to determine the performance of the available control techniques at high values of the Reynolds number Re. We consider active, predetermined strategies based on spanwise forcing (oscillating wall and streamwise-traveling waves applied to a plane channel flow), and explore via Direct Numerical Simulations (DNS) up to Reτ = 2100 the rate at which their performance deteriorates as Re is increased. To be able to carry out a comprehensive parameter study, we limit the computational cost of the simulations by adjusting the size of the computational domain in the homogeneous directions, compromising between faster computations and the increased need of time-averaging the fluctuating space-mean wall shear-stress. Our results, corroborated by a few full-scale DNS, suggest a scenario where drag reduction degrades with Re at a rate that varies according to the parameters of the wall forcing. In agreement with already available information, keeping them at their low-Re optimal value produces a relatively quick decrease of drag reduction. However, at higher Re the optimal parameters shift towards other regions of the parameter space, and these regions turn out to be much less sensitive to Re. Once this shift is accounted for, drag reduction decreases with Re at a markedly slower rate. If the slightly favorable trend of the energy required to create the forcing is considered, a chance emerges for positive net energy savings also at large values of the Reynolds number.