Laboratory and Numerical Modeling of a Stably Stratified Wind Flow over a Water Surface
Journal
Moscow University Physics Bulletin
Journal Volume
73
Journal Issue
6
Pages
710-715
Date Issued
2018
Author(s)
Abstract
Abstract: The objective of this paper was to perform laboratory modeling and direct numerical simulation of a turbulent wind flow over a water surface under stable stratification conditions of the air boundary layer. Laboratory and numerical experiments were carried out with the same bulk Reynolds (Re) and Richardson (Ri) numbers, which first allowed direct comparison between measurements and calculations. A wind flow with an air–water temperature difference of up to 20°C and a relatively low wind speed (up to 3 m/s) were obtained in laboratory experiments in the wind–wave flume of the large thermostratified tank at the Institute of Applied Physics of the Russian Academy of Sciences. This allowed a sufficiently strong stable stratification with a bulk Richardson number of up to 0.04. The air velocity is obtained using both contact (a Pitot tube) and particle image velocimetry methods. At the same time, the air temperature profile is also measured by a set of contact probes. Analogous bulk Richardson and Reynolds numbers are prescribed in the direct numerical simulation, where the turbulent Couette flow is considered as a model of the near water constant-stress atmospheric boundary layer. The mean velocity and temperature profiles obtained in our laboratory and numerical experiments agree well; they are also predicted well by the Monin–Obukhov similarity theory. The experimental results state that sufficiently strong stratification, although it allows a statistically stationary turbulent regime, leads to a sharp decrease in momentum and heat fluxes. For this regime it is demonstrated that the turbulent Reynolds number for the boundary layer (based on the Obukhov length-scale and friction velocity) satisfies the known criterion that characterize stationary strongly stratified turbulence.
Type
journal article