Abstract

Particle-laden boundary flows occur in many geophysical and industrial environments yet are difficult to understand and quantitatively describe because the interactions of an often turbulent boundary layer flow with surface and particle dynamics are complex. The SLF wind tunnel allows the investigation of snow and sand particle laden boundary layer flows over their natural surfaces with and without the presence of a grass canopy.The experimental results are used to investigate the two possible approaches in describing the surface dynamics: (i) Models of particle transport, which assume a stationary flow situation and predict a mass flux for an hypothetical equilibrium; (ii) Models that take the temporal variability of the mass flux explicitly into account such as Lagrangian Stochastic particle tracking Models (LSM) on the basis of large eddy simulation (LES) or direct numerical simulation (DNS) of flow and turbulence. This presentation shows that wind tunnel data support the form of semi-empirical equilibrium models, which predict mass flux, q, as a function of the mean wind speed or the friction velocity, u, and a threshold velocity, uth: q=a(u-uth)x. For the exponent "x", a value of approximately 3, as based on theoretical considerations, is consistent with the data. This simple form of equilibrium models as well as more complicated equilibrium models are all based on the hypothesis that the surface shear stress induced by a fluid on the ground during sediment saltation is constant, i.e. independent of the magnitude of the particle mass flux (Owen's second hypothesis). Our surface shear stress measurements in a drifting-sand wind tunnel show a constant value of the fluid shear stress for saltation layers of various mass-flux magnitudes, directly validating Owen’s second hypothesis for the first time. The equilibrium models, however, only insufficiently describe the full dynamics of particle-laden flows. The second part of the presentation therefore discusses non-equilibrium features such as a high variability of the particle mass flux caused by flow turbulence and surface heterogeneity. Mass flux intermittency is primarily observed around the threshold value uth. Using a combination of LES and LSM models, we show how the simulation of individual feed-back processes leads to a more complete understanding of the mechanisms behind the flux variability.

Details

Actions