(642b) Relative Dispersion for Turbulent FLOW IN A Channel and Differences BETWEEN Backwards and Forwards Dispersion IN Time

Understanding the movement of particle clouds, say in case of pollutant dispersion in the atmosphere or effluent disposal into lakes, is of utmost importance. Two different scenarios arise when studying the motion of a cloud of particles around its center of mass (i.e., turbulent relative dispersion). The first is the case of clouds of particles that arise from a point and move out from the center of mass, which can be termed as forwards relative dispersion, because the particles move forwards in time. The second is the case where particles over a period of time move about and arrive at a particular location, which can be referred to as backwards relative dispersion. This study utilizes a unique approach, which couples a direct numerical simulation of turbulent flow with the tracking of scalar markers in the flow field, termed as Lagrangian scalar tracking, to analyze the effects of relative dispersion. Turbulent flow in an infinite channel with dimensions 4?Îh x 2h x 2?Îh in the x, y, z directions (where h is the half channel height, and h =150 in viscous wall units) is considered for the study. Four main regions of turbulent channel flow, namely the viscous sub-layer, the transition region, the logarithmic layer, and the outer region of the flow at the center of the channel are examined. Turbulent dispersion statistics are determined at these locations for forwards relative dispersion, by tracking the cloud of markers diffusing from a point, and for backwards relative dispersion, where the statistics of the cloud of markers that arrive at a particular region are calculated. Prior results for the case of isotropic turbulence indicated that backwards relative dispersion is faster than forwards relative dispersion [1], but single particle dispersion data for anisotropic turbulence indicated that there is a dependence on the distance from the wall and on the Prandtl or Schmidt number of the dispersing particles [2]. In this presentation, relative dispersion statistics will be presented as a function of the turbulent flow region and as a function of the particle Prandtl or Schmidt number, ranging from Pr=0.1 to Pr=100.

References cited [1] B.L. Sawford, P.K. Yeung, M.S. Borgas, Comparison of backwards and forwards relative dispersion in turbulence, Physics of Fluids 17 (2005) 095109. [2] Srinivasan, C., and D.V. Papavassiliou, "Backwards and forwards dispersion of a scalar in turbulent wall flows," Int. J. Heat Mass Transf. 53, 1023-1035, 2010.