Turbulent dispersion from an elevated line source: Markov chain simulations of concentration and flux profiles

Abstract

AbstractA Markov chain model that simulates the trajectories of fluid particles in turbulent flow is described. It incorporates fluctuations in vertical and streamwise velocity, and can be used with velocities that have a skewed probability distribution. The model is used to estimate the distribution of heat downwind from an elevated line source in the boundary layer above a rough surface, and the results are compared with those from a wind tunnel experiment. The simplest model, with fluctuations in vertical velocity only, underestimates the plume depth by 4% to 22% at various distances downwind, and underestimates the rise in the mean height of the plume by 1·5 mm in 25·5 mm. The inclusion of streamwise velocity fluctuations increases the vertical plume width by 6% and also improves the prediction of the mean height of the plume. The model also gives profiles of streamwise eddy flux of heat u'θ that agree closely with those measured. Skewed velocity distributions can be included in the model, but in the boundary layer above a rough surface Gaussian velocity distributions are quite adequate.