When a material spreads in the turbulent flow, its instantaneous concentration distribution becomes not homogenous in space, and areas with high concentration containing sudden change of the concentration are formed by the local intense turbulence. In this study, we extracted such localized high-concentration areas by a conditional sampling technique, and observed the behavior of the areas.
In order to realize the scalar diffusion in quasi homogeneous isotropic turbulent flow, a model experiment was performed in a water channel flow. Fluorescent dye was introduced from a nozzle located at the center of the channel, and the concentration images of the dye were obtained at several downstream positions by PLIF measurement. To extract areas of high concentration containing a sudden change of the concentration, three types of analysis techniques including the conditional sampling technique were applied to PLIF images.
By the conditional sampling technique, we can extract areas of high concentration, which cannot be identified by the other two kind methods, and the effectiveness of this technique was proved. It was found that the areas extracted by the conditional sampling technique appear as lumps. The spanwise numerical probability distribution of the lumps roughly follows the Gaussian distribution, and the peak of the distribution decreases while its standard variation increases as streamwise distance increases. This result implies that performing extraction of areas with high concentration appropriately would enable to realize the material diffusing process as a Lagrange particle and provide a prediction method about the material diffusion source and the damage of the pollutant.