Flow structure, droplet formation and dispersion when an interface separating immiscible fluids is impinged with a vertical negatively buoyant jet.

摘要

This dissertation summarizes a series of experiments to determine the effects of impinging a water and diesel fuel interface with a negatively buoyant water jet. The maximum depth that the jet penetrates is only a function of a pipe exit Richardson number divided by a jet spreading factor, which is only a function of distance between the jet exit and the interface. As long as the jet is turbulent, the penetration depth is independent of Reynolds and Weber numbers. A universal correlation developed in this study relates the penetration depth of jets in any fluid systems.; The macroscopic flow structure is categorized into distinct flow regimes that are dependent on an interface Richardson number, Ri_i. At high Ri_i the jet forms a smooth and stable deformation in the interface with no flow separation and no droplet formation. At moderate Ri_i the deformation becomes taller and steeper and a flow separation develops at its edge. An oil lip created at the flow separation sometimes detaches to form oil droplets in the water below. At lower Ri_i the deformation becomes unstable and alternately collapses and reforms. Droplets result when the collapsing deformation impacts the interface and drags down fingers of oil. Droplet size distributions are lognormal and depend on Ri_i, R_e , viscosity ratio and Morton number. Mode diameters range from 0.6mm to 1.5mm.; Experiments demonstrate that the mean droplet rise-rate, Ū , varies between 0.2 to five times the rise-rate in quiescent fluid, U_q, increasing with decreasing diameter. At high turbulence intensity, u′, Ū asymptotically approaches 0.25u′, while at low u′, Ū asymptotically approaches U_q. At intermediate u′, Ū is a strong function of Stokes number, St. In this case, at low St, Ū = U_q, while at high St, Ū U_q. Data from this study is the first to show an enhancement of rise-rate of droplets that are less dense than the continuous phase. Trends are explained in terms of trajectory biasing and non-linear drag effects.