Primary Breakup in Gas/Liquid Mixing Layers for Turbulent Liquids

Abstract

An experimental study of primary breakup of turbulent liquids in gadliquid mixing layers is described. The experiments involved mixing layers along large liquid jets (3.6.6.4 and 9.5 mm dia.) injected at various velocities into still air at atmospheric pressure with fully-developed turbulent pipe flow at the jet exit. Liquids studied included water, glycerol (42% glycerin by mass) and n-heptane. Pulsed shadowgraph photography and holography were used to find conditions where turbulent primary breakup was initiated and drop sizes and velocities after primary breakup. Drop sizes after primary breakup satisfied Simmons' universal root normal distribution and can be characterized solely by their SMD. Mass weighted mean streamwise and crosstream drop velocities after primary breakup were comparable to mean streamwise and crossweam rms fluctuating velocities in the liquid, respectively, with effects of mean velocity distributions in the jet passage reflected by somewhat lower streamwise drop velocities near the jet exit. Conditions for the initiation of turbulent primary breakup and the variation of SMD with distance from the jet exit were correlated reasonably well by a phenomenological analyses considering effects of surface tension and liquid turbulence properties alone.

Conditional damped random surface velocity model of turbulent jet breakup

“…Sitkei [4] and Lebedev [8] use a simpler energy balance, obtaining the same scaling. Wu et al [9] use similar arguments combined with inertial range scaling to estimate D 32 . Broadly, these arguments have three problems: 1. the arguments implicitly assume that the droplets have zero velocity at d 0 r axis spray…”

“…Wu et al [9] and Kerstein et al [13] estimate the breakup onset location x i by finding the time required for breakup to occur, but they estimate this time differently. Kerstein et al is more consistent with the data aside from the proposed Re l0 dependence.…”

“…Only initial droplet diameter and velocity measurements are compared because the constant k and Λ approximations may be inaccurate downstream. For initial D 32 , three data sources are available [1,9,27]. For initial droplet radial velocity, the only available data is from Wu et al [9, …”

Liquid Breakup in Dense Sprays

“…Extensive work on this primary breakup process has been carried out earlier in this laboratory. This work was limited, however, to the primary breakup of round turbulent liquid jets in still gases with fully-developed turbulent pipe flow at the injector exit, see , Sallam et al (2002), Faeth (1993,1995) and Wu et al (1992. Present measurements were carried out considering round liquid jets for various liquids produced using long injector passage lengths (greater than 50 injector diameters) with sufficiently-large injector Reynolds numbers to ensure fully-developed turbulent pipe flow at the injector exit, based on criteria developed by .…”

Flow characteristics of a large-size pressure-atomized spray using DTV

“…higher than 500. Atomization in this case is governed by the turbulence of the liquid: the liquid surface is disturbed by the scales of turbulence having sufficient energy to overcome surface tension forces ( Wu et al, 1992;Wu and Faeth, 1993 ). It is the smallest of these scales that forms ligaments on the surface of the liquid at the nozzle outlet.…”