Low pipeline velocities lead to stratification of liquid-liquid flows, resulting in phase
slip where the in situ phase fractions differ from the input fractions. This prevents
samples being obtained that are representative of these flow’s phase fractions, and
thus, mixing is used to overcome stratification. This work investigates pipeline jet
mixers, obtaining measurements for comparison with theory.
A review of jets for mixing is presented which establishes that few studies
have been published on liquid-liquid jet mixing. Additionally, a review of prior
refractive index matching experimental techniques is included, covering the
techniques used within liquid-liquid and liquid-solid experimental systems.
Importantly, only two liquid-liquid-solid experimental systems are found to be triply
refractive index-matched and none are used for studying pipeline flows.
A refractive index-matched facility is developed, consisting of a 10 mm2/s
silicone oil, a 51 wt/wt % water/glycerol solution and an ETFE pipe. This allowed
the mixing processes to be observed through using laser-based optical techniques,
namely Particle Image Velocimetry to measure 2- and 3-component velocity fields
and Planar Laser-Induced Fluorescence to measure phase distributions, interface
heights and droplet sizes. Measurements are made of the incoming flows, the jet
interactions and the resulting dispersions for a range of flow conditions, jet
velocities and pipeline positions.
Results demonstrate that mixing occurs in two zones: the first during
initial jet interaction where Kelvin-Helmholtz instabilities dominate; the second
occurs downstream due to secondary flows. Mixing is found to decreases as the
jet velocity decreases and the aqueous phase fraction increases. Increasing the
crossflow velocity decreases mixing while simultaneously pushing the point of
optimum mixing downstream. Coalescence is found to dominate breakup by ten
pipes diameter downstream of the jet.
The results presented within will aid the understanding of these flows and
provide sufficient detail to improve the predictive accuracy of computational models.