Experimental and Numerical Study of Multi- Phase Flow in a Coaxial Air Jet

Loading...
Thumbnail Image
Date
2014-09-03
Authors
Payne, Graham
Keyword
Numerical Fluid Dynamics , Experimental Multiphase Fluid Dynamics
Abstract
The ultimate goal of this work was to optimize a polymer flame deposition process by developing an experimentally verified numerical model of it. This process consisted of injecting a polymer powder into an air/propane flame where the powder was heated enough to soften it before it was projected on to a substrate to provide a protective coating. Intermediate goals towards the final goal were identified. This thesis was based on four papers with each paper reporting the results of the work on an intermediate goal. The first paper reported on work with the turbulent coaxial air jet created by shutting off the flow of propane and polymer particles to the torch head. Experimental measurements of the flow field were compared to the results of numerical simulations using different Reynolds Average Navier Stokes (RANS) models. The k-ε realizable model provided the best agreement with the axial velocity measurements, but all the RANS models predicted a recirculation zone immediately downstream of the central jet that did not exist. The next paper described two phase flow created by the introduction of polymer particles into the air stream to the central air jet. A sheet of laser light and high speed digital imaging were used to determine the particle velocities, which varied widely because the particle size and shapes varied widely. The radial particle distribution was roughly Gaussian. The third paper described the use of the large eddy simulation (LES) turbulence model in Fluent to try to improve the agreement between experiment and numerical results. The LES results did not predict any recirculation zones, but agreement with experimental axial velocities was worse than with the RANS simulation results. Overall, no numerical model’s results agreed with all experimental results within 10%, so their further development was abandoned. The fourth paper described mixing when propane was replaced with air or CO2 and seeded with fine water/glycerine droplets. These jets were rapidly but irregularly entrained in the coaxial air jet. Gas sampling showed this irregular entrainment was related to significant variations in the concentrations of CO2 off the jet centerline. Introducing a particle stream had little effect on the CO2 concentrations.
External DOI