 |
Mapping of crossflowing jet velocity, scalar and associated fluctuation fields
With Prof. M. Godfrey Mungal, Stanford University
(Figures: flow schematic sample images sample quantitative results) (Publications)
The crossflowing turbulent jet, in which a round jet is injected into a perpendicular fluid stream, has considerable practical significance in engineering systems. Major combustion applications, for example, include aerospace propulsion and gas-burning power generation. Significant effort has been devoted to modeling and simulation of the crossflowing jet. Problems confronting modeling work include the complicated vortical structure of the flow, and a lack of detailed experimental measurements for model validation and assessment.
Figure 1. Schematic of the crossflowing jet, where the jet issues in the x-direction into a crossflow that moves in the y-direction, showing the characteristic vortical structures.
Figure 1 shows a schematic of the flow arrangement, showing the horseshoe vortices that form at the windward side of the jet exit, the jet shear layer instability on the jet windward surface, the wake vortices, and the counter-rotating vortex pair that dominates the downstream flow organization. The goal of this work is to map comprehensively the velocity, scalar, and associated fluctuation fields in the developing region of the flow. Particle image velocimetry (PIV) measures the velocity fields, while simultaneous planar laser-induced fluorescence provides the scalar field information. Measurements are made in the jet symmetry plane (the x-y plane) as well as in parallel planes offset in the out-of-plane, z-direction. In this work air is both the jet and crossflow fluid, where the jet is seeded with acetone for the PLIF, and the crossflow is seeded with glycerol-fog particles for PIV. A 308 nm XeCl excimer laser excites the fluorescence, while an Nd:YAG laser at 532 nm provides the light sheet for PIV. The jet nozzle diameter is d = 4.6 mm, the jet exit Reynolds number is 5000, and the jet-to-crossflow velocity ratio is r = 5.7.
Figure 2. (a) A representative scalar field from the jet center plane (the x-y plane) measured by PLIF. (b) The mean two-dimensional streamline pattern in the center plane, determined from the PIV measurements of the velocity field.
Analysis of the full set of the velocity and scalar field results demonstrates that the velocity field observes fairly closely the expected pure jet scaling properties near the jet exit, and approaches the expected wake scaling properties in the far field, while the scalar field does not follow the canonical scalings. This discrepancy can be attributed to strong departures from the expected similarity profiles in the scalar field. Comparisons with the results of other studies also suggest that the jet exit profile strongly affects the flow development (the present experiments used a pipe-flow exit profile). More detailed discussion of the flow physics can be found in the publications below.
The present work also contributes significantly to modeling and computation efforts in the crossflowing jet geometry by providing extensive mapping of the velocity and scalar fields, and associated turbulence quantities. For computation of turbulent mixing, key quantities include the turbulent scalar flux components, namely the products of the scalar fluctuation term and the velocity component fluctuations. Figure 3 shows fixed-y and fixed-x profiles of the scalar flux components u'C' and v'C' in the symmetry plane.
Figure 3. Fixed-y (top) and fixed-x (bottom) profiles of the mean values of the scalar flux components u'C' and v'C' in the jet symmetry plane. The three-dimensional nature of the data set permits the extensive mapping of mean and turbulence quantities throughout the flow.
Among the notable features of these profiles are the sign changes in both the u'C' and v'C' components, which reflect the importance of both the jet outer and inner boundary in the scalar mixing; the higher magnitude of the u'C' component, suggesting that turbulent scalar transport perpendicular to the crossflow is more significant than transport in the crossflow direction; and the relatively higher u'C', thus stronger turbulent transport, on the windward side relative to the wake side of the flow. Profiles such as these, for quantities such as turbulent stresses etc., are also very useful simply for comparative purposes with computational results. Again, further details can be found in the publications below.
Selected publications (inquiries: email Lester Su, lsu: jhu.edu)
All materials © 2008 Applied Fluid Imaging Laboratory. Last page update 1.2.08.
|
