Evolution of the vorticity distribution in vortex rings
Vortex rings are very efficient at transporting fluid on long distances and can generate large forces, either thrust or drag. These abilities are influenced by the vorticity distribution within the vortex. Previous work on vortices produced by piston-cylinders showed that the vorticity distribution reaches a steady state when the vortex separates from the apparatus. First, we experimentally investigate the evolution of the vorticity distribution independently of the vortex separation. The vortices are created by impulsively accelerating cones immersed in water. In this configuration, the self-induced velocity of the vortex is directed towards the cone and there is no separation. Particle image velocimetry is carried on at Reynolds numbers around 30000. The vorticity distribution is quantified using the non-dimensional energy of the vortex, which is the energy with respect to the impulse and circulation. After three convective times, the volume of fluid recirculating within the vortex ring is filled with vortical fluid and the non-dimensional energy to a value around 0.3. The vorticity produced on the cone circumvents the vortex and a portion of the vortex volume is lost via tail-shedding. The translational velocity of the vortex ring linearly depends on its circulation and non-dimensional energy. This velocity, relative to the cone, also converges after three convective times and is found to be a more reliable scaling parameter than energy or circulation. It consistently reaches values around 0.9. In a second part, we present models to predict the vortex growth in the wake of disks and cones. Two models are developed. The first model reduces the vortex ring to a core of constant vorticity density. The translational velocity of the vortex is deduced and its trajectory integrated. The model accurately predicts the maximum circulation of the vortex. A second model, based on axisymmetric discrete vortex methods, predicts the growth, vorticity distribution and tail-shedding of the vortex. A third model is developed to explain why the non-dimensional energy consistently converges to values around 0.3. Based on the self similar roll-up of inviscid shear layers, a non-dimensional energy of 0.33 is computed for vortices formed by impulsively accelerated disks or pistons. The model also predicts that a linear acceleration profile leads to a more uniform vorticity distribution, decreasing the non-dimensional energy to 0.18. This result indicates that the vorticity distribution can be controlled by varying the velocity profile of the vortex generator. Another control option is to use permeable disks. We impulsively accelerated perforated disks and observed the vortex formation. A portion of the incoming flow bleeds through the disk and does not circulate around the disk edge, resulting in a lower vorticity maximum. The vortex ring has a more uniform vorticity distribution, as well as a more elongated shape. The non-dimensional energy is brought down to 0.14. Finally, vortex rings have a great potential to transport fluid on long distances, such as extinguishing powder. Their resilience to vortical perturbations is critical for the transport and depends on the vorticity distribution within the vortex. Simulations with nested contour methods are performed to assess that resilience. Vortices with lower non-dimensional energy shed less vortical volume when facing perturbations and qualify as better candidates for fluid transport.
EPFL_TH8768.pdf
n/a
restricted
Copyright
5.02 MB
Adobe PDF
7eec08b5c84f91a6177f59e2704448ac