Mixing in Circular and Non-circular Jets in Crossflow

Abstract

Coherent structures and mixing in the flow field of a jet in crossflow have been studied using computational (Large Eddy Simulation) and experimental (Particle Image Velocimetry and Laser-Induced Fluorescence) techniques. The mean scalar fields and turbulence statistics as determined by both are compared for circular, elliptic, and square nozzles. For the latter configurations, effects of orientation are considered. The computations reveal that the distribution of a passive scalar in a cross-sectional plane can be singleor double-peaked, depending on the nozzle shape and orientation. A proper orthogonal decomposition of the transverse velocity indicates that coherent structures may be responsible for this phenomenon. Nozzles which have a single-peaked distribution have stronger modes in transverse direction. The global mixing performance is superior for these nozzle types. This is the case for the blunt square nozzle and for the elliptic nozzle with high aspect ratio. It is further demonstrated that the flow field contains large regions in which a passive scalar is transported up the mean gradient (counter-gradient transport) which implies failure of the gradient diffusion hypothesis.

Numerical Simulations of Imperfect Bifurcation of Jet in Crossflow

“…While the stability of the flow structure depends on a lot of factors, the bifurcation may not be observed in different models. The simulation results from Salewski et al (2008) for a square JICF show that the mixing flow structure still remains symmetric in the case of M=4. However, their simulations were performed for a wide channel with a width of Y/D=20 which is much wider than that used in the current study.…”

“…Fig.5b shows the V x profile at a further downstream location, X/D=3, which presents a double peaked curve. There is a local minimum at Z/D≈0.7, which stems from the region of largest momentum exchange between the jet and the crossflow (Salewski et al, 2008). Between this local minimum and the minimum on the no-slip wall, there is the first peak.…”

“…Liscinsky et al (1996) observed in their experiments that the mean concentration distributions, mean concentration trajectories and turbulent mixing were not affected significantly by hole shape. Salewski et al (2008) also investigated the structures and mixing in the flow field of a JICF with different hole shapes using both large eddy simulations and experiments. They pointed out that the passive scalar distribution could be single or double peaked in cross-sectional planes, depending on nozzle shape and orientation.…”

Effects of aerodynamic particle interaction in turbulent non-dilute particle-laden flow

“…This approach has an increasing acceptance as due to increase in computational resources a larger part of the turbulence spectrum can be resolved and the unresolved kinetic energy therefore declines with increasing resolution. In this respect, LES is an approximation to DNS, and therefore it is conceptually acceptable that no explicit SGS model is required [34][35][36]. SGS models should drain energy at the smallest resolved waves but should be negligible for the largest waves.…”

Flow Characteristics of a Low Reynolds Number Jet in Crossflow with an Obstacle Block

“…The corresponding results showed that the principal vortex systems from the interaction of the jet with the crossflow at low Reynolds number were the same as those observed at larger Reynolds numbers. Salewski [10] studied the vortex structures in the field of a jet in crossflow by using computational (large eddy simulation) and experimental (particle image velocimetry and laser-induced fluorescence) techniques. The corresponding scalar fields and turbulence characteristics were compared for circular, elliptic, and square nozzles.…”