On a Turbulent Mixing Layer Created Downstream of a “Λ” Notch Simulating One Wavelength of a Chevron Nozzle

Abstract

Measurements were carried out in a turbulent mixing layer formed downstream of a splitter plate, that had a -shaped trailing edge. The results revealed that the center of the mixing layer shifts toward the high-speed flow while its sides bend toward the low speed stream at larger distances from the splitter plate. This suggests the existence of a counter rotating streamwise eddies that dominate the flow and substantially increase its level of turbulence relative to the classical plane mixing layer. The change in the orientation of the vorticity, emanating from a chevron nozzle, decreases the susceptibility of the flow to spanwise uniform periodic excitation relative to a classical plane mixing layer.

The application of boundary layer independence principle to three-dimensional turbulent mixing layers

“…In the present case, it is bent further with increasing spanwise distance from the plane of symmetry. This bending was first reported by Zakharin et al (2009). The dominant mixing layer direction (either normal or parallel to the TE) in the vicinity of the trailing edge is determined by the sweep-back of the trailing edge or the nozzle lip.…”

“…This process may explain the effectiveness of the chevron in mixing enhancement and noise reduction and it may provide a needed design criterion for its shape. Provisional explanation for the bending of this mixing layer was given by Zakharin, Kit & Wygnanski (2009) and the present observations provide a possible framework for examining 'simple' three-dimensional turbulent flows.…”

“…Preliminary measurements of amplitudes of phase-locked perturbations in a mixing layer that was forced at 50 Hz were carried out by Zakharin et al (2009) at two spanwise locations. These amplitude distributions were almost identical to one another but their phases differed, suggesting that uniform in-phase actuation along the trailing edge, amplifies eddies in the normal direction only.…”

“…The splitter plate was 2.54 cm thick in order to contain electromagnetic actuators in its interior. Active Flow Control (AFC) slots are shown in figure 2(b), they have been used by Zakharin et al (2009) and more results are forthcoming as the instability of this flow is of great interest. The mean velocities were measured using a rake of Pitot tubes that could be traversed in all three directions.…”