Unsteady behavior of a sweeping impinging jet : time-resolved particle image velocimetry measurements

Abstract

The unsteady behavior of a sweeping impinging jet is measured experimentally using time-resolved particle image velocimetry. For the configuration with a jet-to-wall spacing ratio L/dh = 8, an approximately linear increase in the sweeping frequency is observed with Reynolds numbers (Re) between 2.7 × 103 and 9.3 × 103, especially at high Reynolds numbers. The saturation of the sweeping jet, at which the maximum deflection angle is reached, occurs at Re = 6.7 × 103. Special focus is then placed on the spatial and temporal variations of unsteady flow fields at two Reynolds numbers Re = 4.0 × 103 and 9.3 × 103. The unsteady behavior in the near-exit region is first compared. At a higher Reynolds number, the jet in the near-exit region remains at the maximum deflection angle for a longer time during one oscillation cycle with more concentrated jet momentum, resulting in a faster switching process. In the near-wall region at Re = 4.0 × 103, the sweeping impinging jet exerts a large region of influence due to the oscillation motion of the impact region along the wall. The time-averaged velocity components and velocity fluctuating components near the wall show double peaks on both outer sides and a minimum in the middle of the near-wall region. At Re = 9.3 × 103, due to the rapid and intensive sweeping process, the jet column breaks in the near-exit region, resulting in a weak flow in the middle of the near-wall region. Accordingly, the profiles of the time-averaged velocity components and velocity fluctuating components show higher double-peak values but an even lower minimum value. Finally, the state-of-the-art dynamic mode decomposition method is used to capture the main flow behavior of distinct frequencies at these two selected Reynolds numbers. Bounded by the breaking location of the jet column, the flow with a superharmonic oscillation frequency in the middle of the near-wall region disappears at Re = 9 × 103. Therefore, the energetic flow patterns are found to be more evenly distributed in the near-exit region than in near-wall region. The phase correlation between the captured flow patterns is determined by projecting the phase-averaged flow fields onto the most energetic modes.