Abstract:
The present study proposes to investigate the flow structure of different arrays of
jets impinging on a flat surface enclosed by three walls, creating a channel that forces the
air to leave in one preferential direction, generating a self sustain crossflow. Details of
both the mean and some turbulence quantities for two 7x7 jet arrays are presented. The
first array was a circular orifice array while the second was a "cusped" ellipse array. The
non-symmetry of the cusped ellipse shape allowed two alignments with the crossflow,
one where the crossflow was along the major axis (0°) and one where the crossflow was
along the minor axis (90°). The Reynolds number ranged from 8,500 to 15,900.
Surface flow visualization, jet orifice flow coefficient measurements and PIV
measurements of the entire flow field were used to interpret the complex flow
characteristics. The mean velocity, turbulent kinetic energy, mean vorticity and mean
squared vorticity fields were calculated from the PIV data at three different locations
along the downstream flow direction: close to the back, at the center of the jet array and
close to the exit.
The visual observation of the impinging flow pattern shows similar results for all
of the different geometries and configurations. The jets generated cells defined by
detachment-reattachment zones with a characteristic horseshoe shape around each jet.
The cells become increasingly more oblong in shape with increasing crossflow due to an
uneven shifting of the upstream and downstream cell boundaries, in the downstream
direction. The flow coefficient of each jet decreases monotonically and quasi-linearly
along the crossflow direction while the average value increases only a few percent for
increasing Re. PIV measurements of the entire flow field in the vicinity of the jet exits
reveals complex flow structures. A large rotating vortex created by the merging of the
crossflow and the jet column return-flow, moves towards the jet column as the crossflow
increases in strength. This interaction appears to generate multiple turbulent flow patterns
that may have consequences in improved surface cooling. For the lower values of Re, the
cusped ellipse jet array, placed in the (0°) position, appears to generate significantly more
turbulence than the circular jet array. For higher values of Re, both configurations show
similar results. But when placed in the (90°) position, there is a considerably smaller
increase in the levels of turbulence when compared to the circular jets. Evidence of axis
switching in the jet column for the cusped ellipse jets tends to prove that a transport in the
lateral direction could be a secondary factor to take into account for surface cooling
efficiency.
