|M.Sc Student||Raviv Sayag Michal|
|Subject||Investigation of the Flow Field about a Plate with an|
|Department||Department of Aerospace Engineering||Supervisors||Professor Emeritus Omri Rand|
|Professor Rimon Arieli|
|Dr. Yossef Elimelech|
|Full Thesis text|
This work constitutes a feasibility study for controlling airfoil flow using a small oscillating bump, by examination of the flow field created about the latter. The study was conducted using two research techniques: An experiment performed in the Technion water tunnel, and numerical simulations of a 2D unsteady flow field, performed in two computational domains: a Tunnel Domain, and a Free Stream Domain. A simplified model of a flat plate was designed for this investigation. The central 10% of the plate's chord composed the flexible bump, capable of oscillating sinusoidally in time, varying from a flat surface to a parabolic arc, with a maximal height of 1% chord.
Water tunnel visualization experiments were performed with a velocity of 0.05 m/sec and a non-dimensional frequency St=0.5. Images of streaklines captured during the experiment verified the formation of separation bubbles behind the oscillating bump, which were shed downstream. The photographed streaklines and the calculated ones are in very good agreement both in the shape of the line and the rate of advancement downstream.
Numerical simulations performed at various Strouhal numbers revealed that the dynamics of separated flow highly depends on the frequency of oscillation. Aerodynamic coefficients were calculated, and were found to vary harmonically in time at the same frequency as the bump. While the lift coefficient is in phase with the height of the bump, a shift in phase is evident for the drag on the model. The time averaged increase in lift and drag is effectively zero for all frequencies, due to a trend reversal during the ‘descending bump’ portion of the cycle.
The effect of water tunnel walls in proximity to the model was examined by numeric simulation. The presence of walls slightly affects the extent and dynamics of the separated flow region, but the effect on pressure distribution along the model is much more pronounced. This is most likely caused by acceleration of the flow in the tunnel, due to narrowing of effective cross section. The variation in pressure distribution between the two domains leads to a significant difference in aerodynamic coefficients, which is not scalable in the case of drag. This is expressed in both the magnitude of the instantaneous coefficient, and the phase compared to the height of the bump. It is inferred that force measurements conducted in a wind or water tunnel on an oscillating geometry are highly unreliable for predicting free stream behavior.