Boundary restaint effects and correction ruler for two-dimensional separated flow past bodies.

Abstract

It is now recognized that considerable uncertainties and inadequacies exist in methods for effecting blockage or boundaryrestraint corrections in wind and water tunnels. These factors have led to the search for alternative approaches such as the adaptivewall method, boundarymeasurement methods, and the momentumbalance method. In this thesis, studies of the effect of solid and freejet boundaries in twodimensional separated flows have been carried out using freestreamline theory, modeled to yield flows closer to reality than those in the pioneering works of Kirchhoff and Helmholtz. The models employed are the Momentum Surface Model and the Displacement Surface Model. Solid boundaries, freejet boundaries, and partially open solid boundaries have been considered. The studies have shown that modeling plays a vital role in predicting blockage effects, even when both models are quite realistic. In the displacement surface model, at large blockages, the force and moment coefficients acting on a body increase with increasing blockage at the same backpressure coefficient. In contrast, in the momentum surface model, the opposite trend is observed for the aerodynamic coefficients in a closedjet tunnel. Separated flows past bodies in the case of:

a free jet issuing from a solid nozzle, and a free jet of infinite length

have also been analyzed using the Displacement Surface Model. These studies confirm that, even for separated flows past bodies in completely closed and completely openjet tunnels, opposite trends of blockage are observed-consistent with the behavior seen in attached flows. Following these analyses, it was conjectured that a suitable combination of rigid and openjet boundaries in a tunnel could yield a blockagefree separated flow. Suppression of boundaryrestraint effects is shown to be feasible for a particular value of the partial opening in the solid wall and the pressure at the opening. In all cases, the back pressure in the nearwake separation bubble emerges as an unknown parameter, which can, in principle, take any value in the tunnel model or in unbounded flow unless additional constraints are imposed to establish a unique relation between the two. One such constraint, of course, is to require that flow separation in both situations be boundarylayer separation. In the inviscid domain, other restraints are suggested to obtain a unique relation between the back pressures on the model in a tunnel and in unbounded flow. When the back pressures are arbitrary, the correction procedure must be semiempirical. Two such methods are proposed, analogous to those used for Reynoldsnumber corrections for drag and lift in airfoil experiments. In addition, brief reviews of:

the development of freestreamline theory for separated flows, and existing boundarycorrection methods

have been presented.