Abstract:
Summary. Experiment has already shown that the aerodynamic damping on half-span models in pitching oscillation is subject to interference effects that may exceed 30 per cent in slotted-wall tunnels at subsonic speeds. An explanatory theory is presented with numerous illustrations and comparisons with measured data. The basic assumptions demand a small ratio of model span to tunnel breadth, small frequency parameter and open instead of slotted boundaries ; nevertheless, the effects of span, frequency and slot geometry are considered theoretically, as well as those of tunnel shape, planform, model size, Mach number and pitching axis. Further experiments have been made to check the theory, and conclusions are drawn from studies in four different tunnels with their slots sealed and open. Direct comparison between prediction and measurement is consistently good with slots sealed and in most cases with slots open. For a particular tunnel large discrepancies in the latter case are attributed to viscous effects that cause a wall with too narrow slots to behave like a closed boundary ; wall interference then changes sign, is less severe, but can no longer be detected by sealing the slots. Although the theory clarifies the problem, the usefulness of slotted-wall tunnels for dynamic measurements is open to question when corrections are very large. With inviscid flow the following conclusions are drawn from the theory: (a) that the lift-damping due to pitching is subject to even more serious interference than the direct pitching damping, (b) that it may be impracticable to obtain adequate reduction in the wall corrections simply by testing smaller models, (c) that complete or more slender models are less seriously affected than half-models of moderate aspect ratio, (d) that with horizontal models in rectangular tunnels the best remedy is to have slotted side-walls but closed roof and floor.