Interference factors are developed as a function of the degree to which the wake is deflected downward. At large wake deflections the interference may be much greater than indicated by classical theory. Methods are given for extending the present numerical results to tests involving multielement and finite-span models. The theory can be at least partially verified by means of available data. Tables of calculated interference factors are presented in NASA Technical Notes D-933, D-934, D-935, and D-936.

A superposition method utilizing a digital computer is developed to obtain wall interference for arbitrary configurations. A variety of specific configurations are treated. Sample numerical results indicate that a large number of variables, such as wind-tunnel configuration, model configuration, wake deflection, model location, span of wing and tail, load distribution, sweep, angle of attack, pivot location, tail length, and tail height, may individually or collectively produce substantial effects on wall interference. Interference is particularly severe at the rear rotor of tandem systems; the maximum size of such systems for reasonable wall effects is discussed.

The wind tunnel boundary interference on a V/STOL model is calculated in a rectangular test section with solid vertical walls and ventilated (perforated or slotted) horizontal walls. The interference is found by applying the small perturbation theory of an incompressible fluid to the boundary value problem. The theory uses an image method in addition to Fourier transforms with an equivalent homogeneous boundary condition on the ventilated wall. The mathematical representation of the V/STOL model accounts for the curvature and decay of the wake. The assumption of a constant wake strength produces a paradox in that the maximum value of the interference factors increases as the initial jet velocity decreases. The most significant aspect of the analysis shows that nonlinear cross-flow effects at the tunnel boundary are important in the V/STOL case, and a quasi-linear approximation to these effects is introduced into the solution providing good agreement with experimental data.