Improved calculations of leading-edge separation from slender, thin, delta wings

Abstract

In previous calculations (Mangler & Smith 1959) of the vortex-sheet model of leading-edge separation, only qualitative agreement was found with experimental observations. Because the numerical treatment of the model was then necessarily incomplete, it was uncertain how far the lack of quantitative agreement was to be attributed to the limitations of the model. The use of an automatic digital computer has now made it possible to reduce the uncertainties in the calculation to a negligible level. The features of interest in the real flow are more accurately predicted and the remaining discrepancies can be ascribed to the deficiencies in the model. The paper describes the method used to locate the vortex sheet and determine its strength in terms of the two boundary conditions on it; assesses the credibility of the results; and relates them to the observations. It is concluded that the model successfully predicts the observed height of the vortex above the wing, though the predicted lateral position is in error by up to 6% of the semi-span of the wing. This error falls as the incidence increases and is less when transition occurs in the boundary-layer upstream of secondary separation. Normal force is predicted accurately as is the distribution of pressure on the lower surface and the inboard part of the upper surface. The observed suction peak below the vortex changes its character when transition occurs in the boundary-layer upstream of secondary separation. The model predicts the suction peak in the turbulent case fairly well, but it is clear that detailed prediction of the suction peak is not possible by a model which is wholly inviscid.