One of the best known results in finite wing theory, which is quoted in aerodynamic text books, states that the distance between two rolled up vortices behind a wing with an elliptic spanwise load distribution is sπ/2, where s is the wing semi-span. Within the framework of linear wing theory this result is correct. But it is of interest to enquire how this result is affected when the standard linear model becomes less valid, for example, at higher wing lift coefficients.

In this note, it is assumed that a continuous trailing vortex sheet rolls up into two discrete vortices. These two discrete vortices are assumed to have cores of finite radius; inside the cores the flow is taken to be solid body rotation, while the flow outside the cores is the standard irrotational vortex field. It is assumed that this rolling up process takes place far downstream of the finite wing which generates the trailing vorticity.