According to Stewartson (1969, 1974) and to Messiter (1970), the flow near the trailing edge of a flat plate has a limit structure for Reynolds number Re →∞ consisting of three layers over a distance O (Re -3/8 ) from the trailing edge: the inner layer of thickness O ( Re -5/8 ) in which the usual boundary layer equations apply; an intermediate layer of thickness O ( Re -1/2 ) in which simplified inviscid equations hold, and the outer layer of thickness O ( Re -3/8 ) in which the full inviscid equations hold. These asymptotic equations have been solved numerically by means of a Cauchy-integral algorithm for the outer layer and a modified Crank-Nicholson boundary layer program for the displacement-thickness interaction between the layers. Results of the computation compare well with experimental data of Janour and with numerical solutions of the Navier-Stokes equations by Dennis & Chang (1969) and Dennis & Dunwoody (1966).

No abstract

No abstract

Thirty-eight data sets from static tests of various 65° delta wings in many water and wind tunnels are compared with four empirical vortex breakdown location prediction methods and the results of two Navier-Stokes computations to assess their range of validity in pitch. Vortex breakdown is the sudden expansion and subsequent chaotic evolution of the otherwise orderly, spiraling, leading-edge vortex flow over the upper surface. Large fluctuations occur in vortex breakdown location at static test conditions making accurate experimental determination difficult.The prediction methods do not account for the seemingly minor geometric details that vary between the models such as thickness, leading-edge bevel angle and radius, trailing-edge bevel angle, sting mounting, instrument housings, etc. These geometric variations significantly affect the position of vortex breakdown and degrade the accuracy of the predictions. The large changes in the flow produced by small geometric changes indicate that an efficient flow control strategy may be possible. Many of the data sets are not corrected for tunnel wall effects, which may account for some of the differences. Data presented herein are as published by the original authors, without additional corrections.The 65° delta wing in pitch and roll at high angles-of-attack was extensively tested by Huang and Hanff (1) and subsequently by Cipolla (2) , Pelletier (9) and Addington (18) who used nearly identical models albeit at much lower Reynolds numbers (Re) based on root chord, c. A large difference in vortex breakdown location is evident when comparing the data of Cipolla (2) (Re = 32,400) for the nearly identical wing to that of Huang and Hanff (1) (Re = 3⋅6m) at zero roll angle (ϕ) and 30° sting angle (α). The vortex breakdown location found by Pelletier (9) (Re = 0⋅1) and Addington (18) (Re = 0⋅29m), also for a nearly identical wing, agrees with that found by Cipolla (2) and the Navier-Stokes computations of Gordnier (20) (Re = 32,000). Table 1, a survey of data and prediction methods for delta wings with 65° sweep with various sharp leadingedge bevels and centre bodies, of varying thickness and in various incompressible flows, shows that vortex breakdown is considerably nearer the apex (further upstream) than found in the experiments of Huang and Hanff (1) . This difference, 0⋅3c, exceeds the inherent uncertainty in experimental data due to the highly unsteady oscillation of vortex breakdown location that may amount to 0⋅2c under steady test conditions (1) . The objective here is to determine if correlations of published data can explain the differences in vortex breakdown location.

No abstract