Modeling, Simulation, and Flight Tests for a T-38 Talon with Wing Fences

摘要

A computational study, a wind-tunnel analysis, and a flight test on the T-38 Talon demonstrated that approach-to-stall characteristics may be improved by the addition of a wing fence. Fences were placed at the location of a preexisting seam, at ±0.825 semispan, and were compared with the T-38 without fences. Reynolds-averaged Navier-Stokes simulations were conducted using the computational fluid dynamics air vehicles unstructured solver to examine the flow around the T-38 at typical full-flap landing conditions (Re = 4.66 × 10~6 per unit length). Solutions were computed over a range of angles of attack from 2 to 15 deg. The fence was found to increase C_(Lmax) by 7% and delay the attainment of C_(Lmax) from a 12 to a 13 deg angle of attack. This effect was achieved through the formation of a discrete streamwise vortex outboard of the fence, which prevented flow separation in the tip region, which (in turn) delayed flow separation over the remainder of the wing, resulting in the higher computed lift. Wind-tunnel results, by and large, were based on aircraft instrumentation and flow visualization for Mach numbers up to 0.65 (Re_(mac) ～ 10~7). The flight-test data showed an increase in C_L of approximately 2 % in the angle-of-attack range of 12-14.5 deg when the wing fence was in place. Onboard accelerometer data suggest the fence reduced the roll-off tendency and the wing-rock amplitude during approaches to stall. Flow visualization on the aircraft wing in both the wind-tunnel study and the flight test suggested that the fence reduced spanwise and separated flow outboard the fence, which agreed with the computational fluid dynamics results.