Current experimental studies concerning three-dimensional separation around a body is discussed in the first half of this review. The evaluations of several measurement techniques are introduced such as laser-Doppler velocimetry, hotwire anemometry, surface hot film, surface pressure, as well as particle image velocimetry. In the second half, the experimental results by the present authors are discussed for the flow with separation over an axisymmetric paraboloid at 50 deg incidence angle at most where smoke visualization techniques and hot wire anemometer with X-type probe were employed. Reynolds numbers is at most 1. 8X 10⁴ referred to the base diameter. Velocity fluctuations and the change of r.m.s. value in various cross-sections are presented.

The supersonic aerodynamic characteristics of lifting bodies with blunt thick delta configuration at high angles of attack are experimentally investigated by using the transonic and supersonic wind tunnels. The results show that the lift-to-drag ratio at the trimmed attack angle can be enhanced by the use of the bent nose and body flaps. Significant temporal nonperiodic fluctuation of the aerodynamic force is observed in the narrow freestream range around the Mach number 1. 5 and attack angle 25 deg. The high speed Schlieren pictures reveal that the intermittent breakdown of the leeward vortex pair causes such unsteadiness. This fluctuation can be abbreviated by the combination of the bent nose and the body flaps located on the upper surface of the body.

Aiming to utilize CFD (Computational Fluid Dynamics) as practical tools for the aerodynamic analysis and design of realistic aircraft, the factors which CFD softwares should possess are pointed out and discussed. On the discussion, in particular, a lot of attention is directed to the applicability of a software to the high angle of attack aerodynamics. Then, one of widely used CFD softwares, “LANS3D,” is validated based on the discussion. Flowfields about a space-plane are simulated using the software at a wide range of the angle of attack. The computational results are compared with wind-tunnel experimental data. They are also examined from the theoretical aspect of aerodynamics. Through the validation, the qualitative reliability of the computational results is confirmed. In addition, the possible usage of CFD as a tool to analyze complex flow phenomena is considered.

Computer codes to compute dynamic stall are developed, then two-and three-dimensional dynamic stall is numerically investigated and compared with experimental results. Both PDNS (Pseudo Direct Numerical Simulation) and RANS (Reynolds Averaged Na vier-Stokes) analysis are used to compute turbulent separated flows. In order to find most appropriate RANS turbulence model, several models are tested at first for static problem and SA (Spalart-Allmaras) model shows best agreement with experiment. Then RANS and PDNS are applied to two-dimensional deep dynamic stall case, in which k=0.1 and angle range 0-20 degree, and both computations show qualitative agreements with experimental results, however, quantitative agreements are not satisfactory. In three-dimensional static case, in which k =0.08 and angle range 12-24 degree, RANS computation with SA model shows good agreement. For dynamic stall, quantitative agreements of both model are not satisfactory, however, RANS computation exhibits better results. Thus, it is expected that RANS will give good estimation for more moderate parameters, because agreement in static case is good and the dynamic case is another extreme in practical case.

The aerodynamic force and PIV (Particle Image Velocimetry) measurements over the wing of a 3-D basic model in pitching motion were conducted in MHI Nagoya's low speed wind tunnel to investigate the dynamic lift and vortex behaviors. In this paper, experimental results of some wings on the various pitch rate conditions are shown. The dynamic effect begins to appear on the high angle of attack because the dynamic vortex is formed and prevents the separation. At that time, the sweepback angle of the wing is the main parameter to maximize the dynamic effect. And the dynamic vortex distribution analyzed from the PIV measurement results shows the followings. The dynamic vortex sheds over the wing from the wing root leading edge to the wing tip trailing edge and starts to break down instantly after the pitching motion is stopped. Moreover by comparing the dynamic vorticity with the dynamic lift, it is turned out that both have a close correlation.