A CFD investigation of wind tunnel interference on delta wing aerodynamics
To explore the influence of wind tunnel test facilities on delta wing aerodynamics, the interference has been separated into two distinct types, wall interference and support structure interference. The wall interference effects have been split into three further components, tunnel blockage, side wall interference, and roof and floor interference. Splitting the tunnel influence in this way allows us to determine the most detrimental interference effects, thus allowing the wind tunnel engineer to design experiments accordingly. Euler and more realistic RANS simulations of tunnel interference have been conducted. To reduce the question of grid dependence when comparing solutions, a common "farfield grid" was created and tunnel grids were extracted. Before doing RANS simulations an analysis of various turbulence models was conducted. It was found that turbulence models have difficulty in predicting turbulence levels in leading edge vortices. As such modifications have been applied to the models which improve predictions. Despite vortex breakdown being widely regarded as an inviscid phenomenon, dependence on turbulence modelling has been exhibited. This is due to the vortex properties being altered with turbulent diffusion of vorticity when turbulence levels are too high. Both 1- and 2-equation models were assessed and it was concluded that a modified 2-equation k-w model was the most suitable of the models available (when compared against experimental results), and was therefore used in all subsequent simulations. From both Euler and RANS simulations it has been concluded that the effect of sidewall proximity significantly promotes vortex breakdown. Side wall induced velocity components increase the mean effective incidence of the wing, the helix angle and the strength of the vortices. The combination of these effects promotes vortex breakdown. Roof and floor proximity has little effect on vortex breakdown as does the frontal area blockage. Pitching simulations have shown that the promotion of vortex breakdown is not consistent on both the upstroke and downstroke. Break-down was observed to be prompted furthest at the higher incidence of the upstroke and on the downstroke. This highlights the dependency of tunnel interference on vortex strength.