An experimental study was conducted at the University of Illinois to understand the effect of supercooled large droplet (SLD) ice accretion on airfoil aerodynamics. The study consisted of a sensitivity analysis of airfoil lift, drag, pitching moment, and hinge moment to different chordwise locations, sizes, and shape of the ridge-ice simulations. Two airfoils were used in this investigation: the NACA 23012m and the NLF 0414.

The forward-facing quarter round (used as SLD ridge-ice simulation) severely altered the flowfield around the two airfoils tested. A small separation bubble formed upstream of the ice-shape simulation, and a much longer separation bubble formed downstream of the ice-shape simulation. The longer bubble grew rapidly with increasing angle of attack and failed to reattach at an angle of attack much lower than that at which the clean model stalled. This led to severe reduction in maximum lift and a large increase in drag. The pitching and hinge moments were severely altered as well.

The most critical ice-shape location on the NACA 23012m was around 10% chord. This corresponded to the location of the maximum adverse pressure gradient of the clean airfoil, just aft of the large leading-edge pressure peak where most of the lift was generated. When the ice shape was located in this region, the bubble that formed downstream had to reattach in a very adverse pressure gradient. This led to a very long bubble and a severely altered pressure distribution, with the elimination of the leading-edge suction peak.

The effects of simulated ice shape on the NLF 0414 were quite different from the NACA 23012m. There was little variation of lift when the simulated ice-shape location was varied from 2% to 20% chord. The large losses in lift occurred when the ice shape was located downstream of 30% chord, and the separation bubble formed over the adverse pressure gradient of the clean airfoil, which started at 75% chord. The effect of the ice shape on the NLF 0414 was not as severe as on the NACA 23012m because the lift was much more aft loaded.

Increasing the ice-shape height decreased lift and increased drag. Streamlining the ice shape increased lift and decreased drag. The presence of surface roughness in the vicinity of the shape did not have large effects on lift, drag, pitching moment, and hinge moment. The presence of gaps in the spanwise ridge ice simulation significantly increased maximum lift when compared to the full span case. There was little effect on ice-airfoil aerodynamics as the Reynolds number was varied from 1 to 1.8 million.