Recent changes in aircraft certification requirements has elevated icing performance to a level that it must now be considered in the earliest stages of aircraft design. However, ice accretion on the swept wings of commercial transport aircraft is highly 3D and the ice accretion physics and resulting iced wing aerodynamics are not well understood. Thus, manufacturers and government researchers and regulators are interested in improving computational and experimental aircraft icing tools. This is particularly critical as the industry attempts to move toward Certification and Qualification by Analysis to improve safety and reduce certification costs.
This talk presents an overview of a multi-year international collaboration to improve our understanding of swept-wing ice accretion and aerodynamics as well as to produce a database of experimental data. The research team was led by NASA with support from the FAA and ONERA. The University of Washington, the University of Illinois at Urbana-Champaign, and the University of Virginia participated as well as Boeing.
A full-scale aircraft based on the NASA Common Research Model was used as the baseline. State-of-the-art computational ice accretion and aerodynamic methods were used as well as testing in three wind tunnels to conduct the program. New computational and experimental methods were developed to undertake this complex experimental program. Ice accretion testing was performed in the NASA Icing Research Tunnel on three hybrid wind tunnel models designed using 3D N-S methods including the effect of the wind tunnel walls. Ice accretion shapes were laser scanned and the digital representations manipulated to generate 3D, full-span ice accretions. Semispan wings with varying fidelity of ice shapes were tested in atmospheric and pressure tunnels in seven different test campaigns. Balance, surface pressure, 3D full wake profiles and surface flow visualization were used to explore the aerodynamic performance.
Significant reductions in maximum lift, pitching moment and large drag increases were seen. As in earlier airfoil studies, the aerodynamics with ice shapes were relatively independent of Re and M as compared to the clean wing. Many cases showed complex leading-edge vortex separated flow that exhibited two distinct flow patterns depending on ice shape fidelity of the 3D features. In general, the ice shapes generated using current state-of-the-art computational methods provided non-conservative results. Attempts to classify the ice shapes by the dominant aerodynamic phenomena were hampered by a lack of understanding of the flowfield of the full 3D ice shapes and suggests more fundamental studies are needed.
