Aerodynamic Fidelity of Ice Accretion Simulation on a Subscale Model

This paper presents the results of a study designed to determine the level of fidelity required to accurately reproduce the performance effects of ice accretion on airfoils. Castings of streamwise and spanwise-ridge ice accretions were obtained from tests in an icing wind tunnel. The streamwise-ice accretion was characterized by a smooth conformal shape on the leading edge followed by rime feathers downstream. For the spanwise-ridge ice accretion the leading edge of the airfoil was smooth. There was an upper surface ridge at x/c = 0.10 and a lower surface ridge at x/c = 0.17. The high-fidelity casting simulations were attached to the leading edge of a subscale (457.2-mm (18-inch) chord) NACA 23012 model and aerodynamic performance testing was performed at a Reynolds number of 1.8×10⁶ and a Mach number of 0.18. Lower-fidelity simulations were developed for each ice accretion to determine the fidelity required to match the performance of the airfoil with castings. For the streamwise ice, it was found that the addition of roughness to the lower-fidelity simulations was important to accurately model the aerodynamics. This paper reports the details of roughness size and concentration studies performed for the streamwise-ice simulations. It was found that the addition of roughness to the lower-fidelity simulations of the spanwise-ridge ice resulted in less aerodynamic fidelity. The flowfield associated with this ice shape was found to be highly three-dimensional, and therefore, very difficult to model with simple methods. This paper describes the flowfield and implications for aerodynamic simulation. The performance results are summarized and compared to previous work for horn-ice simulations.