| dc.description.abstract | The performance of a diffusing S-duct inlet (M2129) is computationally studied for the effects of inlet icing. Different ice accretion shapes, predicted by numerical analysis in the literature reviewed, are simulated on the inlet lip. Two commercial codes, FLUENT and STAR-CCM+ are used for the steady- and unsteady-state computations. The shear-stress transport (SST) k-e turbulence model and large eddy simulation (LES) turbulence model are applied in the computations. The glaze ice shape, which is characterized by intrusive horns, degrades inlet performance, while the effect of the streamlined rime ice shape is negligible. At the free-stream Mach number of M=0.23, the glaze ice causes a 3.2 percent decrease in the total pressure recovery and a 26 percent reduction in the inlet mass flow rate. This result comes from the massive flow separation and flow blockage from the glaze ice horns. The total pressure recovery is further decreased by 22.8 percent, as the free-stream Mach number increases to M=0.85, due to the increased internal blockage and formation of internal shocks in the S-duct inlet. Also, the glaze iced inlet induces 6.6 percent increase in the engine thrust loss and the specific fuel consumption at M=0.25. The level of the ice-induced flow blockage by the ice accretion is also important for the inlet performance. The symmetrical glaze ice that covers the entire inlet lip portion causes a nearly 11.8 percent decrease in the total pressure recovery at M=0.475, whereas the top- or bottom-asymmetrical glaze ice that accretes on a ¼ portion of the inlet lip leads to just a 2.5 percent decrease. Also, the dynamic inlet distortion level, which is represented by the total pressure fluctuation at the engine face, is almost doubled with the symmetrical glaze ice when compared to the asymmetrical glaze ice. Therefore, the dynamic inlet distortion is proportional to the total pressure recovery that corresponds to the steady-state inlet distortion. Furthermore, the application of local angles of attack and local sideslip angles for the iced S-duct inlet contributes to the further degradation of the inlet performance, regardless of the ice shapes. However, the angles that provide the most distortion for each ice shape all differ due to the combined effects of the angle of attack or sideslip angle, icing location, and downward duct curvature. In addition, both the steady-state inlet distortion and dynamic inlet distortion become most severe at the highest angles tested: symmetrical (angle of attack=+20º), top-asymmetrical (angle of attack=-20º), bottom-asymmetrical (angle of attack=+20º), and side-asymmetrical glaze (sideslip angle=-20º). Finally, a strongly coupled temperature-total pressure distortion is created at the engine face under the icing condition. This coupling, as measured by the total pressure distortion parameter, increases the engine face distortion by 6.97 percent in the glaze iced inlet at M=0.85 when the inlet wall is heated to 350 K. | |
