Analytical and Experimental Determination of Airfoil Performance Degradation Due to Ice Accretion

Abstract

A physics-based, empirical correlation between icing conditions and corresponding drag coefficient was developed for NACA 0012 airfoil and compared to other three existing prediction methods. The correlation was developed based on experimental aerodynamics database of iced airfoils and derived using statistical methods. The correlation model also provides drag coefficients for varying angles of attack under a given icing condition. The calculated drag coefficients matched (33.40% mean absolute deviation) with reference data from three different experimental databases. To validate the proposed degradation model and to further extend the database for helicopter rotor performance degradation, rotating ice accretion experiments were conducted. Four ice shapes obtained at the NASA Icing Research Center were reproduced on a 21-inch chord, 4.5-feet radius NACA 0012 rotor blade at the Adverse Environment Rotor Test Stand facility. Ice shape molding and casting techniques were introduced to capture delicate ice features such as ice feathers. The iced airfoil castings were tested in a dry-air wind tunnel. Performance degradation was compared between the four iced models and the clean airfoil. The effect of ice feathers on drag degradation was investigated. Ice feather formation can account for up to 25% of the drag introduced by ice accretion prior to stall. Comparison between the proposed analytical determination method and experimental results from both rotor ice testing and icing wind tunnel testing showed to be satisfactory, ranging from 25% to 6% depending on the icing condition. NomenclatureA c = Accumulation parameter, dimensionless AOA, α = Angle of attack, degree C = Airfoil chord, m Cd = Drag coefficient, dimensionless Cl = Lift coefficient, dimensionless Cm = Pitching moment coefficient, dimensionless d = Characteristic Length, twice the leading edge radius K 0 = Modified inertia parameter, dimensionless LWC = Cloud Liquid Water Content, g/m 3 Ma = Mach number, dimensionless MVD, δ = Water droplet Medium Volume Diameter, µm r = Airfoil leading edge radius, percentage of chord Re = Reynolds number, dimensionless RPM = Rotational speed, revolution per minute T = Static Temperature, K V = Free-stream velocity of air, m/s β 0 = Collection efficiency at stagnation line, dimensionless ΔCd = Drag coefficient increment, dimensionless ρ = Density, kg/m 3 τ = Icing time, minute

Effect of icing roughness on wind turbine power production

“…Cold‐climate regions typically contain a high wind resource, and the higher density air allows for more extraction of power from the wind; however, these cold‐climate regions expose wind turbines to atmospheric icing conditions. Many studies show that wind turbine icing events lead to severe power degradation . This results in significant losses of revenue for wind turbine operators during the winter season and an Annual Energy Production (AEP) loss of up to 20% .…”

“…The driving force behind icing power loss is degradation of airfoil performance due to icing roughness and accumulation. In an icing event, super‐cooled water droplets accrete to wind turbine blades, causing an alteration to both the shape and surface roughness of local blade airfoil sections . Bragg et al .…”

“…Other research on wind turbine performance degradation due to icing includes experimental studies of many types . Similar to numerical studies, some of these experimental studies use ice accretion codes to generate simulated ice shapes for experimental wind tunnel or water‐tunnel testing .…”

“…Once again, roughness must be simulated on these shapes by a generic grit, and the resulting airfoil drag differs depending on the chosen value for the surface roughness height . Other studies experimentally generate actual ice shapes for aerodynamic analysis . The chosen airfoils for these studies (S809 and NACA 64‐415 ) are for stall‐regulated wind turbines at lower operational Reynolds number (Re) than what is typical for a utility‐scale pitch‐controlled wind turbine.…”

Large Eddy Simulation of Airfoil Ice Accretion Aerodynamics

“…This may be be due to calibration bias in the measurements since it also appears for the clean foil (not presented here). 9 Nevertheless it is clear that CFD model results do not reproduce the high-α lift characteristics. )…”