Comparison of Wind Tunnel Airfoil Performance Data With Wind Turbine Blade Data

Abstract

Horizontal axis wind turbine (HAWT) performance is usually predicted by using wind tunnel airfoil performance data in a blade element momentum theory analysis. This analysis assumes that the rotating blade airfoils will perform as they do in the wind tunnel. However, when stall-regulated HAWT performance is measured in full-scale operation, it is common to find that peak power levels are significantly greater than those predicted. Pitch-controlled rotors experience predictable peak power levels because they do not rely on stall to regulate peak power. This has led to empirical corrections to the stall predictions. Viterna and Corrigan (1981) proposed the most popular version of this correction. But very little insight has been gained into the basic cause of this discrepancy. The National Renewable Energy Laboratory (NREL), funded by the DOE, has conducted the first phase of an experiment which is focused on understanding the basic fluid mechanics of HAWT aerodynamics. Results to date have shown that unsteady aerodynamics exist during all operating conditions and dynamic stall can exist for high yaw angle operation. Stall hysteresis occurs for even small yaw angles and delayed stall is a very persistent reality in all operating conditions. Delayed stall is indicated by a leading edge suction peak which remains attached through angles of attack (AOA) up to 30 degrees. Wind tunnel results show this peak separating from the leading edge at 18 deg AOA. The effect of this anomaly is to raise normal force coefficients and tangent force coefficients for high AOA. Increased tangent forces will directly affect HAWT performance in high wind speed operation. This report describes pressure distribution data resulting from both wind tunnel and HAWT tests. A method of bins is used to average the HAWT data which is compared to the wind tunnel data. The analysis technique and the test set-up for each test are described.