Aeroelasticity in Dynamically Pitching Wind Turbine Airfoils

Transcription

1 Aeroelasticity in Dynamically Pitching Wind Turbine Airfoils Andrew Magstadt, John Strike, Michael Hind, Pourya Nikoueeyan, and Jonathan Naughton Dept. of Mechanical Engineering Wind Energy Research Center University of Wyoming, Laramie, WY

2 Motivation Wind turbines are growing larger Market demands cheaper energy Increased blade size (blades proposed ~ m) Decreased relative stiffness Aeroelasticity becomes major concern Varying inflow conditions produce well-known unsteady aerodynamics Shear & turbulence in Atmospheric Boundary Layer Yawed operating conditions Aeroelastic system comprised of nonlinear components Complex & difficult to understand Tough to numerically model Need for experimental investigation Validation of models Elasticity Aerodynamics Inertia

3 Motivation Validation & Verification Wind Turbine Aeroelastic Validation Roadmap / Aero - Elastic Control 2 - D Static More Complexity 2 - D Dynamic 3 - D Static 3 - D Dynamic 3 - D Rotating W W W Naughton 3 / 11 Higher Fidelity Data

4 Objectives & Approach Research Goals Investigate and characterize the effects of elastic compliance Airfoil Response Aerodynamics Understand how compliance affects the wind turbine system Dynamic Loading Stall Flutter, LCOs? Experimental Means Single d-o-f system considered I b k M All forces ~ Order of Magnitude Driven pitch oscillations produce dynamic stall in wind tunnel spring section designed & characterized and compliant systems contrasted

5 Experimental Setup Mechanical System UW s Low Speed Wind Tunnel Operational range up to m/s.61 x.61 x 1.22 m test section.3% free-stream turbulence Airfoil Encoder Pitching System 24V DC driving motor PID algorithm, flywheel, & PWM maintain constant frequency Cam & push rod for sinusoidal pitch cycles Compliance Section Driving Motor Encoder Test Section Pitching Motor Compliance Variable Spring Stiffness 16.2 N.m/rad to 389 N.m/rad Maximum Allowable Differential Φ max = 4.4 o

6 Experimental Setup Instrumentation Position Two rotary encoders Airfoil Encoder Compliance Section Airfoil Driver System response Test Section Pressure Remotely measure unsteady surface pressures Spatial & temporal distributions Lift & moment coefficients Nd:YAG Lasers Flow-field Dual PIV Vector fields merged & phaseaveraged Structure CCD Cameras Driving Motor Encoder Pitching All 3 necessary for understanding Motor the physics of aeroelastic system!

7 Experimental Setup Dynamic Pressure Distribution x/c x/c x/c Case 9: α = o ± 12 Hz (k =.17), k φ = N. m/rad

8 Experimental Setup Test Cases All cases operated under the following test conditions: Chord, C =.3 m Reynolds Number, Re c = 4.4x Flexural Axis, FA = c/4 Moment of Inertia, I = 6.x -3 kg. m 2 Spring stiffness was arrived at by maximizing the differential angle while oscillating & compliant data taken for each case Cases 2-4 suggested flow structure may start to deviate Focus on Case : α = o ± Hz α = 12 o ± Hz

9 Static Equilibrium Mean AoA shifts Aeroelastic Airfoil Response Inertial & Elastic Interactions More extreme AoAs experienced Peak lag Classic harmonic oscillator Aerodynamic Influence Departure from sinusoidal curve Falling slope is steeper than rising Requires further investigation

10 Dynamic Pressure Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

11 Dynamic Pressure Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

12 Dynamic Pressure Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

13 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

14 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

15 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

16 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

17 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

18 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

19 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

20 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

21 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

22 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

23 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

24 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

25 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

26 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

27 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

28 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

29 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

30 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

31 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

32 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

33 Pressure & Flow-field Case : α = 12 o ± Hz (k =.21), k φ = N. m/rad

34 Integrated Case 1. Fully attached 2. Trailing edge stall begins setting up 3. Trailing edge separation initiates w/ secondary vortex 4. Minor TE stall, with weak secondary vortex. 6. Vortices shed Flow reattachment C l C m Angle of Attack, Integrated Case 1. Fully attached 2. Trailing edge stall begins setting up 3. Trailing edge separation initiates w/ secondary vortex evident 4. TE stall. Additional structure in front of TE vortex. Suction side vortices merge 6. Secondary vortex sheds first, primary follows 7. Remnants of merged vortex 8. Flow reattachment Angle of Attack,

35 Analysis of Integrated C l 1. C m Angle of Attack, Evidence of increase in hysteresis -.3 Increased dynamic loading More extreme AoAs -.4 Angle of Attack, Change in aerodynamic structures Exotic stall observed in compliant case may result from non-sinusoidal pitch frequency Moment increase is more involved Result of asymmetric AoA schedule See paper for in-depth explanation Asymmetric AoA schedule Changes in Instantaneous reduced frequency Airfoil sees higher k momentarily and stalls accordingly

36 Summary Fully Attached Intermediate Stall Deep Stall C m -.2 C l 1. C m -.2 C l -.3, Angle of Attack, Angle ( ) of Attack, C m Angle of Attack, Angle ( ) of Attack, Angle of Attack, Asymmetric AoA No stall Asymmetric AoA Strengthened stall Additional structure High c m lowers AoA max Stalls prior to AoA max Deep stall insensitive to small AoA change

37 Conclusions Coupling of surface and flowfield measurements critical to understanding complex flow Presence of compliance affects: AoA schedule Flow structures Dynamic loading Hysteresis increased Lift & Moment increased High sensitivity to operating conditions α = o ± o α = 12 o ± o Varying inflow may push blade into this region Adverse consequences Fatigue of components Possibility of more complex phenomenon with plunge Flutter, LCOs Demonstrate potential for aerodynamic control Minimize negative aspects, possibly improve performance Large gains from little changes Little (intelligent) effort required

38 Funding for this work Acknowledgements Grant number DESC1261 from the Department of Energy monitored by Timothy J. Fitzsimmons A gift from BP Alternative Energy North America, Inc. Background Photo Credit Barrow Offshore Wind Farm, Irish Sea _wind_turbines.jpg Full Work AIAA Paper Magstadt, A. S., Strike, J. A., Hind, M. D., Nikoueeyan, P., and Naughton, J. W., Compliance Effects in Dynamically Pitching Wind Turbine Airfoils, AIAA Paper , Jun , 43rd Fluid Dynamics Conference, San Diego, CA. Chapter DOI:.214/

39 Thank you. Questions?