LOAD REDUCTION USING PRESSURE DIFFERENCE ON AIRFOIL FOR CONTROL OF TRAILING EDGE FLAPS

1
LOAD REDUCTION USING PRESSURE DIFFERENCE ON
AIRFOIL FOR CONTROL OF TRAILING EDGE FLAPS

• Peter Bjørn Andersen Ph.D. student, Risø DTU,
  Denmark
• Mac Gaunaa Senior Scientist, Risø DTU, Denmark
• Joachim Christian Heinz M.Sc. (the Amazon,
  Brazil)
• Leonardo Bergami M.Sc. (Monterrey, Mexico)
• EWEC 2009
• Parc Chanot, Marseille,
• France 16 19 March 2009

2
Outline

• Motivation
• 2D Thin Airfoil Model (TAM)
• TAM versus CFD
• TAM versus Wind tunnel experiment
• Results ?P control versus y-control and
  a-control
• Conclusion

3
Motivation

• Main driver of fatigue loading on a wind turbine
  unsteady flow conditions
• (Wind turbulence, Shear, Tower shadow, Yaw/tilt
  misalignment, etc)
• Using blade pitch (Up to 28 load reductions
  reported)
• Not especially fast, harder for flexible blades,
  expensive(far away from where the action
  happens)
• Using local flow control
• Fast and local control possible (close to
  where the action happens)

4
Motivation

• Why not a plain rigid-body trailing edge flap?
• Surface discontinuity triggers stall
• ?
• Noise issues
• Bad L/D leading to loss in
• power production
• Flap losing its potential
• load reducing effect
• ?
• Go for the continuously
• deforming one
• (smooth deformation shape)!

5
Motivation

• Risø, Delft, UC Davis, Sandia
• Commercial interest (Vestas, LM, )

6
Outline

• Motivation
• 2D Thin Airfoil Model (TAM)
• TAM versus CFD
• TAM versus Wind tunnel experiment
• Results ?P control versus y-control and
  a-control
• Conclusion

7
2D Thin Airfoil Model

• Assumptions
• Thin Airfoil ? Camberline
• Potential flow ? Fully attached

• Gaunaas model.
• Unsteady 2D aerodynamic forces and distribution
• Arbitrary motion and camberline deformation.

8
2D Thin Airfoil Model ?P?P(chord position)
9
Outline

• Motivation
• 2D Thin Airfoil Model (TAM)
• TAM versus CFD
• TAM versus Wind tunnel experiment
• Results ?P control versus y-control and
  a-control
• Conclusion

10
TAM versus CFD (Ellipsys)
11
TAM versus CFD using airfoil in springs
12
TAM versus CFD (step change in wind speed)
13
Outline

• Motivation
• 2D Thin Airfoil Model (TAM)
• TAM versus CFD
• TAM versus Wind tunnel experiment
• Results ?P control versus y-control and
  a-control
• Conclusion

14
TAM versus Wind tunnel experiment
15
TAM versus Wind tunnel experiment
VELUX Wind Tunnel Measurements Inflow
velocity 40m/s Chord length (incl. flap)
0.66m
16
TAM versus Wind tunnel experiment
17
TAM versus Wind tunnel experiment, ?P(chord
position at 40)

• Gaunaas TAM predicted a near perfect correlation
  between lift and ?P for 40-50 chord position
  using the Risø B1-18 camberline.

18
Outline

• Motivation
• 2D Thin Airfoil Model (TAM)
• TAM versus CFD
• TAM versus Wind tunnel experiment
• Results ?P control versus y-control and
  a-control
• Conclusion

19
The ?P control
at some sensor chord position
20
The ?P control, advanced version
21
The y control
y, dy
22
The a control
23
The ?P control FY - fatique reduction 60m/s
24
The ?P control (time delay in control)FY -
fatique reduction 60m/s
25
ComparisonFY - fatique reduction 60m/s
26
Conclusions

• Validation with a reimplementation of Theodorsen
  and Garricks method show excellent agreement
• Comparison against 2D CFD aeroelastic model show
  excellent agreement
• Comparison with wind tunnel experiment show good
  agreement
• The pressure sensors should be placed closed to
  the leading edge for the simple ?P control, not
  important for the advanced ?P control but
  knowledge of the rate of torsion is needed in the
  control.
• The a control, ?P control and y control show FY
  fatique reductions well above 70