Wind Tunnel Tests of the Risø-A1-18, Risø-A1-21 and Risø ... · This report contains 2D...

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

You may not further distribute the material or use it for any profit-making activity or commercial gain

You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Jul 24, 2020

Wind tunnel tests of the Risø-A1-18, Risø-A1-21 and Risø-A1-24 airfoils

Fuglsang, P.; Dahl, K.S.; Antoniou, I.

Publication date:1999

Document VersionPublisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):Fuglsang, P., Dahl, K. S., & Antoniou, I. (1999). Wind tunnel tests of the Risø-A1-18, Risø-A1-21 and Risø-A1-24 airfoils. Denmark. Forskningscenter Risoe. Risoe-R, No. 1112(EN)

https://orbit.dtu.dk/en/publications/ee9ef2b3-4c96-485e-b066-7886b8fc9ce1

Risø-R-1112(EN)

Wind Tunnel Tests of the Risø-A1-18,Risø-A1-21 and Risø-A1-24 Airfoils

Peter Fuglsang, Kristian S. Dahl, Ioannis Antoniou

FINALDRAFT

Risø National Laboratory, Roskilde, DenmarkJune 1999

AbstractThis report contains 2D measurements of the Risø-A1-18, Risø-A1-21 andRisø-A1-24 airfoils. The aerodynamic properties were derived from pressuremeasurements on the airfoil surface and in the wake. The VELUX open jetwind tunnel was used having a background turbulence intensity of 1%, a flowvelocity of 42 m/s and a Reynolds number of 1.6×106. The airfoil sections had achord of 0.60 m and a span of 1.9 m and end plates were used to minimise 3Dflow effects. The measurements comprised both static and dynamic inflowwhere dynamic inflow was obtained by pitching the airfoil in a harmonicmotion. We tested the influence of leading edge roughness, vortex generatorsand Gurney flaps both individually and in combination.For smooth surface conditions, all three airfoils had the desirable properties ofconstant lift curve slope and low drag coefficient until the maximum lift ofabout 1.4 was reached. The Risø-A1-18 airfoil had a smooth post stall whereasthe Risø-A1-21 and Risø-A1-24 airfoils had a significant drop in the liftcoefficient after stall. Test on all airfoil sections mounted with zigzag tapeshowed that the airfoils were insensitive to leading edge roughness. Howeverwith a drop in the maximum lift coefficient to about 1.2. Mounting of deltawing shaped vortex generators and Gurney flaps showed that there was roomfor a significant increase in the maximum lift coefficient, which was increasedto 1.90 for Risø-A1-24 with vortex generators located at 15% chord. Thecombination of vortex generators and Gurney flaps increased the maximum liftcoefficient to about 2.0.

The Danish Energy Agency funded the present work under the contract,ENS-1363/98-0038.

The work was carried out within the EFP 98 project, ‘Experimental verificationof the Risø-A airfoil family’ with the following partners:

• LM Glasfiber A/S. Peter Grabau,Peter Hansen,Steen Z. Dyremose,Niels Brønnum

• Vestas Wind Systems A/S Kaj Morbech,Mark Slot

• The Techical University of Denmark Stig Øye• Risø National Laboratory Peter Fuglsang

Ioannis AntoniouKristian S. Dahl

Risø-R-1112

ISBN: 87-550-2538-2ISBN: 87-550-2539-0 (internet)

ISSN: 0106-2840

Information Service Department, Risø, 1999

Risø-R-1112(EN) 3

Contents

Nomenclature 5

1 Introduction 6

2 Experimental set-up 7

2.1 Testing facility 72.2 Wind tunnel boundary corrections 92.3 Wind tunnel flow conditions 102.4 Calculation methods 10

3 Airfoil sections and aerodynamic devices 12

3.1 Airfoil sections 123.2 Vortex generators 133.3 Gurney flaps 143.4 Leading edge roughness 14

4 Results 15

4.1 Testing conditions 154.2 Numerical calculations 15

5 Results for Risø-A1-18 16

5.1 Risø-A1-18 Smooth flow (run017) 165.2 Risø-A1-18 LER (run016) 205.3 Risø-A1-18 VGs (run014, 020) 225.4 Risø-A1-18 VGs at 0.2 (run014, 015) 255.5 Risø-A1-18 VGs at 0.25 (run020, 021) 275.6 Risø-A1-18 Dynamic stall (run022) 29


6.1 Risø-A1-21 Smooth flow (run025) 336.2 Risø-A1-21 LER (run005) 376.3 Risø-A1-21 VGs (run003, 024) 396.4 Risø-A1-21 VGs at 0.2 (run003, 004) 426.5 Risø-A1-21 VGs at 0.25 (run023, 024) 446.6 Risø-A1-21 Dynamic stall (run026) 46


7.1 Risø-A1-24 Smooth flow (run032) 507.2 Risø-A1-24 LER (run029) 547.3 Risø-A1-24 VGs (run034, 027, 031, 081) 567.4 Risø-A1-24 GFs (run039,run040) 597.5 Risø-A1-24 VGs at 0.1 (run064, 065) 617.6 Risø-A1-24 VGs at 0.15 (run034,035) 637.7 Risø-A1-24 VGs at 0.2 (run027,028) 657.8 Risø-A1-24 VGs at 0.25 (run030,031) 677.9 Risø-A1-24 VGs at 0.30 (run081) 697.10 Risø-A1-24 GFs of 1% (run037,040) 717.11 Risø-A1-24 GFs of 2% (run038,039) 737.12 Risø-A1-24 VGs at 0.10 double spacing (run067,066) 75

Risø-R-1112(EN)4

7.13 Risø-A1-24 VGs at 0.15, GFs of 1%, LER (run036) 777.14 Risø-A1-24 Dynamic stall (run043) 79

8 Discussion 83

8.1 Risø-A1-18 838.2 Risø-A1-21 858.3 Risø-A1-24 868.4 Summary 87

9 Conclusions 89

References 91

A Measurement survey 92

A.1 List of symbols 92A.2 Measurement types 92A.3 Data file naming convention 95A.4 Data file formats 95A.5 Performed measurements 100

Risø-R-1112(EN) 5

Nomenclature

c [m] Airfoil chordh [m] Jet heightk Reduced frequency∆p [Pa/m] Pressure lossp [Pa] Static pressurepo [Pa] Total pressure headq [Pa] Dynamic pressures Airfoil surface co-ordinatet [s] Pitch motion timex Co-ordinate in chord directiony Wake rake vertical co-ordinate, airfoil vertical co-

ordinate

Α [°] Pitch motion amplitudeCD Drag coefficientCL Lift coefficientCM Moment coefficientCN Normal force coefficientCP Airfoil pressure coefficientCT Tangential force coefficientRe Reynolds numberT [°C] Air temperatureV [m/s] Velocity

α [rad] [°] Angle of attackε Speed-up factorρ [kg/m3] Air densityω [rad/s] Pitch motion angular velocity

Subscripts1-3 Pitot tube measurementa Airfoil section measurementatm Atmospheric valuej Jet outlet measurementm Mean valuemin Minimum valuemax Maximum valuep Pressure measurementt Measured value (uncorrected)w Wake rake measurement∞ Free stream reference for normalisation of airfoil forces

Risø-R-1112(EN)6

1 Introduction

This report concerns 2D wind tunnel measurements of the Risø-A1-18,Risø-A1-21 and Risø-A1-24 airfoils. These airfoils are members of an airfoilfamily that was recently developed by Risø National Laboratory for use onwind turbines [1]. The airfoils were specially designed for wind turbines witheither stall, active stall or pitch regulation. The operational design Reynoldsnumbers were around Re = 3.0 million depending on the relative thickness.This corresponds to a rotor size of around 600 kW. The measurements werecarried out in the VELUX wind tunnel, which has an open test section with abackground turbulence level of 1% and a maximum flow velocity of 42 m/s. Alltests were carried out at the highest possible Reynolds number of 1.6 million.The angle of attack range was between -5° and 30°. Pressure distributionmeasurements were taken on the airfoil section together with wake rakepressure measurements. The testing facility is described in detail in Fuglsang etal., 1998 [2].

The test matrix included:• Steady and quasi-steady inflow measurements where mean values were

obtained for the airfoil aerodynamic coefficients. The angle of attack waschanged continuously at an average rate around 0.3°/s. Alternatively theangle of attack was changed in steps of 2° and a 20 s duration time serieswas obtained for each angle of attack.

• Dynamic inflow measurements with the airfoil in pitching motion atamplitudes around ±2° and reduced frequencies around 0.1. The hysteresiseffects on the aerodynamic coefficients were derived.

The airfoils were tested under the following configurations:• Smooth surface, referred to as, ‘smooth flow’.• Vortex generators on the suction side to delay separation and increase the

maximum lift coefficient referred to as, ‘VG’.• Gurney flaps on the pressure side trailing edge to increase the maximum lift

coefficient referred to as, ‘GF’.• Leading edge roughness to simulate the change of the aerodynamic

coefficients from dirt and dust accumulation referred to as, ‘LER’.• Different combinations of vortex generators, Gurney flaps and leading edge

roughness.

Risø-R-1112(EN) 7

2 Experimental set-up

The experimental set-up is briefly described in this chapter. A more completedescription can be found in Fuglsang et al., 1998 [2].

2.1 Testing facilityThe VELUX wind tunnel is of the closed return type with an open test sectionwith a cross section of 7.5×7.5 m and a length of 10.5 m, Figure 2-1. The crosssection of the jet blowing into the test section is 3.4×3.4 m. The maximum flowvelocity is 42 m/s.

10 50

7.504.00

3.40

4.00

.50

.50

.75

.75

2.651.75

Pitot 2

Pitot 1Pitot 3

.85

Airfoil section

Wake rake

Figure 2-1 The wind tunnel test section with the test stand seen in a top view.with the flow coming from the left.

A test stand was built for 2D airfoil testing, Figure 2-2. The test stand wasinserted in the tunnel test section. The airfoil section with a span of 1.9 m and achord of 0.6 m was mounted 1.7 m from the tunnel floor and 3.2 m from thenozzle outlet. End plates were fixed to the stand at the ends of the airfoilsection to limit 3d effects.

Three Pitot tubes measured static and total pressure at different locations in thetest section, Figure 2-1. These Pitot tubes were used to measure the wind tunnelreference pressures and to estimate the turbulence level and the stability of thewind tunnel flow.

Quasi-steady measurements at continuously varying angles of attack as well asdynamic inflow measurements were possible. Dynamic inflow was obtained bypitching the airfoil section at different reduced frequencies up to k = 0.15 andamplitudes between ±1° < A < ±5° with the pitch axis located at x/c = 0.40, seesection 2.4.

Risø-R-1112(EN)8

Figure 2-2 The test section with the test stand and the wake rake downstream ofthe airfoil section.

The wake rake consisted of 53 total pressure probes and five static tubes. Thevertical span was 0.456 m, Figure 2-3. The distance between the airfoil trailingedge and the wake rake was 0.7 airfoil chords and the centre of the wake rakewas placed at the height of the trailing edge at 0° incidence and behind thecentre line of the airfoil section. The rake was not traversed in the horizontal orthe vertical directions.

Figure 2-3 The wake rake seen from the side in front of an endplate.

The HyScan 2000 data acquisition system from Scanivalve Corp. was used.Two ZOC33 pressure-scanning modules recorded the pressure signals. For theairfoil surface pressures, 40 psi and 24 2.5psi range sensors were used. For thewake rake and the Pitot tubes, 10´´ H20 sensors were used. The ZOC modulefor the airfoil pressures was mounted on the test stand side just outside theairfoil section. Equal length tubes were lead from the airfoil section through ahollow axis to the pressure module. The pressure module used for the wake andthe Pitot tube measurements was placed on the floor next to the wake rake. AZOCEIM16 module was used for the acquisition of the electrical signals.

Risø-R-1112(EN) 9

A total of 134 signals were measured by the data acquisition system during themeasurement campaigns:

• 64 airfoil surface static pressures, pa(s)• 5 wake rake static pressures, pw(y)• 53 wake rake total pressures, pow(y)• 3 Pitot tube static pressures, p1-3

• 3 Pitot tube total pressures, po1-3

• Angle of attack, α• Air temperature, T• Atmospheric pressure, patm

• 2 strain gauges for recording shaft bending corresponding to the lift anddrag forces experienced by the airfoil section.

• Electric motor frequency

2.2 Wind tunnel boundary correctionsWind tunnel corrections should be applied for streamline curvature and down-wash. Horizontal buoyancy, solid and wake blockage could on the other handbe neglected because the test section configuration corresponds to an open jet,which is free to expand, Ray and Pope, 1984 [3]. The application of windtunnel boundary corrections for the VELUX wind tunnel was verified inFuglsang et al., 1998 [2].

Streamline curvature is introduced to the flow, especially in the case of opentest sections. Solid walls do not bound the flow, which is then free to divergedownstream of the airfoil section. The curvature of the flow induces drag andinfluences the effective angle of attack over the airfoil. In the case of theVELUX tunnel, the presence of the floor close to the jet bottom boundary willinfluence the streamline curvature and will introduce uncertainty on the windtunnel corrections. This influence was assumed to be negligible and the appliedcorrections for streamline curvature do not account for it.

Down-wash is introduced to the flow when the jet dimensions exceed the airfoilsection span. The airfoil section corresponds to a finite wing and trailingvortices appear at the ends of the span although reduced by the end plates. Thetrailing vorticity induces a down-wash velocity in the case of positive liftcoefficient. Due to the down wash the angle of attack is reduced and additionaldrag is induced.

Both down-wash and streamline curvature result in a change in the angle ofattack due to the induction of a velocity normal to the flow direction and theairfoil section. It is assumed in this case that down-wash is insignificantcompared with streamline curvature because of the presence of end plates.

For the correction of streamline curvature, the method of Brooks and Marcolini,1984 [4] was used.

Risø-R-1112(EN)10

22

48

⋅=

h

cπσ

The corrected angle of attack, α, is found from:

(2-1)

Where

(2-2)

The drag coefficient, CD, is calculated from:

(2-3)

The moment coefficient, CM, is obtained:

(2-4)

For details see Fuglsang et al., 1998 [2].

2.3 Wind tunnel flow conditionsIn Fuglsang et al., 1998 [2] the wind tunnel flow conditions are investigatedand it is found that:• The turbulence intensity at the test section inlet is 1%.• Between the inlet and the airfoil section, there is a speed-up of, εj-a = 6.9%,

and a static pressure drop of ∆pj-a = 15 Pa/m.

The wind tunnel references for static, p∞ and total pressures, po∞ were derivedfrom Pitot 1 measurements, Figure 2-1. The flow acceleration between Pitot 1and the airfoil section, ε1-∞ = 5.9% and the static pressure drop between Pitot 1and the airfoil section, ∆p1-∞ = 15 Pa/m were determined in Fuglsang et al.,1998 [2] and they are taken into account at the calculation of p∞ and po∞.

2.4 Calculation methodsThe airfoil pressure coefficient, Cp(s), around the airfoil surface, s, is calculatedfrom:

(2-5)

Where

(2-6)

The normal force coefficient, CN, and the tangential force coefficient, CT, arefound from integration of the CP(s) distribution along the x- and y-axis as seen

LLDtD CCCC

−+=

πσ3

LMtM CCC2

σ−=

( ) [ ]radCCC MtLLt 423

πσ

πσ

πσαα −−−=

( )∞

∞−=q

pspsC a

p

)(

∞∞∞ −= ppq o

Risø-R-1112(EN) 11

in Figure 2-4. The airfoil lift coefficient, CL, and drag coefficient, CD, are foundby resolving CN and CT perpendicular to and parallel with the oncoming flow:

( ) ( ) TNL CCC αα sincos +=

(2-7)

( ) ( ) NTD CCC αα sincos +−=

The moment coefficient, CM, is found from integration of CP(s) at x/c = 0.25.

Figure 2-4 Sign convention for aerodynamic coefficients.

The total airfoil drag is the sum of skin friction and pressure drag. By assuminga control surface, which surrounds the airfoil section, the total drag can becalculated from the balance of the momentum flux entering and exiting thecontrol surface. The momentum profile entering is assumed uniform and iscalculated from the wind tunnel free stream reference pressures. Themomentum profile exiting is calculated from the pressures measured by thewake rake.

Assuming that the flow is 2D, the total wake drag coefficient, CDw, is calculatedfrom Rae and Pope, 1984 [3]:

(2-8)

In the analysis of dynamic loads, while the airfoil is in pitching motion, thepitching motion is described by the equation:

(2-9)

The pitching motion is related to the reduced frequency:

(2-10)

∫

−−⋅

−=

∞∞

max

min

)()(1

)()(2y

y

ooDw dy

q

ypyp

q

ypyp

cC

( ) mtA αωα += sin

∞

=V

ck

2

ω

Risø-R-1112(EN)12

3 Airfoil sections and aerodynamicdevices

The tested airfoils were the Risø-A1-18, Risø-A1-21 and Risø-A1-24 airfoilsfrom Fuglsang and Dahl, 1999 [1].

3.1 Airfoil sectionsFor all airfoil sections, the span was 1.9 m and the chord was 0.60 m. VestasWind Systems A/S manufactured the Risø-A1-18 model and LM Glasfiber A/Smanufactured the Risø-A1-21 and Risø-A1-24 models. Risø carried out theinstrumentation of pressure tabs, end pieces and strain gauges. Each model wasmanufactured in two pieces as an upper and a lower shell to facilitateinstrumentation. The models were made of GRP in moulds. The pressure tapswere holes drilled directly in the model surface with the exception of theleading and trailing edges where tubes were installed through the modelsurface, flush with the surface. Inside the model metal tubes were mountedparallel to the drilled holes and flexible plastic tubes were connected to themetal tubes. When the instrumentation was completed the two shells wereassembled. The pressure tubes were taken outside of the model through ahollow axis at one side of the airfoil.

The airfoil sections were equipped with 62 pressure taps of 0.5 mm innerdiameter in the centre line region. The taps were placed along the chord at thecentre line of the model in a staggered alignment to minimise disturbances fromupstream taps. Additional taps were drilled close to the centre line as a back upto taps at important positions, e.g., the leading and trailing edges, and in orderto allow measurements away from the centre line.

The position of the pressure taps on the model was decided by looking on thetheoretical pressure distributions derived by numerical calculations. Thedistribution of the pressure taps reflected the expected pressure gradients andthe tap spacing was dense at leading edge. There was higher concentration onthe upper surface compared to the lower surface. After the model waspermanently assembled the model dimensions and the tap positions werechecked for compliance with the theoretical ones, with the help of a CNS flat-bed machine.

Only minor differences were found between the theoretical and the measuredco-ordinates. For all airfoil models, it was concluded that this would not resultin significant errors in the pressure distribution and in the derivation ofaerodynamic loads.

Risø-R-1112(EN) 13

3.2 Vortex generatorsVortex generators (VGs) are often used at the inner part of wind turbine bladeslocated on the blade suction side between 10% to 30% of the chord countedfrom the leading edge. They increase the maximum lift coefficient by delayingseparation on the airfoil suction side to higher incidences. At the same time,they increase the drag coefficient.

A parametric study was conducted where VGs of height 6 mm were used atdifferent chordwise locations. However not all combinations were tried for allthe airfoils. The design of the VGs followed the guide lines from Hoerner andBorst, 1975 [5] and was similar to those used for numerous airfoil tests byTimmer, 1992 [6], at Delft University.

Figure 3-1 shows the shapes and dimensions of the used VGs. They have aheight of 6 mm a length of 18 mm. The angles relative to the chordwisedirection are ±19.5°. The leading edge spacing between two VGs is 10 mm andthe distance between two consecutive pairs is 25 mm. They are of the Deltawing type with a shape of orthogonal triangles and they are placed with theirright-angle perpendicular to the airfoil surface and their height increasestowards the trailing edge. The presence of the VGs results in the formation ofcounter-rotating vortices, which transfer high momentum fluid down to theairfoil surface and thus delay separation. To achieve this VGs are arranged inpairs at equal and opposite angles relative to the chord of the blade.

The VGs were constructed from 0.2 mm thick stainless steel. Each VG was cutout and bent perpendicular to the surface. The VGs were glued on the airfoilmodel surface separately. The thickness of the gluing surface will slightlydisturb the measurements since the flow has to enforce the edge of the gluingsurface. In particular the drag coefficient at low angles of attack will beincreased.

25 10

19.5˚19.5˚ 6

18.97

Side view

18

Figure 3-1 Vortex generators of height 6 mm, length 18 mm.

Risø-R-1112(EN)14

3.3 Gurney flapsA Gurney flap (GF) is a small flap rigidly attached to the trailing edge of thepressure side of the airfoil. The intended purpose of a GF is to improve theairfoil performance by increasing the lift coefficient without introducing acommensurate increase in drag coefficient.

A parametric study was conducted for the Risø-A1-24 airfoil where GFs ofheight 1% and 2% of the chord respectively were used.

The GFs were constructed from 0.2 mm thick strip, bent in a 90° angle so thatthey were mounted perpendicular to the pressure side trailing edge surface.

3.4 Leading edge roughnessTrip tape was mounted to the airfoil model surface to simulate the effects fromleading edge roughness (LER). LER appears when dirt, bugs or soil areaccumulated on the wind turbine blades in dirty environments.

The used trip tape was originally intended for use on gliders and weremanufactured as fibre enforced plastic tape that was glued to the airfoil modelsurface. The trip tape was mounted at x/c = 0.05 on the suction side and at x/c =0.10 on the pressure side.

Figure 3-2 shows the used 90° zigzag trip tape with a 90° angle, a width of 3mm and a thickness of 0.35 mm.

90°3

11

Thic kne ss 0.35 m m

Figure 3-2 Trip tape with 90° zigzag of 3 mm width and 0.35 mm thickness.

Risø-R-1112(EN) 15

4 Results

All shown results were corrected for wind tunnel effects and the aerodynamicforces were referenced to the wind tunnel free stream flow by use of Pitot 1taking into account corrections for speed-up and pressure loss.

The measurements for each airfoil are reported in Chapter 5 – 7 and themeasurements are discussed in Chapter 8.

The different types of conducted measurements are described in Appendix A.

4.1 Testing conditionsThe testing conditions are shown in Table 4-1.

Table 4-1 Testing conditions

Airfoil chord c = 0.60 mFlow velocity v = 42 m/sReynolds number Re = 1.6×106

Dynamic inflowAngular velocity ω = 12.9 rad/sReduced frequency k = 0.092Amplitude 1.4o < A < 2.0o

4.2 Numerical calculationsThe measurements were compared with numerical calculations when it waspossible. The Ellipsys2D Navier-Stokes code, Sørensen, 1995 [7], with the k-ωSST turbulence model, Menter, 1993 [8], was used for turbulent flowcalculations. Free transition was modelled using the Michel transition criteria,Michel, 1952 [9].

The leading edge roughness measurements were compared with numericalcalculations. The Ellipsys2D Navier-Stokes code was used with turbulent flowon the entire airfoil to simulate leading edge roughness.

Measurements with vortex generators and Gurney flap were not compared withnumerical calculations.

Risø-R-1112(EN)16

5 Results for Risø-A1-18

5.1 Risø-A1-18 Smooth flow (run017)

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

Smooth measurement, α=0o




-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





Figure 5-1 CP at different angles of attack for Risø-A1-18 smooth measurement(run017).

Risø-R-1112(EN) 17

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c


EllipSys2D TransitionEllipSys2D Turbulent

Figure 5-2 CP at α = 2° for Risø-A1-18 smooth measurement compared withEllipSys2D calculations with transition modeling and turbulent flowrespectively (run017).

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




Risø-R-1112(EN)18

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α

Smooth measurementEllipSys2D TransitionEllipSys2D Turbulent

Figure 5-4 CL-CD for Risø-A1-18 smooth measurement compared withEllipSys2D calculations with transition modeling and turbulent flowrespectively (run017).

Risø-R-1112(EN) 19

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α

Smooth measurement

Figure 5-5 CL, CD and CM for Risø-A1-18 smooth measurement compared withEllipSys2D calculations with transition modeling and turbulent flowrespectively (run017).

Risø-R-1112(EN)20

5.2 Risø-A1-18 LER (run016)

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

LER measurement, α=8o


EllipSys2D Turbulent

Figure 5-6 CP at α = 8° for Risø-A1-18 LER measurement compared withsmooth measurement and EllipSys2D calculations with turbulent flow (run016).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α

LER measurementSmooth measurementEllipSys2D Turbulent

Figure 5-7 CL-CD for Risø-A1-18 LER measurement compared with smoothmeasurement and EllipSys2D calculations with turbulent flow (run016).

Risø-R-1112(EN) 21

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α

LER measurementSmooth measurement

Figure 5-8 CL, CD and CM for Risø-A1-18 LER measurement compared withsmooth measurement and EllipSys2D calculations with turbulent flow (run016).

Risø-R-1112(EN)22

5.3 Risø-A1-18 VGs (run014, 020)-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

VG 0.25 measurement, α=10o



Figure 5-9 CP at α = 10o for Risø-A1-18 VG measurements compared withsmooth measurement (run014,020)

-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




Figure 5-10 Cp at maximum CL for Risø-A1-18 VG measurements comparedwith smooth measurement (run014,020).

Risø-R-1112(EN) 23

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25

α

VG 0.25 measurementVG 0.20 measurementSmooth measurement

Figure 5-11 CL-CD for Risø-A1-18 VG measurements compared with smoothmeasurement (run014,020).

Risø-R-1112(EN)24

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 5-12 CL, CD and CM for Risø-A1-18 VG measurements compared withsmooth measurement (run014, 020).

Risø-R-1112(EN) 25

5.4 Risø-A1-18 VGs at 0.2 (run014, 015)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

VG 0.20 LER measurement, α=17o



Figure 5-13 Cp at maximum CL for Risø-A1-18 VG 20% Smooth and LERmeasurement compared with smooth measurements (run014,015).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α

VG 0.20 LER measurementVG 0.20 measurementSmooth measurement

Figure 5-14 CL-CD for Risø-A1-18 VG 20% Smooth and LER measurementcompared with smooth measurements (run014,015).

Risø-R-1112(EN)26

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 5-15 CL, CD and CM for Risø-A1-18 VG 20% Smooth and LERmeasurement compared with smooth measurements (run014,015).

Risø-R-1112(EN) 27

5.5 Risø-A1-18 VGs at 0.25 (run020, 021)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α



Risø-R-1112(EN)28

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN) 29

5.6 Risø-A1-18 Dynamic stall (run022)

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α

Smooth measurementk = 0.092, 1.4o< A < 2.0o

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 5-19 CL, CD and CM hysteresis loops for Risø-A1-18 smoothmeasurement at k = 0.092, A between 1.4° and 2.0° (run022).

Risø-R-1112(EN)30

0.2

0.4

0.6

0.8

-2 -1 0 1 2C

Lα

k = 0.092, αm = 0.0o, A = 1.4o

0.6

0.8

1

1.2

2 3 4 5 6

CL

α

k = 0.092, αm = 3.8o, A = 1.4o

1

1.2

1.4

1.6

7 8 9 10

CL

α

k = 0.092, αm = 8.5o, A = 1.5o

1.2

1.4

1.6

1.8

9 10 11 12

CL

α

k = 0.092, αm = 10.4o, A = 1.6o

1.2

1.4

1.6

1.8

10 11 12 13

CL

α

k = 0.092, αm = 11.4o, A = 1.9o

1.2

1.4

1.6

1.8

11 12 13 14

CL

α

k = 0.092, αm = 12.5o, A = 1.9o

1

1.2

1.4

1.6

1.8

11 12 13 14 15

CL

α

k = 0.092, αm = 13.2o, A = 2.0o

1

1.2

1.4

1.6

13 14 15 16

CL

α

k = 0.092, αm = 14.7o, A = 2.1o

1

1.2

1.4

1.6

15 16 17 18 19

CL

α

k = 0.092, αm = 16.8o, A = 2.1o

1

1.2

1.4

1.6

17 18 19 20 21

CL

α

k = 0.092, αm = 18.8o, A = 2.1o

Figure 5-20 CL hysteresis loops for Risø-A1-18 smooth measurement at k =0.092, A between 1.4° and 2.0° (run022).

Risø-R-1112(EN) 31

-0.05

-0.025

0

0.025

0.05

0.075

-2 -1 0 1 2

CD

α

k = 0.092, αm = 0.0o, A = 1.4o

-0.05

-0.025

0

0.025

0.05

0.075

2 3 4 5 6

CD

α

k = 0.092, αm = 3.8o, A = 1.4o

-0.05

-0.025

0

0.025

0.05

0.075

7 8 9 10

CD

α

k = 0.092, αm = 8.5o, A = 1.5o

-0.025

0

0.025

0.05

0.075

0.1

9 10 11 12

CD

α

k = 0.092, αm = 10.4o, A = 1.6o

0

0.025

0.05

0.075

0.1

0.125

10 11 12 13

CD

α

k = 0.092, αm = 11.4o, A = 1.9o

0

0.025

0.05

0.075

0.1

0.125

11 12 13 14

CD

α

k = 0.092, αm = 12.5o, A = 1.9o

0.025

0.05

0.075

0.1

0.125

0.15

11 12 13 14 15

CD

α

k = 0.092, αm = 13.2o, A = 2.0o

0.05

0.075

0.1

0.125

0.15

0.175

13 14 15 16

CD

α

k = 0.092, αm = 14.7o, A = 2.1o

0.1

0.125

0.15

0.175

0.2

0.225

15 16 17 18 19

CD

α

k = 0.092, αm = 16.8o, A = 2.1o

0.125

0.15

0.175

0.2

0.225

0.25

17 18 19 20 21

CD

α

k = 0.092, αm = 18.8o, A = 2.1o

Figure 5-21 CD hysteresis loops for Risø-A1-18 smooth measurement at k =0.092, A between 1.4° and 2.0° (run022).

Risø-R-1112(EN)32

-0.1

-0.08

-0.06

-0.04

-0.02

-2 -1 0 1 2C

Mα

k = 0.092, αm = 0.0o, A = 1.4o

-0.12

-0.1

-0.08

-0.06

-0.04

2 3 4 5 6

CM

α

k = 0.092, αm = 3.8o, A = 1.4o

-0.12

-0.1

-0.08

-0.06

-0.04

7 8 9 10

CM

α

k = 0.092, αm = 8.5o, A = 1.5o

-0.12

-0.1

-0.08

-0.06

-0.04

9 10 11 12

CM

α

k = 0.092, αm = 10.4o, A = 1.6o

-0.14

-0.12

-0.1

-0.08

-0.06

10 11 12 13

CM

α

k = 0.092, αm = 11.4o, A = 1.9o

-0.14

-0.12

-0.1

-0.08

-0.06

11 12 13 14

CM

α

k = 0.092, αm = 12.5o, A = 1.9o

-0.14

-0.12

-0.1

-0.08

-0.06

11 12 13 14 15

CM

α

k = 0.092, αm = 13.2o, A = 2.0o

-0.14

-0.12

-0.1

-0.08

-0.06

13 14 15 16

CM

α

k = 0.092, αm = 14.7o, A = 2.1o

-0.14

-0.12

-0.1

-0.08

-0.06

15 16 17 18 19

CM

α

k = 0.092, αm = 16.8o, A = 2.1o

-0.14

-0.12

-0.1

-0.08

-0.06

17 18 19 20 21

CM

α

k = 0.092, αm = 18.8o, A = 2.1o

Figure 5-22 CM hysteresis loops for Risø-A1-18 smooth measurement at k =0.092, A between 1.4° and 2.0° (run022).

Risø-R-1112(EN) 33



-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c






Risø-R-1112(EN)34

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




Risø-R-1112(EN) 35

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α



Risø-R-1112(EN)36

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α

Smooth measurement


Risø-R-1112(EN) 37

6.2 Risø-A1-21 LER (run005)-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α



Risø-R-1112(EN)38

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 6-8 CL, CD and CM for Risø-A1-21 LER measurement compared withsmooth measurement and EllipSys2D calculations with turbulent flowrespectively (run005).

Risø-R-1112(EN) 39

6.3 Risø-A1-21 VGs (run003, 024)-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c


VG 0.20 measurement, α=10.5o


Figure 6-9 CP at α = 10o for Risø-A1-21 VG measurements compared withsmooth measurement (run003, 024).

-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




Figure 6-10 Cp at maximum CL for Risø-A1-21 VG measurements comparedwith smooth measurement (run003, 024).

Risø-R-1112(EN)40

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25

α


Figure 6-11 CL-CD for Risø-A1-21 VG measurements compared with smoothmeasurement (run003, 024).

Risø-R-1112(EN) 41

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 6-12 CL, CD and CM for Risø-A1-21 VG measurements compared withsmooth measurement (run003, 024).

Risø-R-1112(EN)42

6.4 Risø-A1-21 VGs at 0.2 (run003, 004)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α



Risø-R-1112(EN) 43

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 6-15 CL, CD and CM for Risø-A1-21 VG 20% Smooth and LERmeasurement compared with smooth measurements (run003 004).

Risø-R-1112(EN)44

6.5 Risø-A1-21 VGs at 0.25 (run023, 024)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α


Figure 6-17 CL,-CD for Risø-A1-21 VG 25% Smooth and LER measurementcompared with smooth measurements (run023,024).

Risø-R-1112(EN) 45

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN)46


-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


-0.1

0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN) 47

0

0.2

0.4

0.6

-2 -1 0 1

CL

α

k = 0.092, αm = -0.6o, A = 1.4o

0.6

0.8

1

1.2

2 3 4 5

CL

α

k = 0.092, αm = 3.3o, A = 1.4o

1

1.2

1.4

1.6

6 7 8 9 10

CL

α

k = 0.092, αm = 8.1o, A = 1.5o

1.2

1.4

1.6

1.8

8 9 10 11 12

CL

α

k = 0.092, αm = 10.0o, A = 1.6o

1

1.2

1.4

1.6

11 12 13 14

CL

α

k = 0.092, αm = 12.5o, A = 2.1o

0.8

1

1.2

1.4

13 14 15 16 17

CL

α

k = 0.092, αm = 14.9o, A = 1.9o

0.8

1

1.2

1.4

16 17 18 19

CL

α

k = 0.092, αm = 17.6o, A = 1.8o

0.8

1

1.2

1.4

18 19 20 21

CL

α

k = 0.092, αm = 19.6o, A = 1.8o

0.8

1

1.2

1.4

22 23 24 25 26

CL

α

k = 0.092, αm = 24.2o, A = 1.9o


Risø-R-1112(EN)48

-0.05

-0.025

0

0.025

0.05

0.075

-2 -1 0 1C

Dα

k = 0.092, αm = -0.6o, A = 1.4o

-0.05

-0.025

0

0.025

0.05

0.075

2 3 4 5

CD

α

k = 0.092, αm = 3.3o, A = 1.4o

-0.025

0

0.025

0.05

0.075

0.1

6 7 8 9 10

CD

α

k = 0.092, αm = 8.1o, A = 1.5o

-0.025

0

0.025

0.05

0.075

0.1

8 9 10 11 12

CD

α

k = 0.092, αm = 10.0o, A = 1.6o

0.05

0.075

0.1

0.125

0.15

0.175

11 12 13 14

CD

α

k = 0.092, αm = 12.5o, A = 2.1o

0.075

0.1

0.125

0.15

0.175

0.2

13 14 15 16 17

CD

α

k = 0.092, αm = 14.9o, A = 1.9o

0.1

0.125

0.15

0.175

0.2

0.225

0.25

16 17 18 19

CD

α

k = 0.092, αm = 17.6o, A = 1.8o

0.125

0.15

0.175

0.2

0.225

0.25

18 19 20 21

CD

α

k = 0.092, αm = 19.6o, A = 1.8o

0.2

0.225

0.25

0.275

0.3

0.325

22 23 24 25 26

CD

α

k = 0.092, αm = 24.2o, A = 1.9o


Risø-R-1112(EN) 49

-0.1

-0.08

-0.06

-0.04

-0.02

-2 -1 0 1

CM

α

k = 0.092, αm = -0.6o, A = 1.4o

-0.12

-0.1

-0.08

-0.06

-0.04

2 3 4 5

CM

α

k = 0.092, αm = 3.3o, A = 1.4o

-0.12

-0.1

-0.08

-0.06

-0.04

6 7 8 9 10

CM

α

k = 0.092, αm = 8.1o, A = 1.5o

-0.14

-0.12

-0.1

-0.08

-0.06

8 9 10 11 12

CM

α

k = 0.092, αm = 10.0o, A = 1.6o

-0.16

-0.14

-0.12

-0.1

-0.08

11 12 13 14

CM

α

k = 0.092, αm = 12.5o, A = 2.1o

-0.14

-0.12

-0.1

-0.08

-0.06

13 14 15 16 17

CM

α

k = 0.092, αm = 14.9o, A = 1.9o

-0.14

-0.12

-0.1

-0.08

-0.06

16 17 18 19

CM

α

k = 0.092, αm = 17.6o, A = 1.8o

-0.14

-0.12

-0.1

-0.08

-0.06

18 19 20 21

CM

α

k = 0.092, αm = 19.6o, A = 1.8o

-0.16

-0.14

-0.12

-0.1

-0.08

22 23 24 25 26

CM

α

k = 0.092, αm = 24.2o, A = 1.9o


Risø-R-1112(EN)50



-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c






Risø-R-1112(EN) 51

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




Risø-R-1112(EN)52

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α



Risø-R-1112(EN) 53

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α

Smooth measurement


Risø-R-1112(EN)54

7.2 Risø-A1-24 LER (run029)-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α



Risø-R-1112(EN) 55

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 7-8 CL, CD and CM for Risø-A1-24 LER measurement compared withsmooth measurement and EllipSys2D calculations with turbulent flow (run029).

Risø-R-1112(EN)56

7.3 Risø-A1-24 VGs (run034, 027, 031, 081)-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c






Figure 7-9 CP at α = 10o for Risø-A1-24 VG measurements compared withsmooth measurement (run081,031,027,034).

-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c






Figure 7-10 Cp at maximum CL for Risø-A1-24 VG measurements comparedwith smooth measurement (run081,031,027,034).

Risø-R-1112(EN) 57

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25

α

VG 0.30 measurementVG 0.25 measurementVG 0.20 measurementVG 0.15 measurementSmooth measurement

Figure 7-11 CL,-CD for Risø-A1-24 VG measurements compared with smoothmeasurement (run081,031,027,034).

Risø-R-1112(EN)58

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 7-12 CL, CD and CM for Risø-A1-24 VG measurements compared withsmooth measurement (run081,031,027,034).

Risø-R-1112(EN) 59

7.4 Risø-A1-24 GFs (run039,run040)-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

GF 1% measurement, α=9o



Figure 7-13 Cp at maximum CL for Risø-A1-24 GF measurements comparedwith smooth measurement (run040,039).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05

CL

CD

-10 -5 0 5 10 15 20

α

GF 1% measurementGF 2% measurement

Smooth measurement

Figure 7-14 CL-CD for Risø-A1-24 GF measurements compared with smoothmeasurement (run040,039).

Risø-R-1112(EN)60

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


Smooth measurement

0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


Smooth measurement

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Smooth measurement

Figure 7-15 CL, CD and CM for Risø-A1-24 GF measurements compared withsmooth measurement (run040,039).

Risø-R-1112(EN) 61

7.5 Risø-A1-24 VGs at 0.1 (run064, 065)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α



Risø-R-1112(EN)62

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN) 63

7.6 Risø-A1-24 VGs at 0.15 (run034,035)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α



Risø-R-1112(EN)64

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN) 65

7.7 Risø-A1-24 VGs at 0.2 (run027,028)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c





-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α



Risø-R-1112(EN)66

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN) 67

7.8 Risø-A1-24 VGs at 0.25 (run030,031)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c


VG 0.25 measurement, α=13oSmooth measurement, α=10o


-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α



Risø-R-1112(EN)68

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN) 69

7.9 Risø-A1-24 VGs at 0.30 (run081)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c



Figure 7-28 Cp at maximum CL for Risø-A1-24 VG 30% measurementcompared with smooth measurements (run081).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20

α

VG 0.30 measurementSmooth measurement

Figure 7-29 CL-CD for Risø-A1-24 VG 30% measurement compared withsmooth measurements (run081).

Risø-R-1112(EN)70

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Figure 7-30 CL, CD and CM for Risø-A1-24 VG 30% measurement comparedwith smooth measurements (run081).

Risø-R-1112(EN) 71

7.10 Risø-A1-24 GFs of 1% (run037,040)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

GF 1% LER measurement, α=9o



Figure 7-31 Cp at maximum CL for Risø-A1-24 GF 1% Smooth and LERmeasurement compared with smooth measurements (run037,040).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α

GF 1% LER measurementGF 1% measurement

Smooth measurement

Figure 7-32 CL-CD for Risø-A1-24 GF 1% Smooth and LER measurementcompared with smooth measurements (run037,040).

Risø-R-1112(EN)72

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


Smooth measurement

0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


Smooth measurement

-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Smooth measurement

Figure 7-33 CL, CD and CM for Risø-A1-24 GF 1% Smooth and LERmeasurement compared with smooth measurements (run037,040).

Risø-R-1112(EN) 73

7.11 Risø-A1-24 GFs of 2% (run038,039)-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c




Figure 7-34 Cp at maximum CL for Risø-A1-24 GF 2% Smooth and LERmeasurement compared with smooth measurements (run038,039).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α


Smooth measurement

Figure 7-35 CL-CD for Risø-A1-24 GF 2% Smooth and LER measurementcompared with smooth measurements (run038,039).

Risø-R-1112(EN)74

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


Smooth measurement

0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


Smooth measurement

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α


Smooth measurement

Figure 7-36 CL, CD and CM for Risø-A1-24 GF 2% Smooth and LERmeasurement compared with smooth measurements (run038,039).

Risø-R-1112(EN) 75

7.12 Risø-A1-24 VGs at 0.10 double spacing(run067,066)

-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

VG 0.10 DS measurement, α=16o

VG 0.10 DS LER measurement, α=16o



Figure 7-37 Cp at maximum CL for Risø-A1-24 VG 10% Double spacing smoothand LER measurement compared with VG 10% and smooth measurements(run066,067).

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α

VG 0.10 DS measurementVG 0.10 DS LER measurement


Figure 7-38 CL-CD for Risø-A1-24 VG 10% Double spacing smooth and LERmeasurement compared with VG 10% and smooth measurements (run066,067).

Risø-R-1112(EN)76

-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α



0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α



-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Figure 7-39 CL, CD and CM for Risø-A1-24 VG 10% Double spacing smoothand LER measurement compared with VG 10% and smooth measurements(run066,067).

Risø-R-1112(EN) 77

7.13 Risø-A1-24 VGs at 0.15, GFs of 1%, LER(run036)

-7

-6

-5

-4

-3

-2

-1

0

10 0.2 0.4 0.6 0.8 1

CP

x/c

VG 0.15 GF 1% LER measurementVG 0.15 LER measurement, α=20o



Figure 7-40 Cp at maximum CL for Risø-A1-24 VG 15% GF 1% LER comparedwith GF 1% LER, VG 15% LER and LER measurements (run036).

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

0 0.025 0.05 0.075 0.1

CL

CD

-5 0 5 10 15 20 25 30

α

VG 0.15 GF 1% LER measurementVG 0.15 LER measurement

GF 1% LER measurementLER measurement

Figure 7-41 CL-CD for Risø-A1-24 VG 15% GF 1% LER compared with GF 1%LER, VG 15% LER and LER measurements (run036).

Risø-R-1112(EN)78

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

-10 0 10 20 30

CL

α



0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α



-0.25

-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Figure 7-42 CL, CD and CM for Risø-A1-24 VG 15% GF 1% LER compared withGF 1% LER, VG 15% LER and LER measurements (run036).

Risø-R-1112(EN) 79


-0.5

0.0

0.5

1.0

1.5

2.0

-10 0 10 20 30

CL

α


-0.1

0.0

0.1

0.2

0.3

0.4

0.5

-10 0 10 20 30

CD

α


-0.20

-0.15

-0.10

-0.05

0.00

-10 0 10 20 30

CM

α



Risø-R-1112(EN)80

0.2

0.4

0.6

0.8

-1 0 1 2C

Lα

k = 0.092, αm = 0.6o, A = 1.4o

0.6

0.8

1

1.2

1.4

2 3 4 5 6

CL

α

k = 0.092, αm = 4.2o, A = 1.5o

0.8

1

1.2

1.4

4 5 6 7

CL

α

k = 0.092, αm = 5.5o, A = 1.5o

1

1.2

1.4

1.6

5 6 7 8 9

CL

α

k = 0.092, αm = 7.1o, A = 1.5o

1

1.2

1.4

1.6

7 8 9 10

CL

α

k = 0.092, αm = 8.6o, A = 1.6o

1

1.2

1.4

1.6

9 10 11 12 13

CL

α

k = 0.092, αm = 10.8o, A = 1.9o

0.8

1

1.2

1.4

12 13 14 15

CL

α

k = 0.092, αm = 13.3o, A = 2.1o

0.8

1

1.2

1.4

14 15 16 17

CL

α

k = 0.092, αm = 15.6o, A = 2.0o

0.8

1

1.2

1.4

16 17 18 19

CL

α

k = 0.092, αm = 17.5o, A = 2.0o

0.8

1

1.2

1.4

18 19 20 21

CL

α

k = 0.092, αm = 19.5o, A = 1.9o


Risø-R-1112(EN) 81

-0.05

-0.025

0

0.025

0.05

0.075

-1 0 1 2

CD

α

k = 0.092, αm = 0.6o, A = 1.4o

-0.05

-0.025

0

0.025

0.05

0.075

2 3 4 5 6

CD

α

k = 0.092, αm = 4.2o, A = 1.5o

-0.05

-0.025

0

0.025

0.05

0.075

4 5 6 7

CD

α

k = 0.092, αm = 5.5o, A = 1.5o

-0.05

-0.025

0

0.025

0.05

0.075

5 6 7 8 9

CD

α

k = 0.092, αm = 7.1o, A = 1.5o

-0.05

-0.025

0

0.025

0.05

0.075

7 8 9 10

CD

α

k = 0.092, αm = 8.6o, A = 1.6o

-0.025

0

0.025

0.05

0.075

0.1

9 10 11 12 13

CD

α

k = 0.092, αm = 10.8o, A = 1.9o

0.025

0.05

0.075

0.1

0.125

0.15

12 13 14 15

CD

α

k = 0.092, αm = 13.3o, A = 2.1o

0.075

0.1

0.125

0.15

0.175

0.2

14 15 16 17

CD

α

k = 0.092, αm = 15.6o, A = 2.0o

0.075

0.1

0.125

0.15

0.175

0.2

0.225

16 17 18 19

CD

α

k = 0.092, αm = 17.5o, A = 2.0o

0.1

0.125

0.15

0.175

0.2

0.225

18 19 20 21

CD

α

k = 0.092, αm = 19.5o, A = 1.9o


Risø-R-1112(EN)82

-0.12

-0.1

-0.08

-0.06

-0.04

-1 0 1 2C

Mα

k = 0.092, αm = 0.6o, A = 1.4o

-0.14

-0.12

-0.1

-0.08

-0.06

2 3 4 5 6

CM

α

k = 0.092, αm = 4.2o, A = 1.5o

-0.14

-0.12

-0.1

-0.08

-0.06

4 5 6 7

CM

α

k = 0.092, αm = 5.5o, A = 1.5o

-0.14

-0.12

-0.1

-0.08

-0.06

5 6 7 8 9

CM

α

k = 0.092, αm = 7.1o, A = 1.5o

-0.16

-0.14

-0.12

-0.1

-0.08

-0.06

7 8 9 10

CM

α

k = 0.092, αm = 8.6o, A = 1.6o

-0.16

-0.14

-0.12

-0.1

-0.08

9 10 11 12 13

CM

α

k = 0.092, αm = 10.8o, A = 1.9o

-0.16

-0.14

-0.12

-0.1

-0.08

12 13 14 15

CM

α

k = 0.092, αm = 13.3o, A = 2.1o

-0.16

-0.14

-0.12

-0.1

-0.08

14 15 16 17

CM

α

k = 0.092, αm = 15.6o, A = 2.0o

-0.16

-0.14

-0.12

-0.1

-0.08

16 17 18 19

CM

α

k = 0.092, αm = 17.5o, A = 2.0o

-0.16

-0.14

-0.12

-0.1

-0.08

18 19 20 21

CM

α

k = 0.092, αm = 19.5o, A = 1.9o


Risø-R-1112(EN) 83

8 Discussion

8.1 Risø-A1-18Measurements of Risø-A1-18 are shown in Chapter 5 and main results areshown in Table 8-1.

Results of the measurements with smooth surface conditions are shown inSection 5.1 . The slope of the CL curve remains constant until a CLmax of about1.43 is reached at α = 10° and when stall begins a well defined upper limit isset on CL. In the post stall area CL is smoothly reduced until deep stall. Themeasurement of CD shows a CDmin of around 0.0097 which is higher thanexpected. The high background turbulence level of 1% will increase CD.However, for this particular airfoil section, the high CD is mainly due to thequality of the airfoil section, where the trailing edge was irregular and the finisharound the pressure tabs was bad. Nevertheless it is positive that CD remainsnearly constant until stall occurs.

The constant steep slope of the CL curve and the low CD nearly until stall is inagreement with the design strategy where the operational point was designed tobe close to CLmax.

The measurements compare well with computational results from the Navier-Stokes solver, EllipSys2D. Figure 5-2 and Figure 5-3 in Section 5.1 show themeasured and computed CP distributions for α = 2° and α = 8°. The latter is justbefore CLmax is obtained. The fully turbulent computed CP distribution is in verygood agreement with the measured CP distribution on the entire airfoil at thelower angles of attack. Differences between measurements and computationsoccur at higher angles of attack. In Figure 5-3 at the first 5% of the suction side,the free transition computation compares better with the measurement. Hence,the shape of the CP curve is in good agreement at low angles of attack exceptfor the location of the transition point, which is predicted too far down streamcompared to the measurement. At high angles of attack the large acceleration ofthe suction side flow around the leading edge suppress the turbulence and thiscauses deviation between the measurements and the turbulent computations.

The measured and the predicted CDmin, Figure 5-4, and the shape of the CD

curve at low angles of attack are both in good agreement. However, CLmax couldnot be correctly predicted when the Michel transition model was used duringthe computation. Good agreement for CLmax was obtained for fully turbulentflow computations compared with the smooth flow measurements.

The agreement for CLmax between the measurement and the fully turbulent flowcomputation indicates that the measured flow is more or less fully turbulentclose to stall. This is in agreement with the design strategy where transitionshould be located close to leading edge to ensure insensitivity of CLmax toleading edge roughness. As a result of the design strategy aiming for aroughness insensitive airfoil a small suction peak on the leading edge appears at

Risø-R-1112(EN)84

α = 8°. The suction peak increases with the angle of attack when getting closerto the stall angle. The peak eventually causes transition from laminar toturbulent flow close to the leading edge making CLmax insensitive to leadingedge roughness. The measurements show that transition occurs at the leadingedge also at angles of attack well below the stall angle indicating that thepressure peak that provokes transition either is too pronounced or occurs at toolow angles of attack.

Measurements with zigzag tape to represent leading edge roughness are shownin Section 5.2 . Compared with smooth flow CDmin is increased to 0.012 andCLmax is reduced to 1.18. The slope of the CL curve is nearly as steep for leadingedge roughness flow as for smooth flow. This is in good agreement with thedesign strategy aiming for CLmax being insensitive to leading roughness.However, the drop in CLmax is somewhat higher than expected from the designcalculations. The zigzag tape used to simulate leading edge roughness is alsoused on small aeroplane wings. It is therefore possible that it represents amassive roughness for the tested airfoil. Hence the deviations from thetheoretically expected values can be attributed to he lack of correct roughnesssimulation.

Measurements with vortex generators (VGs) are shown in Section 5.3 - 5.5 .The CP distributions for the different VG configurations were obtained directlyfrom the pressure measurements, neglecting the 3d variation in the pressure.The CP distribution at α = 10o for smooth flow shows that CP is nearly identicalfor low angles of attack, Figure 5-9. Hence, the VG’s have neglecting influenceon CP when the corresponding smooth flow without VGs is attached. The 3dinfluence on the pressure is negligible so that the measured CP distribution isrepresentative for the VG flow. At higher angles of attack, the VGs reduce thepressure on most of the suction side. Separation is delayed to a higher angle ofattack and CLmax is increased. The slope of the linear part of the CL curve andthe angle of attack for zero CL are not affected by the VGs. The resulting CLmax

depends on the chordwise location of the VGs. For this airfoil x/c = 0.20 andx/c = 0.25 were measured. At x/c = 0.20 CLmax was increased to 1.82. However,the VGs also increased CD in the pre-stall region. Compared to the smooth flowmeasurement in the post stall region CD remains the same for both smooth andVG conditions. In the area where the flow remains attached due to VGs, CD

remains favourably lower compared to the smooth flow measurements, Figure5-12.

Unfortunately the reported CD curves with VGs are not accurate because thewake rake pressures were used for calculation of CD. There will be a spanwisevariation in the momentum downstream of the airfoil section because of thediscrete location of the vortex generators. The wake rake represents a snap shotof the flow in one particular spanwise location. Since the wake rake was nottraversed a representative average value of CD could not be obtained.

Both VG configurations were measured with leading edge roughness. Thepresence of leading edge roughness only marginally reduced CLmax whereasCDmin was increased for angles of attack above 0°.

Measurements with smooth surface condition and dynamic inflow are shown inSection 5.6 . Time series were measured with the airfoil section being in aharmonic motion around x/c = 0.40 at different mean angles with a geometricmaximum amplitude of A = ±2° and a reduced frequency of k = 0.092. The

Risø-R-1112(EN) 85

derivation of the hysteresis loops from time series is explained in Fuglsang etal., [2]. When wind tunnel corrections are applied for the angle of attack, theamplitude of the harmonic motion is changed.

Table 8-1CLmax and CDmin for Risø-A1-18.

Smooth LER VG 0.20 VG 0.20LER

VG 0.25 VG 0.25LER

CLmax 1.43 1.18 1.82 1.75 1.76 1.66CDmin 0.0097 0.012 0.016 0.017 0.015 0.016

8.2 Risø-A1-21Measurements of Risø-A1-21 are shown in Chapter 6 and main results areshown in Table 8-2.

Results of measurements with smooth surface conditions are shown in Section6.1 . The measured CP distribution is slightly irregular on the rear part of thesuction side. This is due to measurement errors but it will not affect thederivation of CL. The overall measurements show similar trends as for the Risø-A1-18 airfoil with a CLmax of about 1.38 and CDmin of about 0.0092.

In the post stall area the CL curve slope is negative and rather steep Figure 6-5.In combination with Figure 6-1 this shows, that the airfoil has a well-definedstall characteristic. Stall occurs at around 10o and moves rapidly to the leadingedge within an angle of attack range of 4o. Due to the thickness of the airfoilthis is not unexpected. However, at the design stage a smaller drop in CL wasintended.

The measurements compare well with the computational results fromEllipSys2D with similar trends as for Risø-A1-18 with even better agreement ofCDmin.

Measurements with leading edge roughness are shown in Section 6.2 .Compared with smooth flow CDmin is increased to 0.014 and CLmax is reduced to1.20. Conclusions regarding roughness insensitivity are the same as forRisø-A1-18.

Measurements with VGs are shown in Section 6.3 - 6.5 . CLmax is increased to1.89 for the x/c = 0.20 configuration and 1.67 for the x/c = 0.25 configuration.CD is higher for the x/c = 0.20 configuration compared with the x/c = 0.25configuration. The choice for the chordwise location of the VGs is crucial forthe obtainable CLmax and CD since large variations occur for small shifts of theirposition. This is in contrast to the Risø-A1-18 airfoil where CLmax was almostthe same for the measured VG configurations

Both VG configurations were measured with leading edge roughness and thisonly marginally reduced CLmax whereas CDmin was increased.

Measurements with smooth surface conditions and dynamic inflow are shownin Section 6.6 . The geometric amplitude is A = ±2° and the reduced frequencyis k = 0.092.

Risø-R-1112(EN)86

Table 8-2 CLmax and CDmin for Risø-A1-21.

Smooth LER VG 0.20 VG 0.20LER

VG 0.25 VG 0.25LER

CLmax 1.38 1.20 1.89 1.85 1.67 1.60CDmin 0.0092 0.014 0.020 0.022 0.017 0.018

8.3 Risø-A1-24Measurements of Risø-A1-24 are shown in Chapter 7 and main results areshown in Table 8-3 to Table 8-5.

Results of measurements with smooth surface conditions are shown in Section7.1 . The overall measurements show similar trends as for the other testedairfoils. CLmax is about 1.36 and CDmin is about 0.010. As for the Risø-A1-21airfoil there is a significant drop in CL during post stall.

The measurements are in good agreement with the computational results fromEllipSys2D with similar trends as for the other tested airfoils.

Measurements with leading edge roughness are shown in Section 7.2 .Compared with smooth flow CDmin is increased to 0.016 and CLmax is reduced to1.17. Conclusions regarding roughness insensitivity are the same as for theother tested airfoils.

Measurements with VGs are shown in Section 7.3 , 7.5 -7.9 and 7.12 . Theobtainable CLmax depends on the chordwise location of the VGs. When the VGsare moved closer to the leading edge CLmax is increased until x/c = 0.15 where aCLmax around 1.90 is obtained. Moving the VGs further upstream results in anentirely different flow pattern with a lower CLmax and a smoother post stall area.When the VGs are moved toward the leading edge CDmin is also increased.

When the VGs increase CLmax to very high values it is inevitable that there willbe a large drop in CL in the post stall area. However, the angles of attack wherethe drop occurs are so high that the flow on a rotor inboard part will have acomplex 3d-flow pattern. This will to some extent counterbalance the 2D dropin CL.

A measurement with larger spacing between the VGs at x/c = 0.10 is shown inSection 7.12 . The larger spacing does not seem to reduce Clmax. However, CL inthe post stall area is reduced. This indicates that the spacing is equallyimportant as the chordwise location when the VG configuration is optimised.

Measurements with Gurney flaps (GFs) are shown in Section 7.4 , 7.10 and7.11 . The measured CP distributions show that GFs increase the area inside ofthe CP curve giving higher CL. To the same angle of attack, CL is increased. Thismoves the angle of attack for zero CL towards negative. The angle of attack forCLmax is not influenced by the GFs. CLmax is measured to about 1.72 at a GFheight of 2% with a CDmin about 0.016. A drawback from GFs is a large drop inCL during post stall.

Risø-R-1112(EN) 87

From the measurement of the significant increase in CD compared with smoothflow it appears that measurements should have been carried out with lowerheight of the GFs compared to 1%. From this it could be expected that CD

would not increase so much but that there would still be an increase in CLmax.

The GF measurements show the importance of the shape of the trailing edge. Adeviation in the trailing edge angle with the chord can cause a different angle ofattack for zero CL and a different CLmax

A measurement with VGs, GFs and leading edge roughness is shown in Section7.13 . The combination of VGs and GFs increase CLmax to about 2.0. VGs andGFs supplement each other well and the measurement is an example of thatvery high CLmax can be achieved by simultaneous application of VGs and GFs.

Measurements with smooth surface condition and dynamic inflow are shown inSection 7.14 . The geometric amplitude is A = ±2° and the reduced frequency isk = 0.092.

Table 8-3 CLmax and CDmin for Risø-A1-24 smooth, LER and GF.

Smooth LER GF 1% GF 1%LER

GF 2% GF 2%LER

GF 1%VG 0.15LER

CLmax 1.36 1.17 1.63 1.50 1.72 1.61 2.02CDmin 0.010 0.016 0.015 0.020 0.016 0.021 0.025

Table 8-4 CLmax and CDmin for Risø-A1-24 VG.

VG 0.10 VG 0.10DS

VG 0.15 VG 0.20 VG 0.25 VG 0.30

CLmax 1.74 1.68 1.90 1.81 1.67 1.59CDmin 0.016 0.024 0.017 0.019 0.016 0.017

Table 8-5 CLmax and CDmin for Risø-A1-24 VG LER.

VG 0.10LER

VG 0.10DS LER

VG 0.15LER

VG 0.20LER

VG 0.25LER

CLmax 1.65 1.69 1.84 1.77 1.66CDmin 0.020 0.024 0.021 0.019 0.021

8.4 SummaryMeasurements of CL and CD for Risø-A1-18, Risø-A1-21 and Risø-A1-24airfoils for smooth surface conditions and leading edge roughness are shown inFigure 8-1 and Figure 8-2 respectively.

All airfoils have very similar characteristics at smooth surface conditionsexcept for post stall where Risø-A1-21 and Risø-A1-24 have a larger drop in CL

compared with Risø-A1-18. With leading edge roughness the variation in CD islarger with higher CDmin for thicker airfoils. However, all airfoils haveapproximately the same CLmax.

Risø-R-1112(EN)88

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20 25

α

RISO-A1-18 SmoothRISO-A1-21 SmoothRISO-A1-24 Smooth

Figure 8-1 CL,-CD for Risø-A1-18, Risø-A1-21 and Risø-A1-24 smoothmeasurements.

-0.5

0.0

0.5

1.0

1.5

2.0

0 0.02 0.04 0.06

CL

CD

-5 0 5 10 15 20 25

α

RISO-A1-18 SmoothRISO-A1-21 SmoothRISO-A1-24 Smooth

Figure 8-2 CL,-CD for Risø-A1-18, Risø-A1-21 and Risø-A1-24 LERmeasurements.

Risø-R-1112(EN) 89

9 Conclusions

The Risø-A1 airfoil family was designed with operational Reynolds numbersaround 3.0 million for a 600 kW wind turbine rotor with the followingcharacteristics:• Maximum lift coefficient around 1.5 in natural conditions for all airfoils• High lift-drag ratio also at high angles of attack just below maximum lift• Insensitivity to leading edge roughness for the maximum lift coefficient• Trailing edge stall, smooth post stall behaviour and no double stall

To address the agreement between the design characteristics and the actualairfoil performance three of the airfoils, Risø-A1-18, Risø-A1-21 and Risø-A1-24 were tested experimentally in a wind tunnel. The testing conditionscorresponded to the maximum flow velocity of 42 m/s giving a Reynoldsnumber of, Re = 1.6×106 compared with the design Reynolds numbers aroundRe = 3.0×106. According to Navier-Stokes calculations the difference inReynolds number would reduce the maximum lift coefficient by approximately0.1 reducing the design maximum lift coefficient from 1.5 to 1.4. The tunnelhas a relatively high background turbulence level of 1% making the resultsmore representative for natural conditions than it would be the case for a lowturbulence wind tunnel.

For smooth surface conditions, all three airfoils had the desirable properties ofconstant lift curve slope and almost constant drag coefficient until maximumlift is reached. The maximum lift coefficient and the minimum drag coefficientfor the airfoils were about 1.4 and 0.010 respectively.

Tests on all airfoil sections mounted with zigzag tape to simulate leading edgeroughness showed that the airfoils were reasonably insensitive to this. Anincrease in the minimum drag coefficient to about 0.014 was inevitable. Themaximum lift coefficient was reduced to about 1.2. The zigzag tape did notaffect the slope of the lift coefficient versus angle of attack. The drop inmaximum lift was somewhat higher than expected during the airfoil design.This indicated that the zigzag tape was not dimensioned properly, relative to thesize of the airfoils.

The measurements in post stall showed that the Risø-A1-18 airfoil had asmooth post stall area with high lift coefficient retained until deep stall.Unfortunately measurements of Risø-A1-21 and Risø-A1-24 showed a largerdrop in the lift coefficient than expected during post stall indicating massiveseparation on the suction side, which was not foreseen to this extent during theairfoil design.

Mounting of vortex generators and Gurney flaps showed that there was roomfor improvement of the aerodynamic properties of the airfoils by relativelysimple means. The improvements can be applied with good results especiallyon the inner parts of wind turbine blades. All airfoils were well suited forapplication of vortex generators, e.g., vortex generators for Risø-A1-24 in 15%chord increased the maximum lift coefficient from 1.38 to 1.90. Thecombination of vortex generators and Gurney flaps increased the maximum lift

Risø-R-1112(EN)90

coefficient to about 2.0. The drag penalty has to be considered individuallyaccording to the application. Unfortunately the measurements of drag at thevortex generator and Gurney flap configurations was uncertain since the wakerake was not traversed to obtain an average value representative to 2-dimensional flow.

The experimental results for smooth surface conditions compare well withcomputational results from the general-purpose Navier-Stokes solver,EllipSys2D. The minimum drag coefficient and the shape of the drag curve atlow angles of attack were in very good agreement. However, the maximum liftcoefficient could not be predicted when transition was included in thecomputation. Good agreement for the maximum lift coefficient compared withthe measurements was obtained for fully turbulent flow computations.

Risø-R-1112(EN) 91

References

[1] Fuglsang, P., Dahl, K.S., 1999, ‘Design Of The New Risø-A1 AirfoilFamily for Wind Turbines,’ Proc. EWEC’99, Nice, France.

[2] Fuglsang, P., Antoniou, I., Sørensen, N.N., Madsen, H. Aa., 1998,‘Validation of a Wind Tunnel Testing Facility for Blade Surface PressureMeasurements.’ Risø-R-981(EN), Risø National Laboratory, Denmark.

[3] Rae Jr., W.H., Pope, A., 1984, Low-Speed Wind Tunnel Testing, SE,John Wiley & Sons, ISBN 0-471-87402-7.

[4] Brooks, T.F. and Marcolini, M.A., 1984, Airfoil Trailing Edge FlowMeasurements and CompaRisøn with Theory Incorporating Open WindTunnel Corrections, AIAA-84-2266, AIAA/NASA 9th AeroacousticConference.

[5] Hoerner, S.F. and Borst, H.V., 1975, Fluid-Dynamic Lift, Hoerner FluidDynamics, P.O. Box 342, Brick Town, N.J. 08723, USA.

[6] Timmer, W.A., 1992, The design and testing of airfoil DU 91-W2-250.Proc. IEA Joint Action. Aerodynamics of Wind Turbines. 6th

Symposium.[7] Sørensen, N.N., 1995, General Purpose Flow Solver Applied to Flow

over Hills, Risø-R-827(EN), Risø National Laboratory, Denmark.[8] Menter, F.R., 1993, Zonal Two Equation k-ω Turbulence Models for

Aerodynamic Flows. AIAA Paper 93-2906.[9] Michel, R., 1952, Etude de la transition sur les profiles d’aile. ONERA

Report 1/1578-A. See White F.M., Viscous fluid flow, p. 442.[10] Fuglsang, P., Antoniou, I., Dahl, K.S., Madsen, H.Aa., 1998, ‘Wind

Tunnel Tests of the FFA-W3-241, FFA-W3-301 and NACA 63-430Airfoils.’ Risø-R-1041(EN), Risø National laboratory, Denmark.

Risø-R-1112(EN)92

A Measurement survey

This appendix describes the performed measurements in detail to make themeasurements stored on CD available for subsequent exploitation. The differentmeasurement types are described and the naming convention for the data files isexplained. The format of the data files is given and each performedmeasurement is listed and described.

A.1 List of symbolsh [cm] Wake rake vertical position, positive toward floor, origin at

wake rake topk Reduced frequencyp [Pa] Static pressurepo [Pa] Total pressure headpatm [Pa] Atmospheric pressureq [Pa] Dynamic pressurex Airfoil chordwise co-ordinate relative to chord, positive

toward trailing edge, origin at leading edgey Airfoil vertical co-ordinate relative to chord, positive toward

ceiling, origin at leading edgeΑ [°] Pitch motion amplitudeCD Drag coefficientCL Lift coefficientCM Moment coefficientCP Airfoil pressure coefficientRe Reynolds numbert [°C] Air temperature

α [°] Angle of attack

ρ [kg/m3] Air densitySubscriptsc Corrected valuep Pressure measurement (opposite to wake rake measurement)w Wake rake measurement∞ Reference for normalisation of airfoil forces

A.2 Measurement typesThere are four different basic types of measurements of the airfoil flow asshown in Table A-1:• STEP• CONT• STAT• PITCH

Risø-R-1112(EN) 93

Table A-1 Overview of the different types of measurements that have beenperformed.

STEPThe lift, drag and moment polar versus angle of attack.• Discrete measurements at different angles of attack.• Angle of attack range: -6° to 30°.• Interval between different angles: 1° to 4°.• Time series length: 20 s.• Sampling frequency: 5 Hz.CONTThe lift, drag and moment polar versus angle of attack. (shorter measurementtime compared to ‘STEP’)• Continuous measurements at different angles of attack.• Angle of attack range: -6° to 30°.• Rate of change of angle of attack: 0.1°/s to 0.5°/s (manually changed).• Time series length app: 250 s.• Sampling frequency: 50 Hz.STATTime series of airfoil flow at different angles of attack, usually in stall.• Stationary measurements at different angles of attack.• Time series length: 20s to 180s.• Sampling frequency: 100 Hz.PITCHTime series of unsteady airfoil flow from pitching motion for determination ofhysteresis loops for lift, drag and moment at different pitching frequencies andamplitudes.• Dynamic measurements at different mean angles of attack with the airfoil

in pitching motion.• Pitching amplitude: 2° to 6°• Reduced frequency: to 0.15• Time series length: 30s to 40s.• Sampling frequency: 100 Hz.

The following table contains a list of all the data files that are available for eachtype of measurement. A detailed description of the data files is given inSection A.4.

Risø-R-1112(EN)94

Table A-2 Overview of the available data files for each measurement type.

STEPrxx.pol Average general results where the average of each sub

measurement is written in rows, Table A-4.Well suited to obtain CL and CD versus α.

cp.pol Average CP distributions where the average of each submeasurement is written in columns.Well suited to obtain CP distributions at different angles ofattack (from rxx.pol).

v_w.pol Average wake rake velocity distributions where the average ofeach sub measurement is written in columns.

rxx-5hz.nnn Raw data as rows in general data file with frames of 5 Hz,Table A-4.

cp-5hz.nnn CP distribution in columns with frames of 5 Hz.v_w-5hz.nnn Wake velocity distribution in columns with frames of 5 Hz.CONTrxx.bin Average results where raw data are sorted in bins of the angle

of attack, Table A-6.Well suited to obtain CL, CD and CM versus α.

rxx.pol Raw data as rows in general data file with frames of 50 Hz.First 15 columns of data file described in Table A-4. All submeasurements are merged together in one file

rxx-1hz.nnn Raw data as rows in general data file reduced to frames of 1Hz, Table A-4.

rxx-10hz.nnn Raw data as rows in general data file reduced to frames of 10Hz, Table A-4.

rxx-50hz.nnn Raw data as rows in general data file with frames of 50 Hz,Table A-4.

cp-1hz.nnn CP distribution in columns reduced to frames of 1 Hz.v_w-1hz.nnn Wake velocity distribution in columns reduced to frames of 1

Hz.STATrxx.nnn Raw data as rows in data file with frames of 50 Hz, Table A-5.rxx-1hz.nnn Raw data as rows in general data file reduced to frames of 1

Hz, Table A-4.rxx-10hz.nnn Raw data as rows in general data file reduced to frames of 10

Hz, Table A-4.rxx-50hz.nnn Raw data as rows in general data file with frames of 50 Hz,

Table A-4.cp-1hz.nnn CP distribution in columns reduced to frames of 1 Hz.v_w-1hz.nnn Wake velocity distribution in columns reduced to frames of 1

Hz.PITCHrxx-loop.bin Average results where raw data are sorted in bins of the phase

angle of the hysteresis loop, Table A-7Well suited to obtain hysteresis loops of CL and CD and CM

versus α.rxx.nnn Raw data as rows in data file with frames of 100 Hz, Table

A-5.rxx-100hz.nnn Raw data as rows in general data file with frames of 100 Hz,

Table A-4.

Risø-R-1112(EN) 95

Wherexx is the measurement run numberyy is the frame resolution in Hznnn is the sub measurement number

A.3 Data file naming conventionThe different data files are stored in the following directory structure:• The name of the airfoil.• The measurement run name.

The naming and the format of the data files is explained in Table A-3.

A.4 Data file formatsThe different data files are shown in Table A-3.

Table A-3 Available data files.

rxx-yyhz.nnn

xx Measurement run numberyy Frame average frequency: 100, 50, 10, 5 or 1nnn measurement sub numberGeneral data file with each measurement frame/average formatted in rows. Thefirst two rows contain the column number and the sensor name. The format ofthe data files is described in Table A-4STEP Measurements are given with 5 Hz frame resolutionCONT Measurements are given with 50 Hz, 10 Hz and 1 Hz frame

resolutions respectivelySTAT Measurements are given with 50 Hz, 10 Hz and 1 Hz frame

resolutions respectivelyPITCH Measurements are given with 100 Hz resolution

rxx.nnn

xx Measurement run numbernnn measurement sub numberRaw data file with each measurement frame/average formatted in rows. Thefirst two rows contain the column number and the sensor name. The format ofthe data files is described in Table A-5.PITCH Measurements are given with 100 Hz frame resolutionSTAT Measurements are given with 50 Hz frame resolution

Risø-R-1112(EN)96

rxx.pol

xx Measurement run numberGeneral data file with overall average values of each sub measurementformatted in rows, as above. All sub measurement for a given measurement runare assembled into one single file. The format of the data files is described inTable A-4.STEP Measurements are given with a 20 s average value for each sub

measurementPITCH Measurements are given with a 30 s average value for each sub

measuremenSTAT Measurements are given with a 180 s average value for each sub

measurement

rxx.pol

xx Measurement run numberGeneral data file with each measurement frame formatted in rows, as above. Allsub measurement for a given measurement run are assembled into one singlefile. The format of the data files is described in Table A-4, but the files containonly the first 15 collumns.CONT Measurements are given with 50 Hz resolution, only column 1 to

15.

rxx.bin

xx Measurement run numberPost processed data file where the frames from all sub measurements are sortedin bins of the angle of attack to obtain the polar curves. The format of the datafiles is described in Table A-6.CONT Measurements where the 50 Hz frames from all sub measurements

are sorted in bins of αc. The angle of attack range is divided into 30bins.

rxx-loop.bin

xx Measurement run numberPost processed data file where the frames from all sub measurements are sortedin bins of the phase of the hysteresis loop. The format of the data files isdescribed in Table A-7.PITCH Measurements where the 100 Hz frams from all sub measurements

are sorted in bins of the phase of the hysteresis loop. The phaserange is divided into 30 bins.

Risø-R-1112(EN) 97

cp-yyhz.nnn

yy Frame average frequency: 5 or 1nnn measurement sub numberData file with each frame/average formatted in columns. The first columncontains the x-coordinates of the pressure tabs. The subsequent columnscontain the CP distributions for the different frames. The angle of attack for theframes can be found in the corresponding, rxx-yyhz.nnn file.STEP Measurements are given with 5 Hz frame resolutionCONT Measurements are given with 1 Hz frame resolutionSTAT Measurements are given with 1 Hz frame resolution

cp.pol

Data file with each average formatted in columns. The first column contains thex-coordinates of the pressure tabs. The subsequent columns contain the averageCP distributions for each sub measurement. The angle of attack for the framescan be found in the corresponding, rxx.pol file.STEP Measurements are given as 20 s average values

v_w-yyhz.nnn

yy Frame average frequency: 5 or 1nnn measurement sub numberData file with each frame/average formatted in columns. The first columncontains the coordinates of the wake rake total pressure tabs. The subsequentcolumns contain the wake rake velocity for the different frames. The angle ofattack for the frames can be found in the corresponding, rxx-yyhz.nnn file.STEP Measurements are given with 5 Hz frame resolutionCONT Measurements are given with 1 Hz frame resolutionSTAT Measurements are given with 1 Hz frame resolution

v_w.pol

Data file with each average formatted in columns. The first column contains thecoordinates of the wake rake total pressure tabs. The subsequent columnscontain the average wake rake velocity distributions for each sub measurement.The angle of attack for the frames can be found in the corresponding, rxx.polfile.STEP Measurements are given as 20 s average values

Risø-R-1112(EN)98

Table A-4 The content of the columns in the general data file.

Col. Symbol Sensor Unit Description1 αc αc ° Corrected angle of attack2 CL cl - Lift coefficient (pressure)3 CDc cdc - Corrected drag coefficient (wake

rake + pressure)4 CMc cmc - Corrected moment coefficient

(pressure)5 CDpc cdpc - Corrected drag coefficient

(pressure)6* CDw cdw Drag coefficient (wake rake)7 α α ° Raw angle of attack8 CD cd - Raw drag coefficient (wake rake +

pressure)9 CDp cdp Raw drag coefficient (pressure)10 CM cm - Raw moment coefficient (pressure)11 Re re Free stream Reynolds Number12 q∞ qref Pa Free stream dynamic pressure13 p∞ ps,ref Pa Free stream static pressure14 T t ° Tunnel temperature15 patm patm Mbar Atmospheric pressure16-71**

CP cp(x) Pressure coefficients correspondingto the coordinates in top row

72-74 p1-3 ps,Pitot() Pa Pitot tube static pressures75-77 po1-3 pt,Pitot() Pa Pitot tube total pressures78-82* pw ps,wake Pa Wake rake static pressures

corresponding to the coordinates intop row

83-136*

pow pt,wake Pa Wake rake total pressurescorresponding to the coordinates intop row

*) At the ‘PITCH’ type measurements, the wake rake was not used. CDW was setto CDP and pw and pow were not written in the data files**) In some measurements one or more of the airfoil pressure sensors wereexcluded because of unstable calibration or because the pressure hole wasblocked by vortex generators or roughness elements. The corresponding columnin the file was then removed and the number of subsequent sensors changed.

Risø-R-1112(EN) 99

Table A-5 The content of the columns in the raw data files.

Col. Symbol Sensor Unit Description1 t t s Running time2 α α ° Raw angle of attack3 CL cl - Lift coefficient (pressure)4 CDc cdc - Corrected drag coefficient (wake



(pressure)7* r ramp rad Hysteresis loop phase angle*) At the ‘STAT’ type measurements, the hysteresis loop phase angle was notused.

Table A-6 The content of the columns in the post processed data files sorted inbins of the angle of attack.

Col. Symbol Sensor Unit Description1 αc αc ° Corrected angle of attack2 CL cl - Lift coefficient (pressure)3 CDc cdc - Corrected drag coefficient (wake



(pressure)

Table A-7 The content of the columns in the post processed data files sorted inbins of the phase angle of the hysteresis loop.

Col. Symbol Sensor Unit Description1 αc αc ° Corrected angle of attack2 CL cl - Lift coefficient (pressure)3 CDpc cdpc - Corrected drag coefficient

(pressure)4 CMc cmc - Corrected moment coefficient

(pressure)

Risø-R-1112(EN)100

A.5 Performed measurementsTable A-8 to Table A-10 contain a list of the performed measurements for thedifferent airfoil sections.

Table A-8 Performed measurements for Risø-A1-18

Run Type Surface conditions Remarks013 CONT VGs at 0.20014 CONT VGs at 0.20015 CONT VGs at 0.20

LER016 CONT LER017 CONT Smooth flow018 STEP Smooth flow019 STAT Smooth flow020 CONT VGs at 0.25021 CONT VGs at 0.25

LER022 PITCH Smooth flow k = 0.092, 1.4° < A < 2.0°


Run Type Surface conditions Remarks001 CONT Smooth flow002 STEP Smooth flow003 CONT VGs at 0.20004 CONT VGs at 0.20

LER005 CONT LER023 CONT VGs at 0.25

LER024 CONT VGs at 0.25025 CONT Smooth flow026 PITCH Smooth flow k = 0.092, 1.4° < A < 2.0°

Risø-R-1112(EN) 101


Run Type Surface conditions Remarks

027 CONT VGs at 0.20028 CONT VGs at 0.20

LER029 CONT LER030 CONT VGs at 0.25

LER031 CONT VGs at 0.25032 CONT Smooth flow033 STEP Smooth flow034 CONT VGs at 0.15035 CONT VGs at 0.15

LER036 CONT VGs at 0.15

GF of 1%LER

037 CONT GF of 1%LER

038 CONT GF of 2%LER

039 CONT GF of 2%040 CONT GF of 1%041 PITCH GF of 1% k = 0.092, 1.4° < A < 2.0°042 CONT Smooth flow043 PITCH Smooth flow k = 0.092, 1.4° < A < 2.0°064 CONT VGs at 0.10065 CONT VGs at 0.10

LER066 CONT VGs at 0.10 double

spacingLER

067 CONT VGs at 0.10 doublespacing

081 CONT VGs at 0.30

Bibliographic Data Sheet Risø-R-1112(EN)Title and authors

Wind Tunnel Tests of the Risø-A1-18, Risø-A1-21 and Risø-A1-24 Airfoils

Peter Fuglsang, Kristian S. Dahl, Ioannis Antoniou

ISBN ISSN

87-550-2538-2 0106-2840

Department or group Date

Wind Energy and Atmospheric Physics Department June 1999

Groups own reg. number(s) Project/contract No(s)

ENS-1363/98-0038

Pages Tables Illustrations References

102 16 98 10

Abstract (max. 2000 characters)

This report contains 2D measurements of the Risø-A1-18, Risø-A1-21 andRisø-A1-24 airfoils. The aerodynamic properties were derived from pressuremeasurements on the airfoil surface and in the wake. The VELUX open jetwind tunnel was used having a background turbulence intensity of 1%, a flowvelocity of 42 m/s and a Reynolds number of 1.6×106. The airfoil sections had achord of 0.60 m and a span of 1.9 m and end plates were used to minimise 3Dflow effects. The measurements comprised both static and dynamic inflowwhere dynamic inflow was obtained by pitching the airfoil in a harmonicmotion. We tested the influence of leading edge roughness, vortex generatorsand Gurney flaps both individually and in combination.For smooth surface conditions, all three airfoils had the desirable properties ofconstant lift curve slope and low drag coefficient until the maximum lift ofabout 1.4 was reached. The Risø-A1-18 airfoil had a smooth post stall whereasthe Risø-A1-21 and Risø-A1-24 airfoils had a significant drop in the liftcoefficient after stall. Test on all airfoil sections mounted with zigzag tapeshowed that the airfoils were insensitive to leading edge roughness. Howeverwith a drop in the maximum lift coefficient to about 1.2. Mounting of deltawing shaped vortex generators and Gurney flaps showed that there was roomfor a significant increase in the maximum lift coefficient, which was increasedto 1.90 for Risø-A1-24 with vortex generators located at 15% chord. Thecombination of vortex generators and Gurney flaps increased the maximum liftcoefficient to about 2.0.

Descriptors INIS/EDB

AERODYNAMICS; AIRFOILS; DRAG; STALL; TEST FACILITIES;TESTING; TURBINE BLADES; TURBULENT FLOW;TWO-DIMENSIONAL CALCULATIONS; WIND TUNNELS;WIND TURBINESAvailable on request from Information Service Department, Risø National Laboratory,(Afdelingen for Informationsservice, Forskningscenter Risø), P.O.Box 49, DK-4000 Roskilde, Denmark.Telephone +45 46 77 46 77, ext. 4004/4005, Telefax +45 46 77 40 13

Wind Tunnel Tests of the Risø-A1-18, Risø-A1-21 and Risø ... · This report contains 2D...

Documents

Transcript of Wind Tunnel Tests of the Risø-A1-18, Risø-A1-21 and Risø ... · This report contains 2D...