large & small blades: design &...

Copyright © 2004 Wetzel Engineering, Inc.

Kyle K. Wetzel, Ph.D.

Wetzel Engineering, Inc.

Lawrence, Kansas

Sandia National Laboratory

Wind Turbine Blade Workshop

February 24, 2004

Albuquerque, New Mexico

Large & Small Blades:Design & Manufacturing

Tailored Twist-Flap Coupling



Tailored Twist-Flap Coupling

• Tailored twist-flap coupling can be used to shed transient loads, thereby reducing fatigue

• Allow an increase in rotor diameter• Increased Energy Capture• Value of increased energy capture exceeds cost of

implementing coupling• Reduction in the cost of energy


Wetzel Engineering Twist-Flap Coupled Blade Studies

•3-m Blades for a 6-kW Turbine•USDOE Distributed Wind Generation Contract

•9-m Blades for Sandia’s 115-kW LIST Turbine•Subcontractor to Wichita State University

•34/37/39-m Blades for GE Wind’s 1.5-MW Turbine•USDOE STTR Contract•Subcontractor to GE Wind (NGT Program)


Normal Operation

Lift

Drag

TypicalOutboard Section

Resultant ForceOn Blade


High-Speed Gust

Lift

Drag

Resultant ForceOn Blade


Lift

DragResultant ForceOn Blade

High-Speed Gust: Pitch Regulation


Lift


High-Speed Gust: Pitch Regulation

• Active Pitch Control will always experience lag


Lift


High-Speed Gust: Twist-Flap Coupling

• Active Pitch Control will always experience lag• Passive flap-induced twist will

always respond faster than an active pitch system


Comment on the Governing EquationsThe Common Formulation of Twist-Flap Coupling Used

in Recent Years is Incorrect

Inverted, this yields

−

−=

x

xx

GJggEI

TM

φκ

−=

TM

GJg

gEI x

x

x

1

1

11

2

2

2 α

α

αφκ

GJEIg⋅

=αwhere

( )21 ακ

−=EIM x

xWhen T=0, then which is incorrect!

It can be shown that the classic, well known relationship

EIM x

x =κ

is always true, even when coupling is present


Comment on the Governing EquationsCorrect Formulation

Inverted, this yields

=

TM

GJg

gEI x

x

x

1

1

2

2

α

α

φκ

−

=

x

xx

GJggEI

TM

φκ

α 211

When T=0, then which is Correct!EIM x

x =κ

Previous Conclusions that Twist-Flap Coupling Introduces an Inherent flapwise softening due to the 1-α2 term are incorrect.

No such softening occurs.Rotating axial fibers will obviously reduce the axial stiffness, but no

additional softening occurs due to coupling.


Comment on the Governing Equations

We could use a formulation whichavoids the use of EI, GJ, or g

=

x

xx

KKKK

TM

φκ

2212

2111

This is essentially what we do when we build the stiffness matrix in ADAMS from ANSYS results

We calculate the K terms from ANSYS and plug them into ADAMS

The error is in how we relate the K terms to EI, GJ, & g


If you use the incorrect formulation consistently between models (e.g., ANSYS & ADAMS), you will obtain consistent and correct answers in terms of

deflections, dynamics, etc.That is, Two Wrongs Seem to Make it Right!

Incorrectly putting the 1-α2 term in the Denominator of the Flexibility Matrix Results in an Overestimation of

EI by the factor of approximately α2. However, putting the 1- α2 term in the denominator overestimates the flexure resulting from a given EI by the same factor!



However, if you calculate E and I from a cross-sectional equivalent beam property model, you must

use the correct formulation of the governing equations in order to get the correct answer

Similarly, if you hand EI data calculated using the incorrect formulation to someone using the classic, correct definition of EI, they will obtain a different,

inconsistent, and incorrect answer.

This problem needs to be corrected.



Load Trimming due to Coupling actually permits flapwise softening of the blade without an increase in tip deflection.

Substantial Reductions in Flapwise Fatigue can be Achieved with Coupling

-0.397α + 0.970-0.352α + 0.936-0.347α + 0.912-0.364α + 0.888-0.333α + 0.856-0.392α + 0.870-0.349α + 0.817

60%

80%

100%0.0 0.1 0.2 0.3 0.4 0.5 0.6

Twist-Bend Coupling Parameter -- α

Nor

mal

ized

Fat

igue

Dam

age

Equi

vale

nt B

lade

R

oot F

lapw

ise

Ben

ding

Mom

ent

}}}

∆m=0

∆m=-15%

∆m=-30%

∆m=+15%

∆K11=0∆K11=-15%∆K11=-30%

90%

100%

110%

120%

130%

140%

-50%-40%-30%-20%-10%0%

Change in Blade Flapwise Stiffness -- ∆EI

Nor

mal

ized

Max

imum

Tip

Def

lect

ion

-- x

tip/x

tip,b

asel

ine

α=0.5

α=0.35

α=0.2

x ∆m=-30%+ ∆m=-15%-- ∆m=0% Results of Parametric Studies of

the Dynamics of 37-m Twist-Flap Coupled Blades on a 1.5-MW Pitch-Regulated Turbine

Thanks to GE Wind & Windward Engineering for their Support

Influence of Coupling on Dynamics


37-m Blade Structural Design Studies

0.000.050.100.150.200.250.300.350.400.450.5095

52G

9550

G95

50C

9532

G95

30G

9530

C97

52G

9750

G97

50C

9732

G97

30G

9730

C90

52G

9050

G90

50C

9032

G90

30G

9030

CX

30C

X50

GC

arbo

n 1

Car

bon

2X

5XG

Cases

Effe

ctiv

e Tw

ist-B

end

Cou

plin

g Coupled Skins & Caps

Skins Only

Spar Only


Influence of Coupling on Cost & Fatigue

X5XG

9730G9732G

9730C

9752G

9530G9530C

9750G

9750C

9532G9550C

9550G9552G

9030C9030G

9050C9032G

Carbon2

9050G

9052GX30CX50G

CarbonBaseline

70%

75%

80%

85%

90%

95%

100%

100% 105% 110% 115% 120%

Blade Cost Relative to Baseline

Bla

de R

oot F

lapw

ise

Ben

ding

Fat

igue

DEL

R

elat

ive

to B

asel

ine

Coupled Skins Only

Coupled Skins & CapsCoupled SparOnly

$8/lb Carbon


Influence of Coupling on COE

70%

75%

80%

85%

90%

95%

100%

9552

G95

50G

9550

C95

32G

9530

G95

30C

9752

G97

50G

9750

C97

32G

9730

G97

30C

9052

G90

50G

9050

C90

32G

9030

G90

30C

X30

CX

50G

Car

bon

1C

arbo

n 2

X5X

G

Cos

t of E

nerg

y R

elat

ive

to B

asel

ine

Coupled Skins & Caps Skins Only Spar Only

Includes influence of coupling and massGlass Baseline is at 104%


Influence of Carbon on COENo Coupling

4

5

6

7

8

9

10

11

30 35 40 45 50 55

Blade Length

Bre

ak-E

ven

Car

bon

Cos

t [$/

lb]

Aggressive Designs

Moderate Designs, Increased Diameter

Moderate Designs, No Increased Diameter

Conservative Designs

Current Carbon Fabric Prices

Automotive Carbon Prices


Influence of Coupling on LoadsPitch-Regulated Design Example

0.0

0.2

0.4

0.6

0.8

1.0

1.2

YawMoment

TiltMoment

TT Thrust Hub YawMom

HubPitchMom

LSSTorsion

Root FlapMom

RootEdgeMom

RootTorsion

Nor

mal

ized

Ext

rem

e Lo

ad M

axim

a

GE37a Class3A KW37 Class2A & 3A KW39 Class2A & 3ABaseline 37-m 37-m TBC 39-m TBC

Class IIIA Class IIA Class IIA Baseline 37-m 37-m TBC 39-m TBC

Class IIIA Class IIA Class IIA

Demonstrated Potential to Reduce or Maintain Key Extreme Loads Close to Baseline Values

0

0.2

0.4

0.6

0.8

1

1.2

Flap Moment EdgeMoment

Hub YawMoment

Hub PitchMoment

Yaw Moment Tilt Moment Tower TopThrust

Nor

mal

ized

DEL

37a Class3A KW37 Class2A KW39 Class2A

Key DELs Kept Below Class3A Baseline Values, Even at Class 2 Mean Wind Speeds and Larger Dia.

Baseline 37-m 37-m TBC 39-m TBCClass IIIA Class IIA Class IIA


9.2-m Blade Flapwise FatigueStall-Regulated Turbine (Sandia LIST)

86%

88%

90%

92%

94%

96%

98%

100%

102%

5 7 9 11 13 15 17 19 21 23 Total

Wind Speed Bin [m/s]

WSU 9.2cWSU 9.2TBC10

m=10

WS9.2c is an Uncoupled Carbon/Glass Hybrid BladeWSU9.2TBC10 is a Twist-Flap Coupled Carbon/Glass Hybrid BladeNPS9.2 is the Baseline, Uncoupled All-Glass Blade


Peak Carbon Fiber Strain, 9.2-m Blades

Coupled Blade

Max 2301 µstrain

52% Margin on Static Axial Strain

65% Margin on Axial Strain Fatigue

Uncoupled Blade

Max 3444 µstrain

2% Margin on Static Axial Strain

8% Margin on Axial Strain Fatigue


Shear Stress, 9.2-m Coupled Blade20° Carbon Fibers

32.2 MPa Peak Shear Stress

8% Margin on Static Shear Strength

Shear Fatigue Allowable is Unknown

0° Glass Fibers

21.6 MPa Peak Shear Stress

-7% Margin on Static Shear Strength

Shear Fatigue Allowable is Unknown


Twist-Flap Coupled Blade StudiesConclusions to Date

•Cost of energy benefits seem to improve with increasing coupling – no interim sweet spot has been found•Shear strength and fatigue issues will probably be

the key design challenge•Coupling in the spar caps and skins is preferable

to confining it to either region•Coupling in the spar caps only is preferable to

coupling in the skins only•Must maintain some ±45° material in the skins•Benefits of coupling for stall-regulated turbines is

questionable

large & small blades: design &...

Documents