I ti Bl dInnovative Blade Research: Passive Load

Reduction & 100-m Off h D iOffshore Design

Thomas D. AshwillWind and Water Power Technologies

Sandia National Laboratories

Fourth Sandia Blade WorkshopJuly 21, 2010

tdashwi@sandia.gov(505) 845-8457

Innovative Blade Research: Major FocusInnovative Blade Research: Major Focus

• Innovations that lead to longer &Innovations that lead to longer & lighter blades that reduce COE

• Designed & built several blade h lprototypes over the past several years

to demonstrate a variety of innovations

• This presentation focuses on results of two blade projects that demonstrate

i l d t l dpassive load control and a• Very large in-the-works 100-m baseline

blade design for offshoreg

OverviewOverview Background on Development of Bend-Twist Coupling

ConceptsConcepts TX-100 9-m research blade

• Goal: demonstrate passive load reduction using off-axis materials K&C 27-m STAR blade K&C 27 m STAR blade

• Goal: demonstrate passive load reduction using geometric sweep• Further the concept of growing the rotor with same loads

Both Projects R f Th D A h ill “P i L d C t l fj• Parametric studies• Design/Analysis• Fabrication

T ti

Ref: Thomas D. Ashwill, “Passive Load Control for Large Wind Turbines,“ 51st SDM Conference, Orlando, Florida, April 12, 2010

• Testing Static, Modal Fatigue Flight

Results and Summary

TX-100 9-Meter BladeT i C li U i Off A i FibTwist-Coupling Using Off-Axis Fibers

CX-100 baseline

TX-100 with off-axis carbon fiber

Twist Measurements DuringTwist Measurements During Static Load Test

6E-04

4E-04

5E-04

6E 04

/kN

-m)

FEAMeasured

1E-04

2E-04

3E-04k

(rad

/m/

0E+00

1E 04

3 4 5 6 7 8 9Span (m)

TX-100 Flight PerformanceTX 100 Flight Performance vs. CX-100 Baseline

100

120

CX - AverageMaximum

40

60

80

nera

tor P

ower

(kW

)

TX - AverageMaximumCX - 10-minAverageTX - 10-minAverageCX A

0

20

40

4 6 8 10 12 14 16 18

Ge CX_ Average

MaximumTX - AverageMinimum

20

30

nt, k

Nm

CX - BinAverage Max

CX - BinAverage Mean

Wind Speed (m/s)

10

20

Flap

Ben

ding

Mom

en CX - BinAverage Min

TX - BinAverage Max

TX - BinAverage Mean

04 9 14 19

Wind Speed, m/s

F Average Mean

TX - BinAverage Min

TX-100 Load TX-100 Load Reduction -Reduction-Azimuth

AveragesFatigue Cycle Counts

10%

0%

10%

20%

30%

omen

t (%

of A

vg)

8 m/s

1E-04

1E-03

1E-02

1E-01

1E+00

ty o

f Exc

eeda

nce CX-100

TX-100

7-9 m/s

-40%

-30%

-20%

-10%

0 90 180 270 360

Roo

t Fla

p M

o

CX-100TX-100 1E-07

1E-06

1E-05

0 5 10 15 20 25 30Root Flap Moment (kN-m)

Pro

babi

lit

Blade Position (°)

10%

20%

30%

(% o

f Avg

)

CX-100TX-100

p ( )

1E-02

1E-01

1E+00

ceed

ance

CX-100TX-100

11-13 m/s

%

-30%

-20%

-10%

0%

Roo

t Fla

p M

omen

t (

12 m/s1E 07

1E-06

1E-05

1E-04

1E-03P

roba

bilit

y of

Exc 11-13 m/s

-40%0 90 180 270 360

Blade Position (°)

R 1E-070 5 10 15 20 25 30

Root Flap Moment (kN-m)

STAR 27-m Swept BladeSTAR 27-m Swept Blade Part of DOE’s Low Wind Speed Technology (LWST) program to

advance technologies so wind systems can be more competitiveadvance technologies so wind systems can be more competitive in regions of low wind speed (5.8m/s). This will significantly increase available land area for economic development.

In 2004 Sandia contracted with Knight & Carver to In 2004, Sandia contracted with Knight & Carver to• develop a sweep-twist adaptive blade• reduce operating loads and allow for a larger more productive

rotor. Zond 750 chosen as baseline platform

• turbines are in need of blade retrofits• Z-48 (23.5 m blade)• Z-51 (24.5 m blade)

Sweep Design ProcessSweep Design Process

Zuteck, M.D., “Adaptive Blade Concept Assessment: Curved Planform Induced Twist Investigation ” SAND02-Curved Planform Induced Twist Investigation, SAND02-

2996, Sandia National Laboratories

STAR Airfoils & Planform Designf f g

ADAMS Simulations & FEM Analysis

PDF’s of Flatwise Moments & Performance Estimation

Baseline vs. STAR

Fabrication & Static Test

Fatigue TestFatigue Test

Tehachapi Sitep

View of the STAR rotor on turbine RP-09 test site looking west

View of the STAR test site looking north

Fl t i Fli ht T t L d C iFlatwise Flight Test Load Comparison

V G d L d C iVery Good Load Comparison Between STAR Blade & Shorter

Baseline Blades

Average & Maximum Operating Loads

Power Measurements vs PredictionsPower Measurements vs. Predictions

Annual Energy Improvements

Turbine Energy Compare

Group 1 (kWh)

RP-01 243009 101%

RP-02 236301 99%

RP-04 237148 99%

RP-08 242496 101%RP 08 242496 101%

Average 239738 100%

STAR 268711 112%

Advanced Rotor Development:100 m Sandia Blade Design100-m Sandia Blade Design

Goal: Provide technology research to produce innovations and advanced design concepts to develop very large utility-grade blade and rotor designs for offshore and onshore (where possible).

Methodology:gy• Develop and apply scaling laws to scale-up of 5 MW turbine system.• Create 13.2 MW Sandia Baseline (100 m long blade) with detailed composite

laminates• Apply innovative concepts to baseline to reduce weight, and improve

performance & cost effectiveness Partners: European UpWind Program and NREL

SNL Contacts: Tom Ashwill and D. Todd Griffith

Challenges & OpportunitiesChallenges & Opportunities for Large Blade Development

Challenges:

Blade weight growth

Opportunities:

Very thick airfoils for structural efficiency

Manufacturing & reliability issues

Material volumes & cost

Material lay-up & choices

Multidisciplinary design optimization

Blade jointsTransportation

Blade joints

Load alleviation concepts (active & passive)

Other innovations

Advanced Rotor DevelopmentpBlade Scaling and Design Drivers

ULUpscaled Length

3Mass:

Scale factor:

U

B

LUpscaled LengthBaseline Lengt

m m

h L

α

α

= ×

= =

2

Mass:

Rotor Power: U

B

B

U

m mP Pα

α= ×

= ×

Root Bending Moments:•Critical buckling load is independent of scale factor

3

4

Root Bending Moments: Due to Aerodynamic Forces:

Due to Gravitational Forces:

A AU UG GU B

M MM M

αα

= ×

= ×

of scale factor•Panel resonant frequency drops as one over scale factor, as do bending frequenciesq

Griffith, D.T. and Ashwill, T.D., “Development of the Sandia 100-meter Baseline Blade,” Sandia Technical Report, To Be Published 2010.

Typical Blade Cross-section forTypical Blade Cross section for 100-m Baseline

Scaling of Mass and Structural Properties to 10 13 2 & 15 MWProperties to 10, 13.2, & 15 MW

Blade Mass (kg/m)

2500.00000

Edge-wise EI (N*m^2)

2.00E+11

NREL 5MW

1500.00000

2000.00000

unit

leng

th (k

g/m

) NREL 5MWUpscaled 10MWUpscaled 13.2MWUpscaled 15MW

1.00E+11

1.50E+11

(N*m

^2)

NREL 5MWUpscaled 10MWUpscaled 13.2MWUpscaled 15MW

0.00000

500.00000

1000.00000

0 00000 0 20000 0 40000 0 60000 0 80000 1 00000 1 20000

Mas

s pe

r u

0.00E+00

5.00E+10

0 00000 0 20000 0 40000 0 60000 0 80000 1 00000 1 20000

EI

0.00000 0.20000 0.40000 0.60000 0.80000 1.00000 1.20000

Span (%)0.00000 0.20000 0.40000 0.60000 0.80000 1.00000 1.20000

Span (%)Flap-wise EI (N*m^2)

2.00E+11

NREL 5MWUpscaled 10MW

5 00E 10

1.00E+11

1.50E+11

EI (N

*m^2

)

Upscaled 13.2MWUpscaled 15MW

0.00E+00

5.00E+10

0.00000 0.20000 0.40000 0.60000 0.80000 1.00000 1.20000

Span (%)

Development of the Sandia 100-t B li Wi d T bi Bl dmeter Baseline Wind Turbine Blade

Process Future Work•Available models upscaled to provide a means to provide preliminary analysis of certification design loads

•Application of innovations to baseline to reduce weight, and improve performance & cost effectiveness, g

•Detailed lay-up and external geometry designed as part of baseline

•Strength deflection fatigue and stability

including use of

• carbon fiber & other advanced materials

•Strength, deflection, fatigue and stability analyses performed

•GL partial safety factors for loads and materials were applied and design found

• flatbacks & other new airfoils

• geometric sweep

b dimaterials were applied and design found to have acceptable margins

•One conclusion: edgewise strains & buckling are important design drivers

• pre-bending

• passive load control

• active load controlbuckling are important design drivers active load control

• more efficient design architecture

Summaryy Both the TX-100 and STAR blade demonstrated that bend-

twist coupling can be implemented in an effective mannertwist coupling can be implemented in an effective manner The STAR rotor increased average energy capture by 10-12%

compared to baseline Z48 turbines exceeding the design goals of 5-8%• measured moments were equivalent to those measured on

other Zond turbines and compared well with predictionsother Zond turbines and compared well with predictions 100-m Sandia baseline blade is being designed for use in

innovation studies

Acknowledgements to Test TeamsAcknowledgements to Test Teams Sandia/USDA Test Team

• Josh Paquette, SNLNREL Laboratory Test Team

• Scott Hughesq ,• Wesley Johnson, SNL• Adam Holman, USDA• Byron Neal, USDA

g• Jeroen van Dam• Darren Rahn

Byron Neal, USDA Knight & Carver Test Team

• Gary Kanaby, K&CKevin Jackson Dynamic Design• Kevin Jackson, Dynamic Design

• Eric Jacobson, Jacobson Tech. Services

• Mike Zuteck MDZ Consulting• Mike Zuteck, MDZ Consulting• Case van Dam, UC Davis• Terra Gen Staff

THANK YOUTHANK YOU