Modelling of static and fatigue failure in wind turbine blades using a parametric blade model

1. Modelling of static and fatigue failure in wind turbine blades using a parametric blade model A G Dutton, M Clarke1, P Bonnet2 Energy Research Unit (ERU) Rutherford Appleton Laboratory (RAL) Science and Technology Facilities Council (STFC) (now at 1Oxford Brookes University and 2SAMTECH Iberica) Presented at EWEC 2010, Warsaw, 23 April 2010

2. Background: SUPERGEN Wind Research Themes: • Baselining wind turbine performance • Drive-train loads and monitoring • Structural loads and materials • Environmental issues To undertake research to improve the cost-effective reliability & availability of existing and future large scale wind turbine systems in the UK

3. Background: Blade modelling • Which are the best materials? • What is the optimum lay-up? • What is the best internal structure? • What are the size limits for wind turbine blades? • What additional stresses do smart control devices generate in a blade? • How should NDT measurements be interpreted? Picture credit: LM Glasfiber Picture credit: EWEA

4. Parametric blade model:Design strategy • Parametric processing tool for creation and running of the underlying FE model • Suitable for sensitivity analyses, flexibility, documenting, re-usability • Python script front end for automation of the Abaqus FE package • Modular program • Realistic load application, including quasi-static aerodynamic loading • Ultimate strength & fatigue analysis • Developing dynamic implementation

5. a b => parameter sweeps: e.g. aerofoil shape tip deflection or max stress c d d - shear web offset (mm) Parametric blade model:Geometry definition

6. Parametric blade model:Geometry definition

7. Parametric blade model:Lay-up

8. Parametric blade model:Fully distrubuted aerodynamic load

9. Parametric blade model:Variable mesh density... ... at the push of a button

10. Parametric blade model

11. 5 MW (61 m) blade model • Basic lay-up information • Target mass and stiffness distributions • Limitations of lay-up information • Overall mass • Discretisation of lay-up info • Required spar-cap stress profile? • Lay-up modification • Materials variation • Static load case (aerodynamic load distribution) • Fatigue lifetime

12. 5 MW (61 m) blade model:Spar-cap stress distribution (smoothed)

13. 5 MW (61 m) blade model:Materials

14. 5 MW (61 m) blade model:Static strength – skins and shear web • Choice of static failure criteria: • Tsai-Wu • Tsai-Hill • Other (user specified)

15. 5 MW (61 m) blade model:Static strength – skins and shear web • Choice of static failure criteria: • Tsai-Wu • Tsai-Hill • Other (user specified)

16. 5 MW (61 m) blade model:Static strength – bonding paste • Cohesive element model • Normal stress component • Shear stress component • Linear up to characteristic value • Material “softening”

17. 5 MW (61 m) blade model:Fatigue strength estimation • Complex loading • Stochastic / semi-deterministic (cyclic) loading • Biaxial (triaxial) stress state IEC 61400-1 • Fatigue characterisation • Predominantly uni-directional materials data • Uncertainty in how best to combine different stress cycles • R-ratio (minimum:maximum stress in a load cycle) • Combine into constant life diagram…

18. 5 MW (61 m) blade model:Fatigue strength estimation Constant life diagram - Linear Goodman diagram

19. 5 MW (61 m) blade model:Fatigue strength estimation Constant life diagram - Multiple R-values diagram

20. 5 MW (61 m) blade model:Fatigue strength estimation Constant life diagram - Multiple R-values diagram

21. 5 MW (61 m) blade model:Fatigue strength estimation • Complex loading • Stochastic / semi-deterministic (cyclic) loading • Biaxial (triaxial) stress state IEC 61400-1 • Fatigue characterisation • Predominantly uni-directional materials data • Uncertainty in how best to combine different stress cycles • R-ratio (minimum:maximum stress in a load cycle) • Combine into constant life diagram… • …applies to a single material direction • How to deal with complex stress states? Biaxial stress ratio

22. 5 MW (61 m) blade model:Biaxial stress ratio Biaxial stress ratio is the ratio between the two largest magnitude principal stress components

23. 5 MW (61 m) blade model:Fatigue strength estimation

24. 5 MW (61 m) blade model:Fatigue lifetime Min: 1.6 x 1010 Min: 1.3 x 109 High performance glass fibre Baseline glass fibre Uniaxial fatigue

25. Full scale blade testingThermoelastic stress analysis Blade test: blade with defects Isotropic materials: Orthotropic materials:

26. Full scale blade testingThermoelastic stress analysis Blade test: blade with defects Blade model: normal blade Blade model: blade with defects

27. Conclusions • Flexible, parametric blade model for assessment of alternative materials • Simple failure model in blade skin and developing damage model in bonding paste implemented • Fatigue methodology under development • Initial results also available for application to full-scale blade testing, control of smart blades and interpretation of condition monitoring data • Future work planned on dynamic loading – operation in wakes from upstream turbines & “smart” blade devices

28. Acknowledgements EPSRC grant no. EP/D034566/1 SUPERGEN Wind Energy Technologies Consortium • For further information please contact: • geoff.dutton@stfc.ac.uk