Improved Wind Turbine Drivetrain Reliability using a s Pure Torque design improved gearbox...

Click here to load reader

  • date post

  • Category


  • view

  • download


Embed Size (px)

Transcript of Improved Wind Turbine Drivetrain Reliability using a s Pure Torque design improved gearbox...

  • NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.



    Yi Guo1,*, Roger Bergua2, Jeroen van Dam1, Jordi Jove2, and Jon Campbell3

    1 National Renewable Energy Laboratory, Golden, Colorado, 80401, USA 2ALSTOM Wind S.L.U., 08005 Barcelona, Spain 3ALSTOM Wind, Richmond, Virginia, 23225, USA

    Presented at the ASME 2014 IDETC Buffalo, New York

    August 17-20, 2014


  • 2

    National Wind Technology Center National Wind Technology Center in Colorado (Photo by Dennis Schroeder, NREL 25915)

    Blade inspection (Photo by Dennis Schroeder, NREL 27204)

    5-MW dynamometer facility (Photo by Mark McDade, NREL 28218)

  • 3

    Standard design considers only torque on gears Aerodynamic loads, rotor overhung weight, and drivetrain weight are the main

    cause of nontorque loads (bending moments) Nontorque loads are present on three-point suspension, four-point suspension, and

    integrated gearboxes.

    Wind Also Makes Nontorque Loads (NTL)

    Blade/Hub Weight Bending Moment (Nontorque Loads)

    Gearbox Generator

    Main shaft

    Main bearing


    Rotor Hub

    Aerodynamic loads (wind shear,

    turbulence) Three-point mount drivetrain [U.S. Department of Energy, National Renewable Energy Laboratory (NREL)]

  • 4

    Study Approach: Combined Testing and Modeling

    Field testing at Ponnequin wind farm Dynamometer testing

    Computational drivetrain modeling , SIMPACK

    (Photo by Iberdrola Renewables, Inc, NREL 15191)

  • 5

    Many wind turbine gearboxes do not meet designed life At high torque, bearings fatigue and gears pit At low torque, bearings skidding and smear.

    Long-Existing Problem: NTLs Affect Reliability

    Risk of bearing fatigue

    Risk of tooth micropitting

    Planetary load sharing versus pitching moment Risk of bearing skidding

    Nontorque Loads Disturb Planetary Load Sharing and Elevate Tooth Edge Loads Guo, Y. Wind Energy, 2014

    Nondimensional pitching moment = Bending moment

    Rated torque

  • 6

    Reliability Issue: Bearing Fatigue

    Bearing fatigue at high torque Sometimes the highest torque is several times that of the rated torque.

    Experiment: Torque during a start-up (Up to 190% of the rated)

  • 7

    Design Solutions for Improving Reliability

    Solution A: Gearbox Reliability Collaborative (GRC) modified gearbox

    o Address the problem inside the gearbox o Stiff planetary gear section with tapered roller bearings o Rest of the drivetrain remains unchanged o Gearbox bearings carry nontorque loads instead of gears

    Solution B: Alstoms Pure Torque drivetrain o Address the problem outside the gearbox o Hub supports the rotor o Flexible hub coupling diverts nontorque loads to the tower directly o Nontorque loads on the drivetrain eliminated from the source

  • 8

    GRC: Address the Problem Inside the Gearbox

    Load share factor increases with carrier-bearing clearance o Optimizing clearance improves load share factor

    It converges to a constant when clearance is larger than bearing motion.

    Nondimensional clearance = Designed operating clearance Bearing motion amplitude w/ infinite clearance

    Reducing Bearing Clearance Improves Reliability New Design Under Manufactured

    [U.S. Department of Energy, National Renewable Energy Laboratory (NREL)]

  • 9

    Alstom: Address the Problem Outside the Gearbox

    Historically widely used configuration

    Pure torque configuration

    Load separation between drivetrain and structure Rotor supported by two tapered roller bearings on an extended cast-iron frame

    to carry rotor weight and aerodynamic loads Long track record, installed starting 1984.

    [Alstom Wind Power]

  • 10

    Study Approach: Combined Testing and Modeling Field testing at NREL Analytical model for load path

    Computational model for turbine response , SAMCEF

  • 11

    Field Testing

    Wind turbine: ECO100 60 Hz 3-MW Installed and operating since March 2011 Test campaign conducted during 2012 Instrumentation includes

    o 3 x blade-flap bending moment o 3 x blade-edge bending moment o Main shaft torque and bending moments o Tower-bending moments and torque o Nacelle accelerations o Front frame bending moments.

    [U.S. Department of Energy, NREL] Front Frame Gauge Layout Front Frame Gauge BD4

  • 12

    Fully coupled aero-servo-elastic simulation Multibody system with more than 3,000 degrees of freedom

    Computational Model



    Finite element model Super elements (Craig Bampton condensation)


    Dynamic link libraries (DLL)

    Blade element and momentum theory Specific wind turbine corrections

  • 13

    Analytical Model

    Main assumptions: o Rigid main shaft o Hub coupling stiffness is orders of magnitude smaller than main bearings o Carrier bearings have high stiffness and smaller clearances than the gear-set o Upwind carrier bearing heavily loaded compared to the downwind (sensitivity analysis).

    Reduced-Order Model

  • 14

    Result Comparison: Modeling/Testing

    Scatterplot for blade 1 flap and edge bending moments

    Blade 1 pitch stuck at 10 degrees Pitch failure event

  • 15

    Pure Torque Design Validation: Modeling/Testing

    Main shaft nontorque loads < 5% of torque during emergency stop o Orders of magnitude lower

    than other drivetrain designs

    Good model-test correlation.

    Preprint, Guo, Y.; Bergua, R.; et al. Improving Wind Turbine Drivetrain Designs to Minimize the Impacts of Nontorque Loads, Wind Energy

    Alstoms Pure Torque

    Main Frame

  • 16

    Sensitivity Study: Main Shaft Bending Diagram

    Carriers center of gravity (COG) and load share between bearings affect main shaft bending moment

    Main shaft bending moment could be further reduced with: o Locating the carrier COG located further downwind o Using a stiffer upwind carrier bearing than the downwind

    Carrier COG Carrier load share between bearings


  • 17








    Shaft BendingMoment/Rated Torque

    Planetary Load SharingFactor

    Gearbox Motion Alongy Direction

    Angular MisalignmentAround x Axis

    Alstom Drivetrain

    GRC Original Design

    GRC New Design

    Design Comparisons Alstoms Pure Torque design minimizes nontorque loads applied on the gearbox Original GRC three-point suspension is greatly impacted by nontorque loads New GRC gearbox design could be a solution for existing wind turbines.

    GRC original design as the reference



    y Mz



  • 18

    Conclusions Nontorque loads can significantly affect gearbox reliability

    o Gear tooth pitting is caused by fatigue o Bearing fatigue risk at high torque o Bearing skidding risk at low torque

    Three-point suspension drivetrain is sensitive to nontorque loads o Carrier bearing clearance plays an important role

    Alstoms Pure Torque design improved gearbox reliability o Nearly eliminated nontorque loads on the drivetrain (orders of magnitude lower than others) o Reduced gearbox misalignment o Unaffected by wind turbine load conditions

    Modified GRC gearbox design can be a solution for existing wind turbines o Stiff planetary gear section diverts majority of nontorque loads from the gears o Other drivetrain components can remain unchanged.

  • 19


    This work was funded under a Cooperative Research and Development Agreement between NREL and Alstom (CRD-10-400). The GRC design and test data was developed under a project funded by the U.S. Department of Energy under Contract no. DE-AC36-08GO28308 with the National Renewable Energy Laboratory.

    Contact information Yi Guo Email: [email protected] Roger Bergua Email: [email protected]

    Improved Wind Turbine Drivetrain Reliability using a Combined Experimental, Computational, and Analytical ApproachNational Wind Technology Center Wind Also Makes Nontorque Loads (NTL) Study Approach: Combined Testing and ModelingLong-Existing Problem: NTLs Affect ReliabilityReliability Issue: Bearing FatigueDesign Solutions for Improving ReliabilityGRC: Address the Problem Inside the GearboxAlstom: Address the Problem Outside the GearboxStudy Approach: Combined Testing and ModelingField TestingComputational ModelAnalytical ModelResult Comparison: Modeling/TestingPure Torque Design Validation: Modeling/TestingSensitivity Study: Main Shaft Bending DiagramDesign ComparisonsConclusionsAcknowledgments

    /ColorConversionStrategy /LeaveColorUnchanged /ColorImageAutoFilterStrategy /JPEG /ColorImageDepth -1 /ColorImageDict > /ColorImageDownsampleThreshold 1.50000 /ColorImageDownsampleType /Bicubic /ColorImageFilter /DCTEncode /ColorImageMinDownsampleDepth 1 /ColorImageMinResolution 150 /ColorImageMinResolutionPolicy /OK /ColorImageResolution 150 /ColorSettingsFile () /CompatibilityLevel 1.7 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /CreateJDFFile false /CreateJobTicket false /CropColorImages false /CropGrayImages false /CropMonoImages false /DSCReport