Gearboxes in wind turbines have not been achieving their expected design life; however, they commonly meet and exceed the design criteria specified in current standards in the gear, bearing, and wind turbine industry as well as third-party certification criteria. The failures are widespread across manufacturers and turbine sizes, and the majority of these failures are not due to manufacturing issues. One of the basic premises put forth by the National Renewable Energy Laboratory (NREL) Gearbox Reliability Collaborative (GRC) is that this gap between design-estimated and actual gearbox reliability results from the absence of critical elements in the design process or insufficient design tools.
Key goals of the GRC are to improve design approaches and analysis tools, and to recommend practices and test methods resulting in improved design standards for wind turbine gearboxes that lower the cost of energy through improved reliability. The GRC uses a combined gearbox testing, modeling, and analysis approach, along with a database of information from gearbox failures collected from overhauls and investigation of gearbox condition monitoring techniques to improve wind turbine operations and maintenance practices.
Full-scale dynamometer testing builds an understanding of how selected loads and events translate into gear and bearing response, including reactions, load distributions, displacements, temperatures, stresses, and slip. Ideally, the knowledge gained from the GRC will result in any necessary improvements to gearbox design standards and associated modeling tools.
There are four main objectives of the GRC gearbox testing program. Testing to be conducted under this plan addresses some but not all of these objectives:
- Verify the gearbox operates within the basic design standards and assumptions, such as:
- Load distribution and sharing for the gears and bearings
- Structural deflections
- Bearing race temperature gradients
- Cooling capacity.
- Assess the effect of real-world operating conditions on the as-built gearbox and potentially extend the current design standards and assumptions to consider these effects, such as:
- Static and dynamic non-torque loading
- High-speed shaft misalignment
- Bearing rolling element skidding.
- Validate key portions of drivetrain, gear, and bearing analysis tools, such as:
- Dynamic response
- Generator coupling stiffness
- Bearing rolling element skidding.
- Develop a standard dynamometer testing and acceptance process, including tests such as:
- Non-torque loading
- Dynamic (field representative) loading
- Length of testing
- Modal testing.
Testing in the GRC project has included eleven test series to date, including tests that have been grouped under the nomenclature of Phase 1 and Phase 2 tests. This test plan describes the first of two test series under Phase 3. Phase 3 tests are currently planned in the National Wind Technology Center (NWTC) dynamometer as described below:
- Phase 3a. Test of gearbox 2 using the two-speed turbine controller that has been used in prior testing. This test series will investigate non-torque loads, high-speed shaft misalignment, and reproduction of field conditions in the dynamometer. This test series also includes vibration testing using an eddy-current brake on the gearbox’s high-speed shaft. This will permit operation at any generator speed between 100 rpm and 2,000 rpm at torque levels of about 5% of rated—sufficient to load bearings and gear mesh. Vibration testing enables the investigation of gearbox modal behavior and enhances the ability to duplicate high-frequency torque events observed in Phase 1 field testing.
- Phase 3b. Test of the newly designed gearbox 3 (GB3) in the NWTC dynamometer using the variable frequency drive. GB3 is a redesigned and rebuilt version of the GB1. It is expected to demonstrate a significantly better capability to withstand non-torque loads.
Other non-dynamometer tests also were planned for completion in 2013. Each of these tests is described in a separate test plan:
- High-speed flexible coupling stiffness measurement. The coupling is removed from the GRC drivetrain and installed in a test rig, which simulates radial (also known as parallel) misalignment conditions. A range of lateral forces is applied and the resulting deflections are measured, yielding a force-deflection curve that will most likely be non-linear. Determining the coupling stiffness enables a greater understanding of axial and bending loads induced in the high-speed shaft when the coupling is subjected to the radial misalignment. This information can be of direct use in simulating the radial misalignment tests described herein.
- High-speed bearing stiffness measurement. Both the cylindrical and tapered roller bearings used on the high-speed shaft are installed in a test rig, which determines the 5 x 5 stiffness matrix for each bearing. A range of radial forces is applied under different preload conditions and the resulting small deflections are measured, yielding a force-deflection curve. Once completed, the experimentally determined stiffness matrix is compared to theoretical predictions, likely resulting in a modification of GRC gearbox models.
- Static modal test. Accelerometers are attached to shafts inside the GRC gearbox and torsional vibration modes are excited with a hammer. Shafting is locked to prevent rotation and various torque settings are applied. This test is similar to a test previously conducted by Purdue University, but this test provides vibration data from internal components of the gearbox.
- Lubrication flow check. Delivery of lubrication oil to the gears and bearings has been predicted but not yet measured. This test uses several non-intrusive flow meters to measure flow through several of the branches of the lubrication system distribution tree. The flow meters are calibrated relative to a positive displacement flow meter currently used to measure total flow to the gearbox.
Click here for the rest of the 105-page report.
Filed Under: Gearboxes, News, Turbines