A wind turbine is essentially held together with bolts, the reliability of a joint possibly being compromised by bolt relaxation, vibration, fatigue, and corrosion. Because of the nature of wind turbines, they must be designed to tolerate variable and extreme vibration levels. Two studs, located on the blade to hub interface, have failed. The initial investigation involved verifying that the material is within the design specification. Having confirmed this, a root cause investigation was launched.
A wind turbine uses moving air to turn blades, which spin a shaft that is connected to a generator to make electricity. There are basically two types of wind turbine design; horizontal and vertical axis. The most common type is the horizontal axis turbine. The name comes from the fact that the shaft of the turbine is horizontal to the ground. All the components (blades, shaft and generator) are at the top of the tower, and normally face into the wind. Inside the nacelle (the head) is a wind vane (to determine wind direction), anemometer (to determine wind speed) and controller. When the wind changes direction a yaw motor turns the nacelle so that the blades are facing into the wind. In case of extreme winds, the turbine has a braking system that can slow the shaft speed.
There are many different types of electrical and optical sensors used in wind turbines. In general, they detect, monitor, and communicate information about parameters. These parameters include changes in the distance between two components, levels of vibration that, if excessive, can cause major damage and changes in temperature, pressure, and mechanical stresses.
The engineers at One Eighty received three studs – the two failed studs as well as an additional un-used stud.
Figure 1: Representative micrograph of the studs, showing tempered martensite.
• Characterising fracture face defects and morphology on both failed studs (macroscopic level using Stereo Microscope).
Figure 2: Stereo micrographs showing ratchet marks, final fracture and beach markings on Failed Stud 1 (left) and Failed Stud 2 (right).
• Characterising fracture face defects and morphology on both failed studs (microscopic level using Scanning Electron Microscope)
Figure 3: SEM images of ratchet marks (left) and striations (right).
The material is within specification, with no material degradation visible. A summary of the results follows:
Ratchet marks, beach marks and striations are visible on both samples and indicate fatigue. No evidence of fast brittle or yield in ductility was observed. It is therefore concluded that the failure mechanism was due to fatigue.
It is furthermore noted that both samples show low nominal stress characteristics, with medium to high stress concentration. One sample failed because of rotational bending, while the other sample failed because of unidirectional bending or repeated tension-compression cycles.
Figure 4: Schematic diagrams showing the forces active on unidirectional and rotational bending.
Fatigue failure in rotating bending or unidirectional bending is a result of repeated stress cycles applied to the material that are greater than the materials fatigue endurance limit. In order for a fatigue failure to occur the bolts must have been subjected to cyclic stresses for which they were not designed. Since there are no metallurgical defects present in the material it is concluded that the fatigue endurance limit has not been compromised in any way.
Inspection of the thread shows no evidence to support that the bolts were not properly fastened. No chattering, wear or surface damage is evident. However, this does not exclude assembly faults from the root cause of the fatigue failure. If the structure is not clamped securely, the resulting movement causes the studs to bend back and forth by small amounts. The more the bolt is stretched at assembly (producing more clamp load) the smaller the portion of external pulling load transmitted to the bolt and vice versa.
The nature of wind turbine operation is such that very little manual control occurs. Safety systems like sensors and overspeed protection devices limit the turbine to predefined operating parameters. Great pains are taken during the initial set-up phase to design a wind turbine with adequate protection and sensor systems, specific to the location. Operational and control deviations cannot be ruled out, due to the variability of the environment.
The fatigue failure therefore could only have been initiated by a system failure or assembly failure that enabled a cyclic load greater than the fatigue endurance limit on the bolts
Stud tensioning is a complex science that requires an experienced operator with a keen understanding of the process. It is recommended that repeated spot checks occur during installation to verify that best practices are being followed.
A re-evaluation of the design and operating conditions should be undertaken if repeated failures continue to occur.