Figure 1 - Vibration Trend
Figure 2 - Uptower inspection
Figure 3 - Borescope photo
Many designs of wind turbine mainshafts are mounted with double row spherical roller bearings for the main bearing. This bearing supports the majority of the radial load and the thrustload from the blades. The fundamental issue with the bearing selection is that the spherical roller bearing is a poor choice for supporting significant thrust (axial load). Romax have been supporting turbine owners, both in the USA and China, with root cause analyses for these main bearing failures.
Installed vibration equipment is used to perform diagnostics, along with SCADA data. This allows for a short list of turbines with potential problems to be developed. The Romax diagnostics team has found that standard algorithms on most CMS are not particularly good for the main bearing (figure 1) and have developed customized algorithms to improve damage detection.
A recent up-tower inspection identified 3 bearings that had failed and needed replacement this year and 4 bearings that were in earlier stages (Figures 2 and 3). From this information and prior records reliable failure rates are able to be established, supporting the warranty process. While up-tower the engineers test the auto lubrication system and take drivetrain measurements to allow development of a load simulation model for the main bearing and mounting system. A grease extraction port allows for borescoping of upwind and downwind sections
of the rollers, races and cage.
Simulation and experience allow for the root cause to be understood. The bearing, under high axial load, is in a condition where the load is concentrated on the downwind raceway/roller while the upwind raceway
is unloaded. The geometry of this type of bearing also causes significant sliding in the contact zone at both ends of the rollers. The rollers also tend to want to skew, which causes high roller/cage loads and high cage wear. This in turn leads to more sliding in the contact zone.
The combination of high sliding, high load and unfavorable lubricant conditions – such as insufficient fill and cold start, leads to adhesive wear and micropitting on the rollers and raceways. This is the principle first failure mode. The debris generated is then mixed in the grease and causes debris indentation on the rollers, cage and races, and subsequently the bearing suffers from surface fatigue and spalls. At a certain point large sections of raceway break off and this is when the bearing must be replaced.
By tearing down a failed bearing the root cause may be further proven. First the disassembly (Figure 5) must not damage parts and identification markings must be made for the as-assembled positions of rollers, raceways and retainers. The parts are cleaned in a wash booth then carefully photographed. Parts are sectioned with care to not overheat the area of interest and relax residual stresses or cause cracking. Evidence of excessive axial load is evident on retaining rings, raceway ends, rollers and cage. Upwind races/rollers are in ok condition while the downwind components fail. Scanning electron microscope work confirms the principle failure modes (Figure 6) to be adhesive wear, micropitting and ultimately surface fatigue.
The solution to the main bearing issue is a new bearing configuration, such as taper roller bearings. This is an option on the table for new turbines, but not for the installed fleet. Such a retrofit is complicated and expensive. There are solutions to extend the life of the bearing including a program of grease analysis, vibration monitoring, borescoping and thorough flushing of contaminated grease. Also some new main bearing products with diamond like coatings are coming into the market and have potential for longer life when installed
after repair.
Figure 6 - Scanning electron microscope work
Figure 4 - Bearing dismantling
Figure 5 - Downwind rollers