Bearings are one of the critical components in rotary machines such as motors, wind turbines, helicopters, automobiles, and gearboxes. Most of these machine failures are caused by bearing faults. Thus, being able to detect bearing faults and predict remaining useful life (RUL) can help to provide advance failure warnings, plan the maintenance schedule, and avoid catastrophic failures. What are the most common technologies for bearing fault detection and RUL prediction?
In general, the common way of condition monitoring of the bearing includes
(1) measurement of the vibrations using the accelerometers etc,
(2) measurement of the temperature of the races (mostly by noncontact type/ indirect methods)
(3) measurement of increase in friction.
Sometimes we can also do an examination of the lubrication oil/grease for any traces of corrosion products.
Look at the work of Karsten Stahl, TUM. He is the expert in this field.
I would agree on vibration and acoustics as the first sign.
Markus Lienkamp Thank you very much for your suggestion and replies, Dear Prof. Markus! I will search for the work of Prof. Karsten Stahl. Have a good day!
s. C. S. P. Kumar Krovvidi Thank you for your replies, Dr. Kumar. I read some papers that try to monitor the health of bearing in induction motor using bearing current or stator current. Do you have any comments on this?
Around 30 years of work in rolling mill teached me that vibrations and acoustic methods are very difficult to use when the bearings are in gear boxes connected with other machines generating variable loads and vibrations.(like mills or gas turbines)
The early start of damage can be detected measuring the temperature of the bearings or of the lubricating oil (if there is a separate circuit)
Also the analysis of the oil contamination can help
Eliseo Galli Thank you very much for your answer, Dr. Eliseo. For the temperature approach, is there any suggestion on how to install/mount the temperature sensors?
Agreed on all of the above. Thermal measurements are useful (IR sensors for non-contact measurement is probably the easiest way).
If the lubricant is flowing oil (rather than simply greased bearings) any small particulate matter could be caught by a filter / magnet and inspected at regular intervals for visual analysis too.
Dear Zhenyu Zhao, That is indirectly checking the friction in the bearing. Frictional Force is proposnal to torque and in turn with the current drawn. We should note that the relation is not y=mx, but it is of y=mx+C. Because, in any induction motor, there is load current (x) and also magnetization current(C). Before we get a sensible variation in the current, we will find vibration (noise) or an increase in temperature.
Additional methods for specific cases not previously mentioned include:
Measurement of deflection (play) induced by wear of the bearings. In this case a large roller bearing supporting a cantilevered load that rotates through 120 degrees. Monitoring deflection changes between open and closed allows monitoring of bearing performance. it has been noted that prior to failure the deflection trend starts increasing quickly as the hardened surfaces become eroded and deteriorate at an accelearting rate.
Monitoring speed or cycle time is a indirect method of measuring frictional forces.
Monitoring the timing or movement differences between the input and output of the components can track the accumulated wear (backlash) and threshold changes over time can be in indicator of potential failure. although this is more for gears than the bearings in the assembly.
I think it is necessary limit a bit the field of work: are we speaking of bearing in big plants like rolling mills, power plants, ships, etc. or about bearings for fans driven by 0.5W motors inside a PC??
- Badges
- Science topic