Hot Bearings & Lubrication

From Imrtwiki

Return to Medium or High Speed Rolling Element Bearings

Hot Bearings & Lubrication (When just a bit more is too much)

CM & Lube Newsletter Article Dec 2008

Many years ago I was walking through a plant with the local senior operations engineer and as we walked near a fan my attention was drawn by the huge air leak noise. What I saw was an air hose that had been wired on so that the leaking air was blowing across the fan rolling element bearing plummer block. The operations engineer immediately launched into his explanation for the obvious source of noise pollution and wasted energy. It seems that 6 months previously this bearing had developed a temperature problem and the ops engineer suggested with some pride that the application of his direct air cooling had ‘kept it going’ since then. I touched my hand onto the opposite side of the bearing housing from the cooling air and determined that the bearing was about 60 to 65DegC and the vibration was only moderate at between 3 to 4mm/sec. I used my listening rod (screw driver) on the bearing but heard no noises that would indicate a problem. As I was fairly young and inexperienced at the time, I was tempted to accept the provided explanation but it just didn’t seem to fit with what I knew about bearing failures.

Even as an inexperienced maintenance person I had seen a number rolling element bearing failures. Temperature had been the most obvious symptom of these problems and usually only giving a week or two warning before final failure. I had also seen a few journal bearing failures where temperature increase had only given a few minutes of warning before failure. What was occurring that would let a hot bearing live for more than six months? What occurred 6 months prior that caused the original temperature increase? Even though 65DegC is not usually considered a serious bearing problem, I was still curious to find a good explanation and later, started thinking more about what could have caused the temperature increase.

One thing I remembered was a story I was told a year or so before by a maintenance supervisor that looked after an area that included 3 large rolling element bearing fans. The supervisor suggested that these fan bearings previously had a serious temperature problems, until they found the trick of only installing 1/3 the normally recommended amount of grease into the bearing housing cavity. Could this current bearing temperature issue be just a matter of too much grease? I started looking around for other examples of rolling element bearings that were running hotter than expected but were showing no other failure symptoms.

There were three large 10,000GPM salt water service pumps in the power station where I was based. One of the outboard bearings was running hotter than the others and again it showed no other failure symptoms. The bearing housing looked similar to the pump on the right but the bearing end plate had a large screw plug installed in the centre. While the pump was in service I remove this end plate plug and shining my torch into the bearing housing I observed moving grease. I did the same observation for one of the other identical pump bearings that did not have the temperature problem and observed no grease movement. I had discovered what is usually called ‘Grease Churning’.

The most graphic verification of grease churning I have seen was on a fan from a plant a friend worked at. If you checked this fan first thing in the morning, all was OK. Usually at around 11.00am when the day had warmed up the fan bearing would shoot up by 20DegC and stay like that for the rest of the day. The temperature would be back to normal by the next morning. This was because when the grease temperature increased with the heat of the day the grease became more liquid, flowed into the bearing and started to be churned around with the rotation of the bearing. This up and down temperature pattern occurred for a few weeks until someone decided to pump some more grease into the bearing and from then on the temperature stayed high as the cooler temperatures at night were now not enough to stop the churning.

All the above examples of grease churning are from fans and pumps. This is because they are higher speed machines. Grease churning bearing temperature problems are not likely to be an issue unless the bearing Speed Factor [Bearing Internal Diameter (mm) x Shaft Speed (rpm)] > 40,000.

Once I understood what grease churning was, I became unconcerned at letting bearings that had moderate levels of churning stay in service. In many situations it is impossible or not worth the effort to remove excess grease without major invasive disassembly. If the bearing temperatures are less than 80DegC, there tends to be little short term risk with letting them remain in service apart from the energy waste. I have observed over the years that many people are not so comfortable with the explanation of grease churning and often have the desire to change out this type of hot bearing even when there are no other fault symptoms. Having to urgently repair bearing failures, usually at the worst possible time, generates real hardship for maintenance people and I think often indelibly imprints severe negative connotations related to hot bearings. On a number of occasions when explaining to someone that a hot bearing is just caused by grease churning, I can see in their eyes that I have not convinced them.

Identifying Serious and Urgent Bearing Problems

If there is any unusual temperature increase in a rolling element bearing the first question should be “is there a serious problem that has to be addressed urgently”. To answer that question you need at a minimum to have a high frequency vibration or ultrasonic meter available such the SPM unit to the right or one of the many other equivalents. The three types of urgent faults that have temperature symptoms are listed below and each will cause a significant increase in high frequency vibration.

Friction from a bearing fault/ damage/ debris
Excessive loading on the bearing especially from misalignment or misassembly
Inadequate lubricant

It is ideal if the bearing in question is on a vibration monitoring program or if there has been some baseline vibrations previously recorded. This will allow easy identification of a serious problem through parameters such as RMS acceleration, HFD, Spike Energy, Peakview, demodulation etc. If there has been no previous recorded information then as suggested above, a high frequency vibration meter should be used. These meters usually come with a standard for assessing bearing condition but more important are comparisons between the hot bearing and other bearing on the machine and to similar bearings on other machines. It is useful to determine the location on the housing with the highest vibration level and how the high frequency vibration level varies around the housing, as this gives useful information about the loading on the bearing (see figure below).

Housing temperature patterns will also give you bearing loading direction information (see the next section). As with all symptom changes you should ask “what happened at or before the time of the change” as part of your diagnostic process to start focusing on the root cause of the specific fault.

As well as checking the high frequency vibration of the bearing you should also be listening to the bearing. Most vibration meters allow you to connect a set of ear phones to listen and compare sounds while you measure. The other options that some prefer is to use a listening rod or stethoscope. There are many important symptoms such as rubbing noises, clicks, hums etc. that are often not detected through vibration analysis. The most important listening test to perform is the bearing greasing test using a grease gun. When grease is injected into a bearing there is a distinctive noise change and if there was a lack of lubrication there will also be a significant reduction in high frequency vibration. Once the noise change has occurred you should stop any further grease addition in case there is already an excessive grease quantity in the bearing housing.

If the hot bearing passes these tests there could still be a serious issue but it will not likely need to be so urgently addressed as the bearing is not under heavy mechanical stress.

Grease Churning and Other Causes for Hot Bearings

If the hot bearing is not under significant mechanical stress, the bearing Speed Factor is > 40,000 and identical or very similar bearings are running at a much lower temperature, then Grease Churning is the most likely cause. A very useful test to carry out is thermal imaging or more basic temperature pattern checks with a hand held temperature meter. Grease churning will generally give a very even temperature pattern around the housing circumference and will likely to be similar on both sides of the bearing housing (if accessible). An example is shown on the image to the right.

A bearing with a major difference in temperature around the circumference of the housing, as seen in the second bearing thermal image, is not likely to have grease churning. This image is an example of a bearing under a directional radial load and due to the axial offset of the hotter area, probably under an axial load as well. Thermal imaging is a very useful technique giving bearing load zone information, which can be very useful when trying to determine machine fault causes and root causes.

The third bearing thermal image below shows a temperature problem caused by the rubbing of a damaged housing seal. Another cause of rubbing faults is incorrect setting of fixed and floating bearings or housing mounting positions.

Other possible factors that should be considered when investigating hot bearings are:-

Speed Factor – rpm x d (d=Bearing Internal Diameter in mm)
Specification of lubricant (viscosity & grease stiffness)
External temperature sources
Housing design
Where the temperature is measured
Fault with the bearing cooling system (if applicable)
The type, size, clearances and loading of the bearing
Other bearing housing friction sources such as seals or excessive axial shaft float.
Temporary bearing temperature increases due to recent greasing.

Grease lubrication is not the only area that can see excess temperatures due to too much lubricant. Often it is easy to overfill oil compartments such as splash lubricated gearboxes that have rotating elements in an oil bath. For some oil sight glasses it is possible to get false low oil indications or have difficulty in reading the level. Sometimes people just like to put a bit extra in ‘just to be sure’. Again thermal imaging can be a useful tool for verifying compartment oil levels and interpretation of the splash direction and quantity through its effect on housing temperature.

Reducing or Eliminating Grease Churning Issues

The two largest negative effects of grease churning are the energy loss caused by the fluid friction heating and the reduction in the grease life caused by the increased temperature. With the greater awareness of greenhouse effect issues energy wastage has become a much higher profile issue for most businesses. For every 15DegC increase in bearing temperature above 70DegC the life of the grease halves. It is not unusual for bearings with too much lubricant in the housing to fail from inadequate lubrication due to the grease condition deterioration, even thought this sounds like a contradiction.

There are two general strategies for reducing or eliminating grease churning issues. The first is reducing the likelihood of getting too much grease in the bearing housing and the second is having an acceptable method of removing excess grease from bearing housing.

Making Churning Initiation Less Likely - There are a number of approaches listed below to minimise the risk of too much grease building up in the bearing housing. These approaches are desirable as they also tends to reduce the usage of grease and reduces lubrication costs.

Reducing initial housing fill volume
Reducing the frequency of relubrication
Using Ultrasonic Greasing to reduce the grease volume added during lubrication
Using grease flow meters to reduce the grease volume added during lubrication

Bearing housing grease fill volumes on initial assembly are usually specified at 30% to 50% of the housing volume. Some maintenance people seem to follow the rule that “if some is good, more is better” and overfill bearing housings. This is one of the main causes of grease churning. For bearings with a higher Speed Factor that will receive regular relubrication, 30% to 50% volume is much higher than is required and dangerous if the application is operating near the bearings ‘Limiting Speed’ for grease (see your bearing supply catalogue tables for Limiting Speed values). A 15% fill volume is usually acceptable for higher rotational speeds especially if the bearing will be monitored by vibration analysis, which can pick up any lubrication problems. If you think of the very small quantity of grease that is in a sealed bearing that, which lasts for a number of years, a lower initial grease fill volume does not seem such a significant issue.

Often the grease relubrication quantity specified is more than is required or even more often, it is not specified at all. The quantity of grease is usually specified in number of grease gun pumps. If there is any confusion with how to specify the volume to be added, grease flow meters can be added to the grease guns used. You should also think about the unfilled volume of the bearing housing, the length of time you would like it to remain in-service and how much you can add each relubrication to reach a 50% fill volume. If there is any concern that the grease lines to the bearing may not be full, then this can lead to excess grease volume added. Ultrasonic monitoring while greasing of in-service bearings is a recommended and successful way of knowing when grease being added has reached the bearing during relubrication, by the obvious noise change that occurs. If grease is being added next to the bearing (locations 2 & 4 on the housing to the right) then pumping should stop immediately grease enters the bearing. If grease is being added directly to the bearing (location 3 on the housing to the right), then the recommended volume of grease should be added after the grease is heard entering the bearing.

Grease relubrication intervals are typically very conservative and often assume more service hours and a higher operating temperature than is actually the case. Again I give the example of sealed bearings which last for years without lubrication problems. One key thing to be careful of is bearing operating temperature as the higher the bearing temperature the shorter the lubricant life will be. One method to help give the confidence to extend relubrication intervals is using ultrasonic monitoring in association with greasing. Any lubrication stress through extended lube intervals will be indicated immediately by much higher ultrasonic reading at the next relubrication.

Removing Excess Grease – Once grease churning has started the only way to stop it is removing the excess grease.

For plummer block bearing or bearings with accessible end plates, it is an option to open the bearing housing and remove some of grease directly. The key issue in this situation is to be able to guarantee that no contaminant enters the bearing. This requires meticulous cleaning around the housing and ensuring no dust or other debris will enter during disassemble. This process should not be attempted unless there is high confidence than any sealing disturbed during the process can be successfully re-established. Another issue to consider is the loading on the bearing and if the load could be released by opening up the bearing.

For any bearing that will operate with a higher Speed Factor (eg.>100,000) it is good practice to have a grease relief hole installed in the bottom of the bearing housing. Larger motors are usually always fitted with a grease relief holes to their bearing housings (see X-section above). You can also see in this figure an ‘Anti-churning vane’ on the top left side of the housing to discourage grease circulation. Motor bearings tend to have small grease housing volumes compared to plummer blocks and so it is easier for over filling to occur and so a grease relief hole becomes critical. If grease churning occurs then it only requires removing the grease relief plug, removing any hardened grease blocking the relief hole and letting the excess grease drain out during operation when the grease heats up. Adding a grease relief hole to plummer block housings is a standard modification and SKF and FAG bearing housings can be supplied pre-drilled. My recommendation is to locate the relief hole in an accessible housing location and have a threaded plug installed so that it can be used in the same way as that of an electric motor grease relief hole. The grease relief hole should be located on the side opposite the grease entry point to ensure the oldest grease is expelled first (see figure above). Another advantage of grease relief holes is that they can be used for taking grease samples for lube analysis to improve the confidence of any condition monitoring diagnosis or prognosis from vibration analysis data.

Article by Peter Todd - Industrial Maintenance Roundtable Facilitator NSW

Hot Bearings & Lubrication

From Imrtwiki

Contents