Can anyone tell the possible root cause of cage defect like attached? It's a taper roller bearing used on a reducer output shaft. Damaged only found at one column of cage while all others are OK.

Thanks, Kevin Ku.

Taper_roller_bearing_cage_defect.pdf (196 Kb, 199 downloads) Picture

Kevin,

Without additional information I'll be just guessing here. Possibly this is a manufacturing defect.

How was this failure detected? Any vibration data?

Noise heard about 10 minutes right before driving motor tripped out. No other abnormal reported.

Kevin

Kevin,

I've never seen a failure quit like this, but the damage is typical of a lock up of the bearing. This is usually caused by lubrication issues, however it is normally a total lockup that affects more than one section of the cage.

The picture shows what appears to be some corrosion on the surfaces of the rollers, which could have been initiated by water contamination in the lubricant and/or sealing issues.

Based solely on the picture provided my guess would be that a hard foreign material became imbedded in the soft steel of the bearing cage causing the localized damage. Was there anything in the lubricant to suggest this?

Was there any damage to the other bearing components, gears, seals, etc.?

Attached is some Failure Analysis literature from Timken that may be of interest.

Timken_Tapered_Roller_Bearing_Damage_Analysis.pdf (500 Kb, 100 downloads)

Michel,

the cage along with inner ring are only components saved for us study. Mechanics mentioned no oil leak and right oil level. None oil analysis done before. Our information is limited.

Kevin

I'll have to echo that I too have never seen a failure like this. The rollers look like they have been contacting bent cage for some time, so it doesn't seem that would fail a reducer. My estimate is that it could be an installation error, looks like a punch mark on the face of the cage, that smeared across the rail of the cage.

I would also think manufacturing, installation or handling error.

Possibly the cage was bent during stamping, maybe dropped or pressed or hammered onto the shaft rather than being heated.

Kevin,

quote:

the cage along with inner ring are only components saved for us study. Mechanics mentioned no oil leak and right oil level. None oil analysis done before. Our information is limited.

I hope we can all take away an important lesson here, namely

The first step in doing proper failure analysis is to PRESERVE THE EVIDENCE. Else we end up with guesswork and theories that cannot be verified

Cordoning off affected areas or parts, downloading event data, locking up the data files, taking witness statements without delay, photographs, oil samples etc. are all vital steps.

By the way, when the oil level looks perfect after a failure, check whether it is NEW oil, i.e., whether the oil was topped up AFTER the failure!

I'd ensure I was around to see first hand how they install bearing - don't say anything just observe.

Mike.

Thanks All of your input,
To resume production are always maintenance team's first priority. We ask them to carefully preserve the replaced parts but they seldom seriously follow or really have difficulty. Any good suggestion to overcome this problem systematically?

Kevin Ku.

Kevin what are you doing with this information? You can take very sharp pictures, use them in your reports.
Who are you reporting to? Do they go to management and are they piling up in some drawer?

Make a template for you RCA reports, with:
1) Date
2) Equipment
3) As found conditions
4) Probable cause
5) Keep to facts
6) Supply some underlying theory about lubrication, heat generation, contamination etc..
7) Give recommendations.

8) publish it! Print it out, give it to the mechanics, operators, supervisors, everybody who comes close to the equipment.
In the beginning they will try to prove that you are wrong, but after 4 - 5 reports people will call you in to take pictures when they notice that something unusual happened, the key is that when they see that something is done with the information to improve conditions and making their work visible, you will be amazed how much clues will appear.
Probably a lot of people already knew about the problem, but is was never properly documented and acted upon. Use the reports for sharing knowledge. The old saying, "a picture speaks louder than words" is still valid.

Kevin,

quote:

To resume production are always maintenance team's first priority. We ask them to carefully preserve the replaced parts but they seldom seriously follow or really have difficulty.

We never have time to do the right things, but always have time to do the wrong thing many times!

I agree we have to get the production going quickly. But does it take a long time to
- take some pictures
- download event data from the DCS
- talk to the people involved
- take oil samples
- get some initial ideas of probable cause

All these can be done by the operators themselves, but the problem in most companies is the great divide! We all seem to have the them and us problem. To get reliability improvements, we need the operators to work with us, not against us. There are many things maintainers can do to make this happen, but that is another subject.

Steven, you are spot on!

This is our company's product. And I happen to work in the facility where such failures are sent for evaluation. And I happen to evaluate such failures. I showed it to one of our seasoned field engineers. He said, "Yup, the ball was running over one of the cage bars".

It's really anybody's guess as to what caused it. My guess is that the bar was bent before or during installation. That makes it either a manufacturing defect or installation error. I find it hard to believe that the installer wouldn't catch this if it was a manufacturing defect.

Either way, it's an isolated incident. Next time inspect the bearing before installation, and make sure the bearing isn't traumatized during installation.

The bearing damage courses show that over 50% of premature failures in bearings are related to improper mounting practices.

BTW, excellent picture!!

Kestas,

Thanks for providing professional opinions to support those of us amatuers.

I particularly liked you field engineers comments.

Yes agree Danny - its a shame we don't have more experts like this contributing (for no cost)as it enriches the knowledge base.
How about some of you other product vendors lending us a hand? I'm sure you guys are out there watching
You never know it may pay back dividends?

Mike.

Don't sell yourselves short. I've read through some of the other threads, and you guys are pretty good, better than the average industrial user.

I read the thread on the broken shaft. I have good training and experience with fractography and you guys nailed it. There's nothing more I could add other than it is best to use angled lighting when presenting a fracture surface.

Sweet as Kestas! - just as long as you hop in there if we drop the hammer on our foot.

Mike.

As mentioned, cage failures usually are the result of no or inadequate lubrication but the rollers generally consume the entire cage assembly. The rollers exhibit some lines across the face of them similar to barring on a roll. I would like to see the roller path of the cone. There may be some evidence that will help with the autopsy. Cage damage is most likely the origin of the failure, but there appears to be enough clearance (end play) in the cup/cone assembly to allow the bar to ride over the roller. There may have been too much float in the shaft assembly leaving the cone to ‘bounce’ around in the cup. That looseness could have contributed.

As many mentioned, it is difficult to complete an analysis without all the parts and pieces.

Have a good one,

Gary B

One thing to remember when thinking about this failure, WHAT DOES THE REST LOOK LIKE. I see it this way,the cone/cage is damaged (by a rotational style lock up)but the cup section of this bearing was not able to be recovered.
Does this mean that there was excessive clearances as suggested? When looking at the rest of the bearing it does look to have excessive wear and a lack of lubrication (unless the rust marks occurred after the bearing was removed, which I very much doubt due to the positioning of the rust like indications and contact of the rollers on the inner raceway). I would also be making sure the preload on this bearing is set correctly.

What condition monitoring is being carried out on this gearbox or equipment? Do you have a VA program, Oil Analysis or thermo in place?

How to get maintenance team to preserve things etc? Good question Kevin! Try this
Step 1 - Ask lots of Questions every time there is a failure(the crew will come to expect this)
Step 2 - Repeat step 1 but throw in the silly questions like are you sure you drained the oil.
Step 3 - show you are really interested in what they have to say and get them to explain it to you again on completion before you walk away.ie; for this failure. Can I see the broken pieces? where is the old oil?did you flush the gearbox with a light oil to rid of contamination? what time did the failure happen? Was it a crappy job to do or were you happy to do it? what do you (the repairer) beieve was the root cause? can we avoid this from happening again?How?What can I do to make this job easier for you?Did we have the right tools for the job? do you think we could have predicted this failure? if so, how?(get them to tell you).

After all this the maintenance team will start talking about how you came down with heaps of questions. After One day you need to go back to them with some feedback. They crave it, gather those involved and tell them where you are up to. Even if you don't have any results just at that time, make the effort because it is still fresh in their heads and they may even give you a little more information that they couldn't remember the other day.

I guess if you show them you are serious about eliminating these types of failures by listening to them and maybe even making the job a bit easier for the future most people will stive to help. Feedback to the men/women I believe is the key. Even if you just thank them for the effort they put in to get the plant back up and running lifts the morale and shows everyone that the work they do is meaningfull.


Hooch

In direct contradiction to what I said but I'd have to say some really good points Hooch.
Thanks.
Mike.