Journal bearing wiped, l/h liner found loose!! Problem??

[Ayden]

We just wiped a journal bearing at our Turbo-Generator.
We bumped the rotor to verify speed indication on the panel but the machine stopped rapidly.
We decided to slow roll the machine to 500 RPM and monitor for noises, rubs etc... after 3 min. into slow roll at 500 RPM the Bearing RTD went of the chart at > 300.deg. F
Yes we wiped the bearing but why?
We opened up the bearing #1 where we got the alarm from and found upper half wiped, when lower half was removed we noticed the liner was loose!!! Could this be an assembly problem or was this caused by the heat generated from wiping. I really doubt that the liner came loose because of liner heating up.

What do you think??

 

[JJPellin]

I have seen bearings on large compressors fail on slow-roll before. The worst case was the 17,000 HP compressor on our FCC main air blower. We neve did discover the cause of that failure. Some people speculated that 500 rpm was too slow for that massive rotor. I don't agree. It was a new bearing, but had been run for a few days during unit start-up previous to the failure. When you say that the liner was wiped in the top, I am confused. Was it also wiped in the bottom? Of does your machine configuration result in the shaft running in the top half? If the bottom half was wiped, I am not surprised that the lower half of the liner was loose. We regularly see bearing housing bores collapse in at the sides from distortion caused by the heat of a failed bearing. With very limited information, I would believe that the loose liner was a result of the failure, not a cause.

 

[electricpete]

The liner I assume is the steel shell upon which the babbit layer is deposited. Liner would be set into a housing with an anti-rotation pin to ensure liner doesn't rotate within the housing, even if loose. Normally the consequence of loose liner would just be vibration.

Unless there was evidence of liner spinning within the housing (that should be obvious), I have a hard time imagining how loose liner causes bearing to wipe. Either loose liner a consequence of wiped bearing or unrelated observation.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[JJPellin]

You are correct Pete. Even if the liner becomes loose in the bore, it should be prevented from rotating by a pin or other device. A loose liner could be a problem in a few other ways. The liner could be flexing slightly with the looseness. Babbet is easy to fatigue and can crack. Without a good tight fit, the heat generated in the bearing is not conducted away as well and the bearing may run hotter. But most of the heat is carried away by the flow of oil. Relative movement between the liner and housing can result in fretting corrosion. So a loose fit tends to get looser. And the movement can damage the housing requiring expensive welding and machining to repair. But none of this would tend to cause a rapid failure of a bearing on slow roll. So this all leads back to our common belief that the looseness is a result not a cause.

 

[Ayden]

Thanks for the responses but having a loose liner in it's casing wouldn't built the necessary pressure to get that oil film underneath the rotor. The bearing is a pressurized bearing, it has a leading and trailing edge on which the rotor will rotate. Having that loose liner won't build the right pressure and remember that we bumped the turbine to verify speed indication. The turbine speed was recorded at 90 RPM, dropping with 20 RPM at the time to a stop.
This is a new machine, new bearings.
The babbit in both halves have been wiped. The top half liner is secure in his shell but the bottom half is loose.
Heat has been generated at both liners, top and bottom half but only the bottom one is loose.
Again I'm finding it hard to find that the liner being loose is a result and not the cause.
We verified with a pressure gauge that we have enough oil pressure going to the radial bearing.

 

[electricpete]

I only work with plain hydrodynamic sleeve bearings on motors, not hydrostatic (forced oil) bearings.

Just to clarify, can you explain a little more why you said that a loose bearing will interfere with proper pressure?

A - allows the pressurized oil supply to this bearing to leak before it gets to the bearing?
OR
B - Changes the internal geometry of the bearing which interferes with converging oil wedge?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[JJPellin]

The force to lift the shaft off the bottom of the bearing does not come from the pressure of the injected oil. It comes from the rotation of the shaft and the viscosity of the oil. At very low speed, there is not enough oil pulled into the oil wedge beneath the shaft to support the load. That is why we set minimum speeds for slow roll. How slow is too slow depends on the bearing design, the oil properties and the load (primarily the weight of the rotor). We used to coat bearings with STP (very thick, very viscous) in order to try and support the shaft during start-up conditions. We no longer do this since we found the additives in the STP interact with the additives in the oil. 90 rpm is very slow for a bearing of this type. You stated that the babbet was wiped in both halves of the bearing. In the example that I mentioned before, it appeared to be the same. But we found that the babbet had been wiped from the bottom half and deposited into the top half. It would take a very unusual event to wipe the top of a bearing, especially at this low speed. A loose bearing could interfere with the delivery of oil to the bearing if the liner shifted and blocked the oil port which is normally in the bottom half. If it was very, very loose, the oil flow might be diverted through the gap and spill out along the outside of the liner. But this is unlikely. I still feel it is loose from the heat of the bearing. If you still have the bearing, scrape on the top half and you may find that the wiped babbet there was babbet from the bottom half and the top never wiped. You can also measure the thickness of the wiped liner and compare it to the thickness of a new liner. If it is thicker, this is a buildup of deposited babbet and it is not wiped at all. I would look at the condition of the oil, the temperature of the oil. I would consider that the new bearing was defective out of the box and you were doomed from the start. Personallyl, that is what I believe happened with out compressor bearing failure. We were able to rule out everything else. So that was all that was left.

 

[electricpete]

JJPellin - what you describe is hydrodynamic bearings (pressure developed from rotation of the shaft and converging wedge action... dependent upon viscosity times speed over load, not dependent upon any external oil supply pressure). When the original poster said "pressurized bearing", I assumed he meant hydrostatic bearing. Maybe the original poster can clarify the role of pressurized oil in this bearing.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Mendit]

This may have been a plain bond failure of the babbitt lining to the shell. An error in manufacture of the bearing, like forgetting to stress relieve the babbitt layer can result in a bearing which is highly stressed yet passes all tests of manunfacture. The different coefficient of expansions between babbitt and steel can result in high stresses in the babbitt layer. You may want to check the roundness of the bearing shell. If the shell was not stress relieved after casting, the shell will be larger in the horizontal than in the vertical on the lower half. A photo of the disbonded bearing would help?

 

[spciesla]

Do you have any vibration data from proximity probes during the bump and slow-roll events. The DC gap voltage and shaft centerline plots should help you figure out how the rotor was moving relative to the bearing clearances.

Steve

 

[Ayden]

Let me try to explain the complete view of our sistuation.
We have a completely new Turbine rotor with integral thrust collar. We have axial groove bearing at the #1 brg. We modifite the active thrust brg to a tilting pad design. Everything else on the machine stayed the same. The only two modification made are the new rotor and new active thrust brg.
We noticed that the thrust bearing was consuming more oil than normal (bigger clearances. Because of this we installed a pressure gauge at the #1 brg liner to make sure that we have enough pressure at the liner and are not starving the brg. We slow rolled the turbine and wiped the brg at 500 RPM within 3 min. We inspect the damage liner together with bearing manufacturer of the thrust brg and liner brg. Brg manufacturer confirm their design is not the problem after running the numbers and because we were having enough pressure supplied to the #1 liner.
The second liner we installed we made sure together with bearing manufacturer that all clearances are good (max), installation is good, load at the bearing is right, crush good.
Bassically no real root cause found for the first failure.
We installed the second liner and started the slow roll process. This time we didn't reach the 500 RPM mark that the temperature spiked and we tripped at 380 RPM.
We are busy removing the brg cover and to verify liner condition. Pressure to the liner stayed steady (18-19PSIG).

What could be the cause?

 

[diamondjim]

We modified the active thrust brg to a tilting pad design. Everything else on the machine stayed the same.

Why???

 

[TPL]

Summary of Facts so far:

1st Bearing:
Top Half of bearing fixed and wiped
Bottom Half Loose

Bearing Manufacturer 'runs the numbers' and confirms that the design is OK

2nd Bearing: Allegedly OK in terms of fit, but tripped due to high temp at 380rpm during


Never, ever believe the manufacturer - if their design process is wrong, then running the numbers simply produces the same wrong design. Even if they are wrong, it is unlikely that they would admit it.

You should call in an independant specialist to assess the bearing and help to determine failure mode.

For a bearing to fail after such a short time does suggest that it is being starved of oil - the oil serves 2 purposes - to build the supporting oil wedge and remove heat.

In the first case, despite having an allegedly adequate supply pressure, how do you know that the oil wasn't passing through the loosenss/clearance - you might have the flow, but most of the lube oil is going down the path of least resistance.

How do you ensure that the bearing is actually wet (with lube oil) before the shaft begins to turn. A lot of damage can be caused during that intial turning phase if the bearing is dry. Some bearings have grooves to ensure that the oil reaches all parts of the bearing.

Are these brand new bearings or refurbished (rebabbitted)?

Let us know what you find when you get to the 2nd bearing.

 

[JJPellin]

Thank you for the additional detail. I would consider the following possible causes:

* If the oil was too cold, the flow of oil could have been too low and the bearing failed from low oil flow.
* If the oil was too hot, the viscosity of the oil may not have been able to support the weight of the rotor at that low speed.
* If the liner was not installed correctly, the oil passages in the housing might not have lined up properly to the oil passages in the bearing liner. Some liners cannot be reversed end for end. Some times the liner is different on the inboard than on the outboard. The passages in the liner might have been machined in the wrong location.
* If the bearing was defective out of the box, as mentioned, the babbet may not have had a good bond to the backing metal.
* If the turbine was not started up properly, the rotor could have bowed and taken out the bearing. This should have been obvious with very high vibration. Possibilities would be seal steam started with turbine not rolling, vacuum established without seal steam flowing, turbine warmed up when not rolling, rotor not centered in casing resulting in radial rubs.
* Slow roll speed too low. Our 17,000 HP compressor also wiped the bearing at about 500 rpm. We increased the minimum slow roll speed to 1000 rpm. But I find this hard to accept as a cause.
* Oil screens left in place. If the oil was circulated through screens, is there any chance that one was left in place? (downstream of the pressure gauge you mentined)
* Internal blockage in the oil passages in the housing. Even the pressure gauge that was added could have resulted in blockage (nipple too long, excessive thread sealant, installed in the wrong port, etc). If an internal passage plugged up downstream of the gauge, the pressure would still seem normal but no oil would be reaching the bearing.

I am not sure I understand why the new thrust bearing would consume more oil. In most that I am familar with the oil comes in through orificed ports. The oil is retained in the thrust bearing by floating brass rings between the thrust bearing and radial bearing and at the outboard end of the thrust bearing. If one of those brass rings was left out, the oil would pour out of the thrust bearing and might starve the radial bearing for oil.

 

[Ayden]

diamondjim: Everything else stayed the same. Because we had problems in the past with misalignment at turbine rotors we are now modifying to axial bearing designs.
TPL: Thank you for your responses. The looseness of the liner was not present at installation. Dry shaft maybe one problem we are seeing because we found that after having checked and corrected clearances at the liner within its shell, the only thing left was to install the bearing shell in the pedestal. These are brand new bearings.
JJPellin:Oil temperatures were good, pressure supply was good, oil passages were lined up correctly. We verifed and checked all possible scenarios.
JJPelin, you have a good piont on the start up of the machine. We will minimize the time for start up. This machine is an old machine with no turning gear. So there is no way to start without a bow, but we have several of these machines and this is the only one giving problems.
It would make it easier if we could attached pictures to this forum.

Findings second wiped bearing: Talking to other people, they are thinking about a clearance issue here because of the short time lived of the bearing.
We verified every clearance from the liner within its shell and THIS TIME we also checked the crush from the top cover with the bearing shell by using plastic gauges.
We NOTICED that we have excessive crush from the top cover and the bearing shell.
We are opening the clearances from the liner to accommodate the excessive crush from the top cover.

I will keep you posted on the outcome.

Thanks for the input from each one of you.

 

[Tmoose]

Plastigage reads (positive) clearance. Crush is negative clearance. Not sure how to test crush with Plastigage. Were you testing the gaps remaining at the parting face with the cap bolts finger tight?

 

[Ayden]

When using plastigage you install shims at the split of the area you want to check. Install the plastigage on top to get the max and also some at a 45 degree angle since your checking a radius.

* First check thickness of plastigage!

By inserting for example 0.005" of shim at the split and putting plastigage then tightening the bolts carefully you will get like for instance 0.003" of CLEARANCE.

Formula goes like Result = Clearance - Shims.
R= 0.003"-0.005"= -0.002" CRUSH.

This is how you determine the crush by means of comparison.

Regards,

 

[Ayden]

Again, thank you for the responses.

I have the update of this problem.

* We focused on the start up (JJPellin) and didn't warm up for long period of time.
* Because of the low speed failure we started looking at clearance possibilities.
* After having checked all clearanced from the liner with it's shell. We checked clearances out in the field.

bold

 

[Ayden]

disregards previous message I posted it by mistake.


I have the update of this problem.

* We focused on the start up (JJPellin) and didn't warm up for long period of time.
* Because of the low speed failure we started looking at clearance possibilities.
* After having checked all clearances from the liner with it's shell. We checked clearances out in the field.
We found that the top cover was crushing the bearing shell excessively. (0.006" in the 45 deg. angle)

We corrected this by opening up the liner clearances to accommodate the excessive crush from the top cover!!

We have a 5" shaft and as per OEM the clearances should be 0.006-0.008".
We opened the clearances by machining the liner up to 5.011" which is 3 over max at the top. We had like 0.014" at the horizontal split.
When closing everthing completely we had 5.008" at the top and decreasing when going towards the split line of the bearing.

Conclusion:
This failure was contributed by having excessive crush on the bearing. It was a clearance issue!!!

Kind regards,
Ayden

 

[Tmoose]

"The second liner we installed we made sure together with bearing manufacturer that all clearances are good (max), installation is good, load at the bearing is right, crush good. "

Didn't that one wipe too?