Catastrophic bearing failure in Sundyne LMV 331 4

[mojtababalaj]

Hi everybody!
We have got a failure of four Sundyne pumps model LMV-331 (400 hp). It had an unusual pattern that we have never seen before. Bearing damage is evident on the lower input shaft in all cases. The pump had been running for about 10 days since the first start-up of the unit after overhaul. The pump tripped off on high vibration. The Input shaft bearings (Items 125C and D) are spherical roller bearings. 22310 E/C3 SKF for upper bearing and 21308 EW SKF for lower bearing are installed on low speed shaft (3000 RPM). In these failures, the lower input shaft bearings are failed dramatically. Pitting on the upper row of the rollers of the lower input shaft bearing (21308 EW) is evident. The failure is similar to surface fatigue. The pitting of the surface material is seen on the inner and outer cages of the lower bearing (21308 EW), too. It should be noted that other components of the gear box are completely healthy. Numbers of pictures of the failure are attached. the operating power of the pumps is around 240 kW.
How did you resolve this problem? Why may this failure happen?
Has anyone else experienced similar failure on the spherical roller bearings as I have described?
Regards!

 

[EdStainless]

Were these pumps part of the overhaul?
When was the last time that they were apart?
It really looks like either debris in them or they were oil starved.
Though I have see similar bearing damage when bearings were installed with grease that was too heavy and too much of it.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[mkfe2005]

It looks like spalling that is because of overload. Maybe the clearances between bearing and housing is not good. Sundyne says that the maximum bore diameter of lower bearing is 90.07 mm, but for upper, its maximum diameter is 110.02. however, axial load direction is in up direction, it seems that the lower bearing is tolerating the axial in your cases.

 

[Rasoul1250]

Dear mkfe2005
It is correct that sundyne mentions that lower seat bearing fitness should be considered +.07 mm, but this value seems to create some problems due to loose fit that causes relative motion between mating parts. In addition, we know that if this relative motion (between Bearing Outer diameter and its housing) is slight but continuous, FRETTING occurs. As a result, the mounting surface will wear and heat, causing noise and runout problem.
All in all, I suggest that the value between 0.02 to 0.03 mm is the reasonable consideration for lower SRB with the knowledge that the lower SRB should not bear the thrust load.
Best

 

[JJPellin]

I would very much recommend not changing the fit of that bearing within the housing. The bearing fits in a steel insert fitted into an aluminum housing. The failure is most likely caused by one of these:

• This bearing is lubricated by a small orifice drilled in a threaded plug. It might be restricted.
• The start-up procedure requires bumping the motor to build oil pressure unless equipped with a pre-lube pump.
• A motor running in reverse will reverse the thrust on the input shaft. Running backwards can result in limited flow and pressure.
• A properly performing internal oil pump will maintain between 20 and 60 psi.
• A bearing not fully seated on the shaft could overload as a result of thermal growth in operation.
• Incorrect setting of the coupling spacing can result in high thrust load.


Johnny Pellin

 

[Rasoul1250]

Dear Johnny Pellin
1- According to your experiences, what are the best bearing fits between for upper & lower SRB within their housings?
2- how much percentage can be probably related to the fake bearing?
3- would you explain, how can one diagnose coupling spacing has an incorrect setting?! for example in vibration spectra,1x and its harmonic or any diagnosis ways ...
4- In my experience, when we used ball bearings (In the last type of sundyne Gearbox 33x), bearing noise was lower than SRB. Basically is it true, or we are doing some mistakes causing noise and bearing vibrations?

Regards
Rasoul

 

[mkfe2005]

Rasul: you are probably right, 0.07 loose fit is too excessive in a high precision gearbox. In fact, I've never seen values more than 90.04 in a sundyne gearbox.
Lubrication problems aside, I dont agree with mr. Pellin. What I see is clearly depicting overload in axial direction on lower input bearing which is very unlikely to happen. Indeed I am getting confused how a bearing which is not axially loaded at all, is axially overloaded?!!
Considering the possible reason is coupling wrong axial set, this would result in either failure in motor axial bearing or failure in the lower row of lower input bearing (not the upper row).
Mojtababalaj are you sure the upper row is damaged?, If yes, there can be something holding lower input bearing outer race from upward axial dosplacement.
As previously mentioned, the only cause can be very tight housing fit, subsequent overload, excessive heat and finally seizure of the lower input bearing outer race.
Please let us know about sundyne experts diagnosis.
mkfe

 

[JJPellin]

I do not have a recommendation for the bearing fits. We only use the Sundyne specification for these fits. Because of the design of the steel inserts in the aluminum housing, it is nearly impossible to do anything to change these fits. I do not understand your second question. I don’t think you intended to use the word “faking.” Coupling spacing would not have characteristic vibration. But it can affect thrust loads which can contribute to bearing failure. I am not aware of changes in bearing noise associated with the change from ball bearings to roller bearings.

Mfke2005 is correct that the upper row of the lower bearing should not see high thrust load under any conditions. I cannot open your pictures. But, if that row is failed, I am even more suspicious of a lubrication problem. When Sundyne changed these boxes from ball bearings to roller bearings, they changed the recommended oil from an ISO 32 to an ISO 46. What oil are you using?


Johnny Pellin

 

[mojtababalaj]

I would like to thank everybody for the replies.
The bearing is lubricated by a plug at the middle of the bearing housing in front of the bearing lubrication groove. The bearing lubrication groove has three holes for better lubrication.
The unit is equipped with an axillary oil pump during start-up.
During the start-up, the pump pressure and flow were according to the unit datasheet and this means that the motor rotation is in the right direction.
The internal lub pressure was around 60 psig.
The bearings were installed properly so that the clearances between the shaft shoulder and the bearings were zero.
But I concern about the correct installation procedure of the coupling and I do not know how it can result in overload for the lower bearing. It does not make any physical sense for me. May I ask you to explain it more precisely?
The maximum bore diameter of lower and upper bearing is 90.03 and 110.02 mm, respectively.
We uses the recommended ISO VG 46 oil for better lubrication according to the Sundyne recently published documents.
Regards!

 

[Rasoul1250]

mojtababalaj
What is the value of input shaft floating? I guess if this value is considered more than 0.1 mm, upper SRB could not be fully seated on its housing properly (I think that input shaft thermal expansion is below of this value). This may be the reason of excessive thrust load transfer to lower SRB and so on...I do not know what is sundyne specification for this value?! It should be check.

Rasoul

 

[JJPellin]

Incorrect coupling spacing can place the coupling in tension or compression during operation. This can transfer thrust loads to the driver and to the gearbox. If the spacing was set too short, the weight of the motor rotor could be bearing down on the input shaft of the gearbox, adding downward thrust. A 400 HP motor has a heavy rotor. I don't know if it would be possible to overload the lower roller bearing by this method. But, if the lower bearing was overloaded, it should be the lower row of rollers affected, not the upper row.

Most machines of this type that I have seen use a strange coupling (Thomas DBZ) that can be tricky to assemble. It has a combination of thin and thick washers on each side of the disk packs that can be assembled incorrectly if the mechanic is not familiar with this design. If you are using a different type of coupling, then this may not be relevant for your machine. In any case, the spacing between the two coupling hubs needs to be properly set based on the requirements of that coupling to avoid transferring thrust loads to the gearbox input shaft.

Your description of the three lubrication holes sounds strange to me. Is this the standard arrangement from Sundyne or a modification? I checked the Sundyne manual and do not see the three holes you are describing. The manual describes a single lube jet at this location.


Johnny Pellin