I am currently doing a bearing failure analysis on a pulper bearing that failed after only about 2 months service. This is new equipment on a new line. I have never done bearing autopsies on vertical mounted equipment before and I am wondering if some of the things I am seing are to be expected in vetically mounted machines. Should you expect to see the botom halves of both spherical roller bearings to be
worn much more than the top? Should you expect to see 360 degree wear with no clear load zone? Should you expect to see skidding wear marks on the ends of most of the rollers? We had a cage failure on the top bearing that let some of the rollers turn sideways and the brass cage was welded to both races on this half of the bearing, we also had a cracked inner race but don't know if this happened before or after the cage failure. Bot top and bottom bearings have a lot of wear on the bottom half and very little on the top half. It is possible we had a lube oil failure that caused the crash but are not sure. the bearings are 24148
I am attaching some pictures.

3302_A_pulper_bearing_autopsy_pictures.doc (1,034 Kb, 116 downloads)

I would say you would see a heavier load area on the bottom of both bearing's outer and inner race due to the weight of the rotor having to be supported by, basically, one side of each of the bearings in a vertical application as this is.

I would also say there was possibly a lube problem and also a water contamination problem. The races look to have corrision marks scattered over their surfaces(at least the ones that are not totally wrecked)

I see you welded a bead, I guess to relieve the fit to the shaft, around the inner race which appeared to be cracked, also one of the other pictures seems to show, it was also cut in two with a torch (in the upper portion of the picture of the cracked). Wonder why this bearing was so difficult to remove from the shaft being that the inner race was cracked( but I see where you have the picture labeled "inner race crack surface". Does this mean the ring was not completely broken in two?

It is possible the shaft size is a little too big for the bearing which would cause a decrease in the bearings clearance? I am sure without a doubt this is a fairly tight interference fit to the shaft, but too tight for what ever the "C" fit which might have been used, is not good, if the "C" fit was tight to begin with. It might be possible that this bearing is not the best suited for this application. A Timken tapered roller bearing might be a better design for this amount of load.

OMOAIMBW

Hi Randle,

Below are some question that can guide to focus to the posibilities.

1.Can you post the pic of the equipment?Is this a roller drum & the equipment is verticaly or the bearings?For a bigger bearings, tolerance is important for the fits- Shaft/journal/housing.
2.What is the operating temperature/speed? Do you think the new machine has fatigue load and is it up and running after the installation. If not ,how long it has been resting with the load.From the pic,only one of the raceway has been skimed,indicate rolling elemnt not rotating,instead skidding.

3.Temperature differnce has occured,it may due to secondary possiblities.

4. Since the bearing is cylindrical bore & from OEM, we should take in account that correct mounting method applied.i.e measurement of the shaft tolerance and check the bearings internal clearance after installtion shaould fall in a given min to max tolerance range.If you have a doubt, then do check this as per Ralph comment.

5. Do you have any CBM in place & are they getting this? If you do, then it should picked up the the impact signal of crack and if the noise floor is high , then it more on lubrication/skidding of rollers.

Joe

Thanks for your replies,
Ralph, we have encountered installation problems with this unit, the seals were bad allowing water to get into the unit and displace the oil. The OEM's performance and service have been terrible, we can't get any cooperation from them. Of the two identical units we have we have had many serious startup problems with both of them. The shaft fits are within bearing mfg specs but the housijng bores were undersized, one by about .004", and the other was egg shaped. The inner ring was cracked completely in two and the bearing was very difficult to remove from the housing-shaft. I have reservations about the suitability of sperical roller bearings in applications with a lot of axial load but we have a pulper on another machine (different mfg. and smaller) that has been running with this type of bearings for 26 yrs.
Joe, I will attach some pictures of the pulper agitator and drive. Shaft fits were good but housing fits were undersized. We did not check bearing internal clearances as it was interferance fit and no way to adjust clearance. Operating temps were high initially (175 F)but cooled down to around 100 degrees F. Speed of agitator is 140 rpm. The agitator is now running and seems to be very smooth and quiet, vibration data show low amplitudes. We did not have vibration sensors installed on this machine at the time of the breakdown and no way to safely access the equipment for readings. Unfortuanately vibration hardware was the very last thing the project purchased and I am having to install them all myself so a lot of it is not completed. The pulper agitators have taken a sudden jump in priority as you can imagine, and are now complete. There was a lot of heat used during the removal process as well as welding.
Thanks again for your input.

pictures_.doc (112 Kb, 64 downloads)

quote:

We did not check bearing internal clearances as it was interferance fit and no way to adjust clearance.

True there is no way to adjust clearances on this type setup, but, what if the shaft is "too large" from the OEM and the inner race is forced to remain at or near its enlarged size from heating it to install on the shaft, which will reduce the factory built-in clearances inside the bearing between the inner and outer races. This is checked by comparing the bearing ID to the shaft OD with micrometers.

Only a thought when a failure arises from no apparent cause (except maybe the water getting in the housing).

You do not have any data from before the failure on this pulper?

Ralph, thanks for your reply, when we install interferance fit bearings we always check shaft and bore sizes with micrometers but we don't check internal clearances with feeler gauges like you would for a taper bore bearing. We found the shaft fits to be good and when we found the housing fits to be undersized we honed them out to bearing mfg specs. Unfortunately we had no vibration data from before the crash as there was no safe access to the agitator while running and permanent mount accels were not yet installed.

Excellent. Sounds like you guys did a good job of rebuilding the pulper. Hope all goes well with it.

Hi Randle. You are correct that these brgs do not like high axial loads. Axial load should be no more than 1/3 of the radial load rule of thumb, otherwise rollers do not track properly and cage wear results along with point loading of rollers races etc. We have had 2 machines (both paper machine screens, vertical shaft)we have had to convert from this design. One to taper roller brgs and one to axial thrust spherical rollers. Maybe your other pulper has less axial load or higher radial load (from belts?)

Another thing, is the top brg the located brg? Is that the same as your other good pulper. Some manufacturers try to share the load by having all the radial on one brg and all the axial on the other. This wont alway work if using sherical brgs. You could try locating the bottom brg, then it will be getting plenty of radial load from the belts to compensate for the axial thrust. The top brg should be ok just floating axially. Then it will only see radial reaction load from the belts and no axial. Your brg supplier should be able to check the calcs.
Good luck.

Randle,

Can you post the PeakVue TWF from the spectrum you posted earlier?

Thanks again for your input, and sorry for taking so long to reply. Things are very hectic around here with the startup of our new machine.
Planty, the top bearing is the locating bearing in our new machine's pulpers. I haven't been able to determine whether that is true of the old one. It has been operating for 26 years with no problems. It has the belt pulley positioned between the two bearings so they both get radial load. Our bearing supplier has been involved with the rebuild of the new pulper agitator and assures us that it will be fine but I still have my doubts about the long haul.
Danny, I have attached Peakvue time waveforms and spectrums from both of our new pulpers. Pulper A is the one that had the failure and was rebuilt
I have been monitoring both these new pulpers closely and everything is looking good so far on the rebuilt one but the other one is showing some cage and outer race defect frequencies on the bottom bearing at very low amplitudes. It is scheduled to be rebuilt in October and both are to be converted from oil to grease lubrication.
It is thought that grease lubrication will be more reliable beacause it eliminates the possibility of pump failure or lines stopping up. Our bearing rep says we should pack it with a super thick grease and not even try to relube it. Does anyone have any thoughts on that?
Thanks again for all your help.

Document.doc (243 Kb, 15 downloads)

Hi Randle. Is that your new pulper with the pulley at the bottom.(not between the brgs) If it is and as you have confirmed the top brg is located I would suspect top brg is getting too much axial load. Cage failure is a clasic axial load type failure. It can also result from lube breakdown or lack of so you cant rule that out. You should tackle both. Oil is generally better but as you say relys on good sealing, however grease will only delay contamination type failures. Also beware if converting that you get the viscosity right and concider housing design regarding grease retainment and purge. For sealing a good labyrinth type is a must and even then grease/oil should be replaced on the shutdown following any reported gland leak. Another thing to think about is pulper blade setup. Can this induce extra axial load through not being set up properly? Hope this helps. Cheers.