Does anyone know the formula for figuring oil flow? I have a roll with a 232/670 on each end. The roll is variable speed,but usually runs from 155 cpm to 160 cpm. The load is approximately 315,000 lbs per bearing. One bearing rep told us that it should be getting around 5 gal/min. Another suggested 2-1/2 to 3 gal/min. What would be the right flow?

David Eason

I'm not familiar with rolls and probably can't give you a useful answer. But can you clarify what type of bearing is that? Have you monitored bearing temperatures?

Pete,
The bearing is an anti-friction, double-row, cylindrical roller bearing. I'll try to attach a photo of one like it. Yes, I have been monitoring the temperature and it has varied from 119F to 151F over time.

David Eason

Mvc-001f.jpg (57 Kb, 55 downloads)

Since it's not a sleeve bearing, isn't temperature going to dictate the flow rate? Seems like the temps you're running now are fine. Are you having any problems?

All that aside, it sure looks like a big bearing and 5 gpm sounds better than 3 gpm.

Patrick,
Yes, it is a big bearing...$80K for a new one. I listed the high and low for the temps, but I forgot to say that it usually runs 125F-130F. And yes again, we are having problems with it. It has an outer race fault. I'll try to attach a spectrum. It has this strange "roller-coaster" cycle. For 4-5 days it will trend upwards, then for the next 4-5 days it will trend downwards. It is currently in a low.

David Eason

4_stack.doc (48 Kb, 36 downloads)

My uneducated opinion is that Patrick is right -those temperature are fine and an outer race fault wouldn't have anything to do with bearing oil flow or temperature... with the possible exception of going below minimum viscosity at operating temprature.

You wouldn't think a real fault would come and go unless the loading on the bearing is changing.

Also I assume with a bearing this big you are doing oil analysis looking for particles etc.

For the smaller higher speed stuff I work with we wouldn't even call that a fault, but I know your bearing is a different story.

Pete,
You are right; I wouldn't give that spectrum a second look in my small stuff either. Remember though, that the spectrum above was at a low. The waterfall below shows the cursor on that same low along with an indication of the swing it has been making. I also added a spectrum from one of the highs a couple weeks ago. The load has been holding fairly constant until last friday when they backed off of it some after changing out some rolls.

David Eason

trends.doc (84 Kb, 34 downloads)

Again the only obvious link between oil flow and bearing defects would be a minimum viscosity problem. With low speed and high load the minimum viscosity may be a challenge (although large bearing helps). Has anyone done an analysis of the minimum viscosity? What type of oil do you use?

We could begin to do a very rough viscosity analysis using SKF Bearing Installation and Maintenance Handbook. For a 1 meter mean diameter bearing at 150 rpm, their chart shows min viscosity about 22 cSt at operating temperature. At 150F you'd need at least VG68 assuming VI=0.95 mineral oil. Also some margin above the minimum is desired to provide long life and allow for less than optimum cleanliness etc.

Some limitations of the above analysis:
1 - It covers minueral oil but not synthetics. For synthetics there is wide variation in the pressure-viscosity coefficient which needs to be considered.
2 - There is only one chart given and I believe it applies to ball bearings. Cylindrical rollers act different.
3 - It assumes the bearing is lightly loaded.

There is a far more detailed analysis of minimum viscosity given in FAG Publication WL 81 115/4 EA Lubricatio of Rolling Element Bearings available for free at their site. It specifically addresses line contact (rollers) separately from point contact (balls). Also addresses pressure-viscosity coefficient. Also addresses load (you need to know ratio of load to rated load).

Also with oil analysis it would be interesting to know how clean the oil is (ISO cleanliness code). A heavily loaded roller bearing with thin film would be very sensitive to even small particles.

Looking at the picture I had two questions
1 - are you sure it is a cylindrical roller and not spherical roller bearing (barrel-shaped rolling elements). I have seen spherical roller bearings where inner ring can pivot like that but not cylindrical rollers.
2 - If I look through the center of inner ring towards the outer race I see what looks like a big honking spall. I'm sure that would have been noticed. Is that just the way the light hits it?

This message has been edited. Last edited by: electricpete,

E-Pete,
First, it is a spherical roller bearing. I don't know why I said cylindrical. The bearing in the picture I posted came out last year and you are right; that is a big honking outer race spall. That picture doesn't show the half of it and you should see the roller that was in it too. I have been trying to get to the bottom of this oil flow question, but I haven't given any thought to the viscosity. I will be out of the mill next week, but I will look into it when I get back. We recently hired a guy to be our "lube engineer." I'll get him to read this entire thread and see what he thinks.

One more point to make. Catching this fault before failure isn't the issue. A new bearing like this one costs about $80,000. If this bearing is pulled and rebuilt before it is damaged too badly, it can be done for about $40,000. I have the chance to save $40,000 bucks. The bearing in the picture was too far gone and it took only 5 days to get that way. I need to figure out where to draw the line on this one.

David Eason

I suggest you to read :
Design in Mechanical Engineering from
Shigley & Mitchell
Chapter related to lubrication
It will help to understand most of questions of you.
Regards

David, Here at our facility we run a similiar Bearing (232/560 CAK W33)on our 3rd Press section on a LNP roll, our flow rate is 5.2 GPM.

As far as how much the bearing cost, that seems like they might be getting to you a little on that price.
We run even bigger bearings than yours on some other equipment and don't pay near that much.
I would say you shouldn't pay over 60K-65K for that one.
This one here is 52K (232/560 CAK W33)

(EDIT

SKF recommends on our Bearing to be 4 GPM, we run 5.2 GPM.
I thought we were running 6.5 GPM but our Rotometers were not calibrated for our oil, so I had to do some checking, comes out to be 5.2 GPM.

Later, Jason

This message has been edited. Last edited by: Jason Pierce,