A client in Eastern PA has a dust collector fan that lost it's bearings due to heat failure. These two units are sitting outside and drawing ambient air through them. There is no process heat source. The bearings are exposed to outside air which as been in the mid 30's this week.

The new bearings were set by myself to exact SKF spec. (benched at .003 and were reduced .0015") these are SKF 22215 bearings, double rollers.

Pillow blocks were squared to the shaft, plumbed and leveled. Every detail was checked and rechecked. Lube: the bearings were packed and then lubed per SKF standards with Mobile XHP222.

The fan is belt driven at 3035RPM. Belt alignment was perfect using a laser, and the tension on the belts was set a little lighter than the 11 that was called for on the Browning tool....we set them at 8 which gave a deflection of about .70".

The float bearing was set inboard near the fan housing with the fixed outboard near the sheave. The inboard fan float bearing went straight to 160 degrees upon start up and has stayed there since Monday night.

What are the normal things to look for when a new bearing just heats up like that for no apparent reason??? Then what else can we look for to eliminate the heat??? I have assured that the lubrication is not too light or to much...so aside from the actual grease type, I cannot believe that it is the issue.

What say you....I'm really scratching my head here.

Thanks

Just brainstorming some possibilities (I apologize if you have already ruled some of these out):

• Axial loading - Is at least one of the bearings free to move axially.
  
• Radial loading - Are there too many belts. Sheave too small.
  
• Clearance - is the internal clearance reduced through excessive interference at the shaft/housing fits (should not be if these have been measured in accordance with standards)
  
• Grease distribution - bearing/housing packed too full. Or else housing packed but none applied to bearing (I've seen that happen).
  
• Mixing of incompatible greases - can cause undesirable change in consistency.
  
• * Another factor I am increasingly tuned in to is the speed limits of grease lubrication (D*N number). 22215 double spherical roller bearing at 3000 rpm I think is at or beyond the limits of grease lubrication given in some references. Let me know if you want some references on this.

It looks like you have already addressed internal clearance, fits, axial float and grease location.

If you give sheave diameters, number, horsepower, we may be able to estimate loading to doublecheck if it is within bearing load rating.

Did the heat appear immediately upon starting? It's not uncommon for greased bearings to run hot for the first few hours and then cool as you probably know.

Was there a failure? Was a bearing pulled/examined or just noted hot bearing housing?

I would start with the last bullet (is this bearing suitable for grease lubricaiton at this speed)

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

Here is an excerpt from Tribology in Machine Design, Stolarski, ISBN 0 7506 3623 8

quote:

"The speed limitations for greased bearings are due mainly to a limited capacity to dissipate heat, but are also affected by bearing type and cage type. Standard quality ball and cylindrical roller-bearings with stamped steel cages are generally limited to 0.2 to 0.3 x 106 DN, where DN is a speed parameter which is the bore in millimetres multiplied by the speed in r.p.m. Precision bearings with machined metallic or phenolic cages may be operated at speeds as high as 0.4 to 0.6 x 106 DN. Grease lubricated tapered roller-bearings and spherical roller-bearings are generally limited to less than 0.2 x 106 DN and 0.1 x 106 DN respectively. These limits are basically those stated in bearing manufacturers' catalogues. The selection of a type or a classification of grease (by both consistency and type of thickener) is based on the temperatures, speeds and pressures"

Your bearing has a bore of 75mm at a speed of 3000rpm, giving a D*N number of 75*3000 = 225,000. That is more than twice the limit 100,000 given by the above author for spherical roller bearings!.

The appropriate limits may be subject to a little bit of discussion and certainly there are a lot of factors to consider (grease type, bearing dimensional precision have all been stated to affect the limit in various references), but you will find many other authors with similar numbers.

You should certainly give the OEM a chance to explain. You might find it hard to believe that OEM's could provide grease lubrication in a machine beyond the limits of grease lubrication. I don't. I have many of them myself. These are among our biggest trouble makers for long-term bearing health.

My immediate thought was also that this is too fast for grease.

I would abide by the OEM brg spec of limiting speed and clearance per ABEC#. Both grease and oil are listed.

1.5 mils seems too tight. I am thinking out loud as I'm in a hurry and no book by my side. However, by memory I'd say 2.5 min w/4.0 mils as max.

Pete,

Can you explain that formula further? What is the 106 before the DN?

Where did the limit of 100,000 come from?

To me this formula would mean that the limit in rpm for the bearing in question is .1 * 106 * 75 * 3000=2,385,000 rpm. That sounds a little fast to me, so maybe I don't fully understand your example.

Have you considered a bearing with more clearance, say a C3 or C4. Sounds like the one in there might be a C2. Another thing, was this 1.5 thousands clearance on both sides of the bearing? Sometimes the tapered sleeve might pull the thick side up tighter than the thin side, especially if something is a little wrong with the shaft, making that side of the bearing be way less than .0015, maybe even .000000. Which bearing failed this time, inboard or outboard?

May be a little too fast for grease like mentioned, but I have seen some pretty fast fans running grease with no problem, some up to 3600 rpm.

Gentlemen,

I was just researching the grease issue and came back to see that you all were on the same track.

I understand the D*N spec. In this case the bearing diameter is 2 7/16th or 61.9125 mm.

61.9125 X 3035 = 187,904

Most of the data I have seen from Grease manufacturers state a nominal of 200K - 300K for normal high temp grease. I could use some clarification on this issue as it seems to vary from Chevron to Mobile....etc.

Pete...if there is any additional information you can provide on this subject...that would be great.

Sam...These bearings benched at .003. The book calls for .0012" to .0015" reduction. Since I can't measure in 100 thousandths any better than you guys can I set the bearings to .0015" reduction because I can measure to .0015" with my feeler gauges. Therefore if I start with .003" and the book calls for .0015 max reduction then .0015" is where they were set.

I agree with you on the point that that seems like a real tight bearing. I used to set bearings to .002-.003....this bearing is one of the tightest I've ever seen.

Pete...to answer your questions: the inboard float bearing heated up to 160F real fast...about 3 minutes and has stayed there. The fixed bearing (near sheave) heated to about 130 and fell off after that to around 90-100F. The inboard bearing had failed on this fan. Thus the reason for the replacement.

An identical fan that sits not 10 feet from this unit has Link Belt 22215 bearings on it and the outboard bearing keeps failing...the last time it got so hot that the shaft actually bent 45 degrees at the driven sheave and caused alot of damage. This fan is also grease lubricated with the same Mobile XHP 222 and these bearings are currently running well under 90 degrees each.

Thanks for any help you can provide with the lubrication issue or any other thoughts you have on the matter.

quote:

In this case the bearing diameter is 2 7/16th or 61.9125 mm.

I believe for 22215, the bore is 75mm bore (not 62mm). Correct me if I'm wrong.

quote:

Most of the data I have seen from Grease manufacturers state a nominal of 200K - 300K for normal high temp grease.

That would be a typical range for ball bearings and is in fact the exact range stated for ball bearings in the quote directly above (0.2 - 0.3 x 10^6). But spherical roller bearings have lower limits as stated in the quote directly above (0.1 x 10^6).

Another reference: Handbook of Lubrication and Tribology, Vol 1, Chapter on Electric Motor Lubrication by Drew Troyer of Noria:

quote:

A motor equipped with radial ball bearingsmay be lubricatedwith conventional grease if the speed factor [Dm*N] is less than 340,000. If the motor is equipped with spherical rolling element bearings, conventional grease is suitable if the speed factor less than 145,000. If the speed factors exceed the limit for the application, the bearing will run hot and/or the grease will be damaged. If the bearing cannot be lubricated with conventional grease due to speed factor limitations, a specialty high DmN grease may be an option. Otherwise, it must be lubricated with bath oil, force circulation oil, mist, drip oil, or spray, all of which provide higher speed factor limitations.

Note he is using Dm*N which is a little higher than D*N (Dm is based on mean diameter, while D based on bore diameter). For your bearing the Dm is (75 + 130mm)/2 ~ 100mm. The Dm*N is 100 * 3000 = 300,000. It is still a factor of 2 higher than the Dm*N limit given in this reference (140,000). So both references I have quoted thus far have expressed the limits in slightly different forms, but they both predict your machine is a factor of 2 above the limiting speed for greased lubrication. If it were a ball bearing, you would be within the limit (Dm*N = 340,000), but a spherical roller bearing has a much lower limit. I have plenty more references like that in the pile.

My experience with high D*N greased bearings is primarily on 3600 rpm deep-groove ball bearing motors with bearing sizes 6313 - 6316. On the 6316's, we are lucky to get three years before we start seeing random vib or 2*BSF sidebanded by cage. When we change bearings, we normally see discoloration of the grease and staining of the bearing which suggest overheating to me. But we don't often observe the bearing housings to be running hot. That's a little bit of a mystery. I believe the explanation for the mystery is that grease lubrication is a very transient phenomenon. When the conditions get bad enough, the grease starts getting hotter which perhaps allows it to melt, redistribute or release some oil, which improves the lubrication conditions for awhile.

Also I believe these high D*N machines are much more sensitive to lubrication technique. Greasing is preferred often with very small amounts. (rather than infrequent with larger amounts). Housing should be packed only toward the very low limit. When the cavity gets near full, the bearing will die (unlike slower smaller bearings which are more tolerant).

quote:

May be a little too fast for grease like mentioned, but I have seen some pretty fast fans running grease with no problem, some up to 3600 rpm

I would like to explore your comment. Do you know the bearing style and size? And when you say "running grease no problem", do you mean that the machines weren't running hot or showing vibration symptoms at the time you monitored them ? Or that you monitored them over 5-10 years and observed reliable bearing performance?

From various bearing manufacturer's catalogs.
22215 maximum speed grease lubricated

SKF 3400 RPM
FAG 3400 RPM
Torrington 3080 RPM
NSK 2900 RPM

Looks like you are right at the edge. I'm sure that the specs in the catalogs are based on optimal conditions also.

dj

Pete,

I gleaned the answers to my questions from your reply to the OP.

No need for a response to me.

Thanks,

Pete,

That was more help that you know. Thanks much! I am getting a better handle on that formula and concept behind it. At any rate I think the general consensus of opinion is that this application speed borders on non serviceability with grease.

Would you look at changing the bearings or just set up the existing bearings with oil lube.

Also, with regard to your question about the 22215...I personally verified that the bearing was indeed an SKF 22215 PK double spherical roller bearing and the bore was 2 7/16".

Thanks for the info.

I was surprised to see as Djames said, this application is roughly "at" the bearing manufacturer limit for greased bearings, when the thumbules quoted above indicated it was twice as high.

Attached I tabulated some comparisons: the two thumbrules listed above against two manufacturer's limits. I looked at two conrad deep groove ball bearings (6215 and 6315) and two double row spherical roller bearings (22215 and 22315) For the ball bearings, it looks like the thumbrule predicts a more conservative greased speed limit in the neighborhood around 70% of the manufacturer's limit. For the spherical roller bearings, the difference is even more striking. The thumbrules above give a limit only around 40 – 50% of the manufacturer's limits.

In the case of ball bearings, from our experience, I don't believe the manufacturer's limits. They're saying 6316 can be greased up to 4,000 rpm. It may last awhile but it's not gonna last long imo. Again, we have lots of problems all the way down to 6313 at 3600rpm.

I would be interested to know exactly where those limits come from. I know there was an ISO change a few years back that affected the way greased bearing speed limits are presented. Is anyone familiar with that?

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

CompareGreaseSpeedLimitThumbruleVsCatalogue.xls (14 Kb, 31 downloads)

Would I change it? Well, you know my opinion about the long-term reliability already. Unfortunately, I have to admit that we have not done much about our high D*N greased bearing motors.... just lived with it since the plant started 20 years ago (changed a lot of bearings). I had looked for precision bearings and alternate cages but they were not available in this large size. We have been using Mobilith SHC100 grease. We believe there may be some better grease optimized for high D*N speed (does anyone have any suggestions?). Last year in one motor we put in Royal Purple Ultra 32. We thought perhaps we would see lower bearing housing temperature due to lower base oil viscosity but we didn't. The motor has operated with new grease for one year... not enough history to draw any conclusion yet. I'm not sure how much effort would be involved in converting from greased to oiled bearings. Maybe someone else can chime in on that. (I would like to learn what's involved).

As far as the diameter of the bearing bore, I am at a loss to explain it. Attached is a page from the SKF on-line catalogue. The closest thing to your part number is SKF 22215EK with bore of 75mm.

In most bearing part number systems, those last two digits times 5 give the bearing part number in mm. In other words, 6315 has 75mm bore. 6316 has 80 mm bore, etc. I don't understand what's going on with your bearing 22215 = 62mm...maybe someone else can explain it.

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

22215.ppt (163 Kb, 23 downloads)

I'm with Sam.... 0.0015" sounds awfully tight to me. Increasing this clearance can be done pretty easily. Do this -- without changing anything else -- and if the temperature goes down, you have your answer.

quote:

I agree with you on the point that that seems like a real tight bearing. I used to set bearings to .002-.003....this bearing is one of the tightest I've ever seen.

What class clearance is it? C2, C3, C4, C5? Like I said earlier, sounds like it might be a C2, and also did the final clearance measure .0015 on both of the roller/races of each bearing and not just on one side? I would check on this "C" number thing. but that is just me.

SKF sells a set of feeler gages that include metric relabeled as American to give you some odd sizes. Very useful when measuring the bearing free clearance and the drawn up clearance. One other thing, the clearance chart has dimensions for the different clearance classes, C3, C4, etc. You need to use the appropriate chart based on your bearing clearance class. Your unmounted clearance tells me it is a normal class clearance. If your drawn up clearance is .0015", that is an acceptable clearance if your unmounted clearance is .0035" or less. However, the unmounted clearance on this size bearing varies up to .0037", which would make your mounting too tight. High hoop stresses in the inner race and higher than normal temperatures would result. Combined with the speed issue, this could create a problem.

Still one more item. It is possible to meet the clearance reduction and be below the minimum allowable final clearance based on some earlier charts from SKF. What was the clearance on the second row of rollers?

quote:

Also, with regard to your question about the 22215...I personally verified that the bearing was indeed an SKF 22215 PK double spherical roller bearing and the bore was 2 7/16".

Harping on his a bit, but it really does not make sense. As Pete noted before, a 22215 bearing should have a bore of 75mm, or imperial equivalent (a shade under 3 inches). And the PK suffix; K refers to a tapered bore, but the P refers to a polyamide cage. A plastic cage in a spherical roller? At 3000RPM???

The SKF website does not list plastic cages in spherical rollers. Curiouser and curiouser.

maybe the 2 7/16 is the shaft size. Add an adaptor or withdrawal sleeve and you would get close to 75mm. Some catalogues list these taper bore brgs against the shaft size rather than brg bore size.
On the brg overheating front am I correct in saying it is the relatively lightly loaded unlocated brg that is overheating the most? Maybe you are under the minimum load range. Could this cause skidding resulting in more sliding friction rather than rolling. You could try using a self aligning ball brg in this position. This will handle light loads better. As mentioned above oil lube would be the better solution but this may get you out of the woods.
Cheers