Spindle's reliability

[JeanMicheling]

Hello,

I work for a company that make wood planers. We have been having several problems with our spindles for a couple of years now. The spindles fail too fast (sometimes they don't last more than 3 months). Here are the specs :

Motor's speed : 3600 RPM
Motor's HP : 100 to 200 HP

It's an overhung desing with a tool that weights 550 pounds and has a 16'' diameter and a 10'' width. The spindle is motorized with a coupling. Bearins are a 7020 15 degrees precision pair at each end. They have a light preload.

For the load calculation, I used (hp*63025/(rpm*Radius))*2

I used a load factor of 2 because it's a high vibration application.

The boards can go through the planer at 3000 ft/min with a 6'' gap between each board.

I would like to have some opinions on this design.

Thank you

 

[ejc]

Sounds like a very "dusty" application. Is the bearing sealed? Premature failure could be as simple as the dust entering the bearing. Especially if there is cutting fluid as well.
If the bearing is not sealed , I would look at adding a seal there to deflect the dust before it enters the bearing.

 

[JeanMicheling]

You're right! it's a dusty environnement and contamination aren't good for the bearings. The bearings aren't sealed. But the failure that we have seen doesn't come from contamination. My biggest concern is the preload (that is actually light). I think the spindle miss rigidity. I got some pictures that show a failure on some sealed bearings that we tried. It's really uncommon. They lasted only 5 weeks.

 

[EnglishMuffin]

Exactly what type of failure are you seeing ? Are the bearings overheating just prior to failure? I expect so, if they are failing that rapidly and you suspect preload. But without seeing a cross sectional drawing of the spindle, it's a little difficult to say what the source of the problem could be. Do you have two preloaded pairs each assembled back to back, with the rear pair floating axially, or do you have two pairs in tandem ? Are they spring preloaded or solid preloaded, and if the latter is the case, what determines the preload value?

 

[JeanMicheling]

we see different type of failures (if you want I can send you some pictures) but one of the most common is a little beginning of flaking at only one place on either the inner ring or the outer ring that has about 1/8'' diameter. Recently, we tried sealed bearings and when we broke it down after the vibration analysis detected an inner ring defect, we saw a beginning of flaking. Also, there was a kind of false brinelling but on all the width of the inner ring's race path that is to me really bizard (if you have an idea of what does it mean I would find that really great!). On another bearing there was a double race path on the balls which is an indication of wrong preload. It's hard to say if the bearings overheat before falures since we don't have thermocouples on them and the crash is really fast (within 2 minutes). But the failed spindles that we break down are completely destroyed by extreme overheating. The design is two preloaded pairs each assembled back to back. They are both 7020s with the drive end floating axially. They are preloaded with a locknut that we tight with the NSK recommended torque.

 

[svanels]

Try to post your problem on

http://maintenanceforums.com

 

[Pressed]

I once worked with a double headed clutch bearing tester. Was mounted with two flywheels on opposite ends of a high speed shaft. It was overheating very quickly.

First problem was not enough lubrication. Then the guys packed it with grease & oil; too much lube causes overheating too. Packed full it won't run, burns motors.

I went through the bearing application and it calculated to last 25 years at the current rate. So I called an application engineer at the bearing manufacture and he said:

Add a fill hole into the pillow block that houses the bearings and add just enough motor oil to contact the lowest ball or roller by about 1/8". That's it, never had another problem. Your setup may be a little different, but I think the basic machine is similar.

My pillow block held about 5 ounces of oil which was changed after each test. If you don't change your bearing or design you might have to add fluid capacity, a filler hole, a drain hole, an external site level, and a oil change schedule. As yours is a product you might add an oiler unit, but an application engineer should be able to provide a permanent solution for you over the phone.

You can get stick-on temperature gages from McMaster Carr for a song. If I remember right we blew about $250 for bearing set each time before the fix. And the fix was well tested as those machines ran for days on end at high speed.



_______________________________________
Feeling frisky.........

http://www.tailofthedragon.com

 

[EnglishMuffin]

Well, I suppose the next question should be - do all four bearings overheat ? Do you notice any difference between the four(?) failed bearings ? If I had to guess, I would say that the "floating" bearing pair is not floating properly, perhaps because the outer race (or cartridge) fit is such that thermal expansion is causing it to stick. And if it's at the drive end and you have a drive belt, axial friction could be quite high which could make things worse. The bearing arrangement you have there is not a good one from the axial-float point of view, and in addition, bearings assembled back to back and very close together are not very thermally forgiving. They are also very sensitive to the inner race fit. Is this machine an established design, or a new or special one ?

 

[JeanMicheling]

I don't think the problem comes from lubrication because we use a good grease that is Kluber NBU 15 and the run-up is always good. Usually, when we receive a failed spindle, there is only one side that has overheated and I can't say that there is a tendency on one side in particular. I don't think that it's a floating problem because we'd see indentations on both side bearings and we didn't. When you speak about bearings assembled back to back and close together, you mean for the same pair at one end or a pair close to another pair? For you, what would be a 'not thermally forgiving' application? Would increasing the preload be a solution for this thermical issue. This design is standard, we usually see it on grinding spindles. I have seen a lot of spindles with fretting on the shaft, maybe we should raise the shaft's O.D.

 

[EnglishMuffin]

If you have a "thermal runaway" situation, increased preload will only make matters worse. Do you have bearing pairs with light, medium or heavy preload built into them? If you already have light preload ones, there is not much you can do there other than order specially made ones. The fretting on the shaft problem that you mention is quite common, because manufacturer's spindle bearing catalogs often recommend a rather loose fit - rather looser than it ought to be. Of late, some of them have seen the light, and are recommending tighter bore fits, which is even more important at high speeds because of centrifugal expansion of the inner race. However, with tighter fits, you will have more problems, especially with pairs mounted back to back against one another with no spacer. The bore fit has a direct effect on the radial preload of the bearing, and without axial spacers there is no way to compensate for it.
If you have a floating problem, you will not usually see indentations. You will just get sudden thermal runaway, and both sets of bearings will be detroyed. It is possible that you would then see more thermal damage on the the two inside bearings of the back to back pairs, although everything gets so hot so quickly that it might be hard to detect a difference. By "back to back" I mean bearings assembled adjacent to one another with opposed contact angles.

 

[JeanMicheling]

EnglishMuffin, you seem to have a lot of experiences in spindle design. I would really appreciate to speak with you. Would you mind E-mailing me at jean_michels@hotmail.com.

 

[EnglishMuffin]

Well - maybe a little bit. I've never worked in any industry but machine tools.

 

[DDRSS10]

The false brineling mentioned would in my opinion be caused by ball chatter as a result of the flaking of the inner race. Once the race gets any sort of irregularity the balls boud back and forth between the thus propigation the markings that you are seeing. The dia of the bearing is critical as surface speed is also a concern in heat. With grease as the only lube the ability to disperse heat is limited. With the break down of the grease building from day one of installation the temp of the spindle finnaly gets to the ultimate break down point of the grease. The grease quickly departs the contact zone of the bearing, instant over heat then failure. Hope this may give some ideas.

Nick