Tree spindle bearing failure - cause?

As you might recall about 5 months ago I bought a Tree Journeyman 425 that appeared to have never been used before it started to undergo a home-brew control conversion. As is not uncommon with these projects for me it took until recently for me to identify all the unusable new or missing parts so I could transfer the Centroid control from my J325. New power supply transformer, 3 new 40 in/lbf DC servo motors, etc etc.

One of the things I did was put a set of new OEM high-speed spindle and motor pulleys on it, changing it from 3500 to 6000RPM top speed. To do that I had to removed the angular contact bearing race on the original pulley and move it to the new one. I read various comments here about working with spindle bearings and I got some recommended grease and applied it sparingly and then broke in the grease by running the spindle in both directions at ever increasing speeds. After that the area around the bearing warmed up a little but was by no means hot or uncomfortable to touch.

I would hear an intermittent squeek from the spindle pulley area (mostly between 500 and 2500 RPM, below that it was quiet and above that there was more noise from the belt and motor and fan) but I couldn't pinpoint a cause. I pulled the top support plate back off (exposing the bearing) which looked fine and I put on a new drive belt for "just in case." The belt I replaced was the OEM and wasn't worn but it was a bit stiff and had a spot with some "wow" to that was probably where it sat wrapped around the small pulley for years. I did the break in again and the squeek was still there. I found mention in some bearing documents that bearing grease can squeek in cold weather so I hoped that was it since I couldn't identify anything wrong.

A couple days ago I was running it at 6K with a 1/16" EM in aluminum and the EM broke after doing OK for several minutes. As I was changing out the CAT40 toolholder I noticed that the spindle wouldn't turn.

I loosened the tension on the belt and it was clear it was the spindle and not the motor so I pulled the support plate off again to reveal a mess.

This shows the top cap and wavy spring washer that applies some pressure to the outer race of the bearing. There are 3 6mm SHCSs that hold that on and you do need to pull the cap down with the screws as the spring has a fair bit of tension:
http://www.eurospares.com/graphics/metalwork/tree425/Tree425bearing17.jpg

This shows the top cap, washer and the upper support plate that holds the outer race of the bearing:
http://www.eurospares.com/graphics/metalwork/tree425/Tree425bearing18.jpg

Here's the pulley and inner race after the plate and the balls and retainer were removed:
http://www.eurospares.com/graphics/metalwork/tree425/Tree425bearing19.jpg

And here are the pieces of the retainer and a couple of the 8.75mm balls that probably don't clearly show the flat spots on them:
http://www.eurospares.com/graphics/metalwork/tree425/Tree425bearing16.jpg

The upper support plate has two tapered alignment dowels for location. Putting it on the bearing seemed pretty straightfoward -- I either put it on straight and it dropped right down or I had it cocked a little and it would knock a couple of balls out and I'd have to collect and clean them and start over again.

I managed to do it OK the first time when I changed the pulley, and it appears that I messed something up the second time when I removed things to change the belt.

Does the debris, like the sawed-in-half retainer, suggest what I might have done wrong the second time? I suspect that once the retainer started disintegrating at max RPM it didn't take long for things to come to a halt. But it had run for at least a good 15-20 minutes before catastrophe visited. As you would imagine I'd like to avoid having this happen again!

These are the markings on the bearing: RHP England 6Z. on the outer race, and 7010C TSUL P4 on the inner race. RHP is now owned by NSK and I found some code information on their website:
7010 is 50x80x16
C is 15 degree contact angle
T is phenolic cage (retainer)
TYN is polyamide cage (but this is TY and I didn't see that listed)
SU is single universal mounting configuration (not a matched set)
L is light preload
P4 is ISO Class 4 (ABEC7)

I didn't see anyting about the "6Z." on the outer race.

NSK 7010CTY P4 seems to be a current bearing per the NSK website though I saw mention of one in the documentation with the extra "SUL" code in the name.

What's a bearing like this likely to cost me? $100, 200 or ? I'd like to have some idea before I start calling around for a replacement as bearing prices seem a bit variable at times.

I was so pleased to finally be working on parts instead of on the tools. Sigh.

cheers,
Michael

Micheal,
I never have had any luck removing and placing back that type of bearing, or even regreasing.
6000rpm is pretty fast for that size of bearing, and would require ideal conditions, mainly fresh grease on new bearings.
All I can say is, what I would do is, rebuild the entire spindle.
It takes very little contamination for that type of bearing to fail, even using different grease then what was originally used, is considered contamination.
Greased precision bearings, are on borrowed time after several years of use anyway.

Michael Moore said:

or I had it cocked a little and it would knock a couple of balls out and I'd have to collect and clean them and start over again

Click to expand...

Are you saying there, that balls popped out of the bearing on assembly?

What was the name of that red grease?

You don’t necessarily need to chase that exact preload. Whilst light will be easier to find, that wave spring applies the load yes?

If I was after one off bearing like that, I’d go to NTN. Their universal flush ground singles are cheap. That part number would be 7010UC G/GNP4. That being a 15 degree (UC), normal preload (GN), at P4 accuracy.

Regards Phil.

Donie, NSK rates that bearing for 16000 RPM with grease so I wouldn't think that 6000 RPM should be causing it much grief. AFAIK the spindle has only got a few hours of running time on it since new. I've not found any evidence that the machine cut metal or anything else before I got it. The pulley sits above the spindle and this bearing seems to be mainly to support the pulley against the drive belt loads.


Phil, if I didn't put the outer race on straight it was easy to nudge a ball and have it fall out of the retainer. When that happened I cleaned off the ball, lightly greased it and stuck it back in and gave it another try. The balls weren't a tight fit into the retainer -- they needed a smear of grease to hold them in place. When I first took the plate off the balls had no difficulty falling out of the retainer onto the top of the pulley when I touched them. So with a basically new machine and no signs of distress in the bearing I figured that was the way things were supposed to be. When I reassembled things this most recent time I looked inside the bearing before putting on the top cap and made sure that all the balls were in place because it was so easy to dislodge them if I didn't drop the support plate/race down on them juuuust right.

Mobilgrease 28 was recommended to me by another J425 owner who replaced that bearing (the grease was recommended to them by their bearing supplier) so I bought a tube of that and cleaned out all the OEM grease when I disassembled things the first time.

http://www.mobil.com/Canada-English/Lubes/PDS/Pds_Files/iocaengrsmomobilgrease_28.pdf

As best I can tell the inner and outer races seem to be a light fit in the pocket or on the pulley boss, and that wave washer does push down on the outer race when the cap is tightened down. I didn't do anything with the outer race when I swapped the inner race over to the new pulley.

I'll look into that NTN bearing. I found a PDF at the NTN site on "main spindle bearings" and I'll read through that. I don't want to cheap out but then again I don't want to spend a lot more money than I need to because I've spent a lot more than I anticipated getting the mill up and running.

Any ideas on what I could have done that made the bearing so unhappy on the second assembly? The only thing I did was take things apart and stick any balls that fell out back in after cleaning them. I didn't do anything to the races as I was just changing the drive belt and there was no need that I could see to mess with anything else. It was turning smoothly by hand after everything was tightened down.

cheers,
Michael

OEM spindles are assembled in a clean room. Contamination is really a problem for spindle bearings. My guess as to cause is dirt.

Micheal - as you know, we have the same exact setup on our Tree and we had the spindle rebuilt before pressing it into service.

We opted to put a timing belt on it and when we first started running it - we were all surpised at how much noise the belt made - at 6000 - 7000 rpm, it is like standing in front of a firetruck siren . . . yikes!

Several folks recommended cutting a groove in the timing belt pulley to let the air out of the gap and that would quiet things down. Well that requires removing that same bearing. When we pulled the bearing off the top of the pulley shaft - we felt a slight pop as the balls went over center in the bearing race and out the balls came just like you described. We weren't comfortable putting it back together and opted to just buy a new one.

When new - the bearings stay together and will only come apart if forced - and this can cause some brinnelling of the race surface if done cocked.

I am going from memory here - but I am almost certain our bearing had a white plastic cage and the bearing was configured such that I would not come apart with out some extra pulling (nothing huge - probably less than 5 lbs) - but when it comes apart, you can definitely feel the balls rolling over the edge of the race. Under normal circumstances, the bearing stays together.

Bruce, I may have gotten a bit of something in there the second time despite doing my best to clean all the parts but I did put some effort into trying to make sure everything was as clean as I could get it. That top bearing has to come apart in order to change a spindle drive belt. Is it common practice to move an entire mill into a clean room in order to change a drive belt?

The spindle proper is below the pulley. I didn't touch any of that and AFAIK everything in the spindle is still fine as it turns quite freely.

Is it possible to get the phenolic cage in this type of bearing jammed so it can't move freely? With the missing shoulders on the races I can't picture any way that would happen.

cheers,
Michael

Hi Ken,

I don't recall any "pop" like that. Having the balls postively retained would sure make it a lot easier to deal with! I wonder if they changed anything with the retainer designs since the early 1990s when my machine was built?

I thought this kind of bearing was designed to come apart like this since it does it so easily.

That makes me wonder what sequence you used to install the new bearing. Did you put the entire bearing on the pulley, push the top plate over the outside race, and then install the entire assembly down over the drive splines into the top of the machine? Or were you able to put the bearing into the top plate and get it to slide down onto the shaft on the pulley that was already in place? Manhandling the pulley and plate assy together could get very awkward.

cheers,
Michael

Michael Moore said:

Donie, NSK rates that bearing for 16000 RPM with grease so I wouldn't think that 6000 RPM should be causing it much grief. AFAIK the spindle has only got a few hours of running time on it since new. I've not found any evidence that the machine cut metal or anything else before I got it. The pulley sits above the spindle and this bearing seems to be mainly to support the pulley against the drive belt loads.

Click to expand...

A client had angular contact bearing failures very like that. he was turning a much smaller bearing at about 20,000 rpm, which it was rated for.... in fact that bearing was rated for over 100,000 rpm.

But bearing ratings depend on load, and amount of grease, etc..... The client's bearing had an extended break-in time, which they followed to the letter. presumably that is to channel the grease and avoid 'stirring" which might heat it.

Their result was a failure with the same blackened and distorted balls as yours. At least one race looked fine.

The only thing they could find was a possibility that the bearing had got un-loaded and "pounded" a bit. there was essentially no side load. The pounding might have caused the channeled grease to get back into the ball path and cause overheating of the balls, with consequent softening and distortion.

In your case, is it possible that the bearing races were not exactly parallel as-installed? That there was a bit of "camming" action on one of them, and that as a consequence there was some impact every revolution, with possible grease getting into the ball path in a similar manner?

or possibly the grease in question didn't channel well?

I don't know the complete results of the client's investigation, so I can't suggest any more than that.

"Any ideas on what I could have done that made the bearing so unhappy on the second assembly? The only thing I did was take things apart and stick any balls that fell out back in after cleaning them. I didn't do anything to the races as I was just changing the drive belt and there was no need that I could see to mess with anything else. It was turning smoothly by hand after everything was tightened down."

In my younger days, I worked at two shops that dealt with precision bearings, the first shop was not so good at it, the second was very good.
The last shop would never remove and replace an old bearing, not ever.
Simply, you cant expect to pull one of those super precision bearings, have the balls fall out, put a puller on the inner race, put it back together, and expect it to work.
After you removed the inner race, did you inspect the shaft for scratches and remove them under 15X, and the bore of the pulley?
Did you know a finger print or a spec of dust has a greater thickness then the run out of that ABEC 7 bearing?
Not trying to give you a hard time, this is one job that requires the greatest of care.
On cleaning and regreasing precision bearings, why would anyone think it needs grease? Because it made noise is the only reason. If the bearing made noise, its already toast.

I didn't do anything to the outer bearing race in the big plate and the inner bearing race seemed to go onto the shaft on the new pulley smoothly. I wiped all the mating surfaces clean before putting things together but I suppose there might have been something causing some tilt.

The grease did seem pushed to the edges, but I don't have any experience to judge what is normal. When I took it apart the second time things looked pretty much as I recalled them looking when I did it the first time which should have been the first time since it left the factory.

I wore nitrile gloves when working with the bearing to try and help keep things clean but I can't claim I'm absolutely certain that everything was immaculately clean. I gave it my best shot because I knew it was important.

I put a new belt on my 325 that had the same setup (the 325 and 425 use the same spindle and pulleys) and that bearing never gave me any problems with the limited use it saw in the years I had it so I had no reason to think that I couldn't do things the same way this time.

If it is normal practice to discard one of these bearings every time it is disturbed to change a $50 spindle belt I can deal with that (this belt should be good for the rest of my life) but I had no indication that was the normal case. If it is, then I've learned something out of this though I'd prefered to not have had to learn it so expensively.

I figured the sad condition of the balls came about after the cage started disintegrating and contaminating the grease.

If I can put a new and pre-assembled at the factory bearing on and have it live I'll be OK with things. It does get discouraging when I have things like this happen for no obvious reason.

Thanks for taking time to comment.

cheers,
Michael

Michael-
Of course not on the question of moving the mill into a clean room. Most mills these days use a removable spindle cartridge. The cartridge is often sent out to be rebuilt - probably in a clean room - or at least under a clean hood. I am with Donie on this one. I would rebuild the whole spindle with new bearings. Then you are sure of your starting point. You should not need to add any grease. I suspect, as MG says, that the bearings will not fall apart.

Michael

I’d doubt lubrication was your problem, but that grease is right on the border line speed rating of that bearing at 6,000 rpm. Its shown here in the Barden Engineering catalog as G-33. You could specify that for special grease fills by them. It only has a Speed rating dN of 400,000

Whats a dN value I hear you ask. That’s the easiest engineering calculation you will ever make. It’s simply the mean bore diameters of your bearing multiplied by your speed to come up with a factor. In your case 50 + 80 =130, divided by 2 = 65, multiplied by 6000 rpm = a dN of 390,000. That’s too close for comfort for the speed rating of that bearing.

Kluber Isoflex NBU 15 will take that factor up to 700,000 , Kluber LDS -18, 850,000, Arcanol L-75 1,200,000, after that your into oil mist territory. I’d have been putting NBU 15 into it.

The way that bearing seemed to drop balls so easily isn’t my experience with an RHP angular contact bearing. Some thing is going on there. I’m having trouble picturing where you had the bearing mounted, when you had that problem. On the shaft, or in the housing?

Apology if this is a rude suggestion, but you did mount it with the face of the bearing pointing down? That’s the only way it can be mounted with that wave spring arrangement pressing on the back (fat) side of the outer.

If its mounted that way, when your assembling it, it should be applying force through the bearing preventing any chance of the balls being pushed over the retainer lip in the groove. Removing it as you & Ken found will tend to pull them apart.

Regards Phil.

The good shop use special lint free wipes, One thing they did was keep exposure of the bearing to the air at minimum.
An assembly plan was written and followed. Hopefully a factory print, if not a drawing, and noting where the original bearing high marks were located. Every thing was carefully tested with measuring tools set to gage blocks and master rings. Bearings such as Barden are coded for deviation of nominal size, this can be handy, for some spindle components that may be off a little.
Bearing heaters were used to aid assembly, oven for housings, however, an arbor press and needed tooling was on hand and clean, just in case a bearing didnt slide on or in.
They used a clean room, but, nowhere is dust free. So, the bearings and spindle components were kept covered.
No moving air, the door closed, the floor left wet from a mop.
The actual assembly was usually in minutes, but the preparation would take hours.
I mentioned before about removing raised areas on components from scratches, its impossible to remove an old bearing without scratching the shaft and housing, used 15X glass and a pointed arkansass stone to carefully remove the raised area.
Another point I learned was, when the bearing package is opened, the bearing is being contaminated from that point on.
Thats why I am now bald!

Bruce Griffing said:

I would rebuild the whole spindle with new bearings.

Click to expand...

G’day Bruce.

Michael has probably done himself a disservice by calling these Spindle bearings. The failed one just supports the drive pulley on top of the spindle.

The true spindle cartridge has a spined shaft protruding above the upper main spindle bearings. That just slides into the female spline of the pulley with the failed bearing in question.

Hopefully the upper labyrinth has protected the main spindle uppers. I wouldn’t be touching the real spindle bearings just yet.

Regards Phil.

In that case, I agree. For two reasons - first you clearly know this machine much better than I. Second - never mess with the spindle bearings unless you must.

Ok well im going of on a tangent here but i think i know why the bearing died. Rolling elemant bearings must do that. They must roll. Your saying it squeked. Bearings only squeak if there dry which this was not. You said it was greassed. The only other way to make a bearing squeak is for it to stop rolling. Now it could have been contamination. Your grease could have had too much dragg overcoming the friction of the rolling causing it to slide. Othere option is too little loading combined with that thick grease. Personaly i quess that that thick grease with only a wave washer of loading caussed insurficant load to make the balls roll especialy at high speed. Centrifugal force lockeing the balls to the stationary od race and sliding on the id race way. Hence the flat spots on the ball.

There are two bearings in this setup. What does the other one look like?

Double down on the notion that if you ever press a bearing like this off or on
*through* the balls, it is junk after that. The way the phenolic cage came
apart, I would investigate excess preload condition. If for example the
upper bearing inner race did not seat down fully on installation.

jim rozen said:

Double down on the notion that if you ever press a bearing like this off or on
*through* the balls, it is junk after that.

Click to expand...

+10 on that..... Clearly.

And it seems to be common for the bearings to be against a shoulder that prevents getting behind the inner ring to press it off.

Looking at the pictures again, particularly #3, it appears there is no way to remove that bearing without pulling on the outer race, unless you can pull the bearing and the pulley together.

I assume the outer ring tends to stay in the upper support, accounting for the bearing popping apart, since you cannot access it until the upper support plate is off.

Seems strange to have a setup like that which forces a bearing removal to replace a belt, particularly when that is going to ruin the bearing. Are you certain there is no other way to access it?

Another thing is that wave washer.....

If the wave washer is to do ANY GOOD, the outer race MUST SLIDE in the retainer/support. Otherwise the bearing will be loose, and may pound or cause balls to slide, with results such as you saw. The limited pressure from the wave washer must mostly be applied through the outer race to the balls. if the race hangs up, the washer is not able to do its job.

In your case, the race is still in the plate, indicating it does NOT slide well.

if the bearing race is so stuck in the support that it won't come out, then teh wave washer won't apply preload either, and the bearing may be loose.

if it did not START loose, the working of the pulley may have pushed the bearing race UP, and then it didn't slide down under the pressure of the washer. I'd not stop where you are, the BOTTOM bearing may be well toasted also, since it had little more preload, just the weight of the pulley.

It seems that the bearing should be easily pressed out of the retainer when you take off the plate, from above. if not, that would be almost proof that the bearing was no preloaded correctly.

That lack of preload is what apparently killed our client's high speed turbine bearing.

Hi Phil,

Kluber must think pretty highly of that NBU 15, as a 50g tube goes for US$34.50 at one of the web vendors I found. But if the Mobilgrease is marginal I'll make the switch. I presume I can't hope that a new bearing would come pregreased with an appropriate grease. I got some scans out of a 425 manual and it does list 3cc of NBU-15 for the spindle bearings.

I didn't remove the outer race from the big support plate. That plate can go on only in one direction as it has counterbores for the SHCS as well as two tapered alignment dowels.

http://www.eurospares.com/graphics/metalwork/tree425/Tree425_18.jpg

shows the plate, belt and pulley and may make the situation a bit clearer.

I had the pulley installed over the splined drive shaft, put the balls and retainer on the inner race which is mounted on the upper shaft of the pulley and then lowered the plate with the outer race down on them.

JST, I went down to the garage this morning and measured (with a caliper) the bottom edge of the outer race being about 6.92mm above the bottom ground surface of that support plate. With the spring washer sitting on the race the top cap is about 4.25mm above the top surface of the support plate. I tightened the cap down and then measured the distance from the bottom of the race to the edge and got 6.04mm. I came back about 10 minutes later and that had reduced to 5.75mm. It appears that the outer race is able to move a bit in the bore. It does make me wonder what is supposed to keep the race from spinning in the bore if things are that light of a fit.

I extracted the spindle motor -- removal and installation section which talks about this pulley from the manual scans into a separate PDF. They call the top pulley the spindle driver which is a term I'd not run across before:

http://www.eurospares.com/graphics/metalwork/tree425/Tree425pulley.pdf

(I'll note that I didn't do h or i when I swapped the pulleys. Should I plan on lifting the pulley so I can have a peek at that lower bearing which I have not touched?). There is a cross-section drawing in that PDF of this upper bearing assembly with the spring washer. There is also a caution about not separating the bearing.

I suppose that when I get a new bearing I might install it on the upper shaft of the pulley, install the pulley on the machine, remember to put the belt around the pulley (I could easily get focused in on the bearing and forget that) and put the clean support plate in the oven and warm it up to 150-180 degrees to give a bit more clearance to the outer race before dropping it down over the bearing. Does that sound like a workable plan? Hopefully that wouldn't be enough heat to melt all the grease out of the bearing, though I suppose I could have some wet rags handy to try and cool the plate down after it has gone over the bearing.

I appreciate the help with this. I was able to reball the X and Y ballscrews on the 325 with Phil's advice and do a control conversion with the advice of several other people. But I'm not an engineer or a trained machinery tech and I don't always have enough knowledge about some of the specialist stuff to know that I'm coming up short on what I need to know. If I knew it all I wouldn't be here asking questions.

cheers,
Michael