What's new
What's new

Can anyone tell me why did I take apart this Hardinge spindle?

Frigzy

Aluminum
Joined
Apr 26, 2018
Well, it's a Feeler FTL-618EM, but I like so much to call it a Hardinge :-)

The spindle would make funny noises at RPMs above 2000 and TIR was as much as 0.0005" on both - inner and outer tapers. It produced nasty finishes on everything except 4140.

As if my life was not challenged enough, I started to slowly disassemble the headstock with intention to check out grease in the bearings.

In a creative impulse of violence, I didn't notice myself how the spindle was taken apart down to a smallest piece.

The rear bearing feels OK, but the front one gives a lot of roughness, like if balls were square-shaped. It would also spin 10-20 revolutions if I push it by hand.

What is the diagnosis? Re-lubrication? Replacement?

On the last photo is a tool that I made to undo the spindle nut.

Thanks!
IMG_9229.jpgIMG_0808.jpgIMG_3361.jpgIMG_8995.jpg
 
Part of me wants to know what happened to your childhood pets, but part of me doesn't...

OK, setting aside vivisection fears, I'm wondering why you didn't ask about the bad finish and geometric errors before you started taking things apart. If the head was still together you could have done some diagnostics with test indicators and the like.

BTW, your comment on the front bearing being rough but spinning freely is possibly indicative of worn balls and little lube (maybe). Any numbers or other identifiers on the bearing?

But where you are right now, it sounds like the front bearing needs replacing. Unless you're lucky and you can clean and relube the bearing and the races look good (no pitting) and the balls are OK, you'll have to figure out the bearing size and accuracy class, then find a replacement. I'd get some help removing and pressing the new one on, haphazard attempts with this will likely damage the new bearing.

Do you have notes on what order you removed stuff, including orientations? Is there any sort of tension adjustment on the bearings, such that if there's a specific preload that needs to be met you can replicate it?
 
Well, it's a Feeler FTL-618EM, but I like so much to call it a Hardinge :-)

The spindle would make funny noises at RPMs above 2000 and TIR was as much as 0.0005" on both - inner and outer tapers. It produced nasty finishes on everything except 4140.

As if my life was not challenged enough, I started to slowly disassemble the headstock with intention to check out grease in the bearings.

In a creative impulse of violence, I didn't notice myself how the spindle was taken apart down to a smallest piece.

The rear bearing feels OK, but the front one gives a lot of roughness, like if balls were square-shaped. It would also spin 10-20 revolutions if I push it by hand.

What is the diagnosis? Re-lubrication? Replacement?

On the last photo is a tool that I made to undo the spindle nut.

Thanks!
View attachment 247011View attachment 247012View attachment 247013View attachment 247014

I think a more cluttered and disorganized work area is just what you need to work on a precision spindle.
 
Don't panic!

carefully inspect the parts for assembly marks, witness marks and any thing that will help you re orient the main rotating parts.

Inspect the bearings with a magnifying glass , it has always surprised me how even the smallest blemishes can create a terrible noise.

if you're fitting new bearings you are pretty much starting from scratch anyway . do lots of measuring and take notes.

I believe feeler's* are a near clone of a hardinge so the experience of the experts here will be invaluable

(*I have a pdf of the manual somewhere if you need one)
 
Don't panic!

carefully inspect the parts for assembly marks, witness marks and any thing that will help you re orient the main rotating parts.

Inspect the bearings with a magnifying glass , it has always surprised me how even the smallest blemishes can create a terrible noise.
)

All good advice, except you shouldn’t be surprised, a ball bearing only contacts a small area at a time. In fact, the trick is to use a good ten power loupe, not a “magnifying glass”. Then you will really get a look at what you are dealing with at the scale you need to.
Also, clean parts well first, some defects in the race in particular can have grease and steel particles packed into them and seem less serious than they really are, especially on machines that got lightly used and never oiled. Good luck!
 
I'm wondering why you didn't ask about the bad finish and geometric errors before you started taking things apart. If the head was still together you could have done some diagnostics with test indicators and the like.
It all started when I decided to replace the belts. It looked easier to me to detach the entire gearbox to remove a spindle belt and then everything just snowballed.

Any numbers or other identifiers on the bearing?
It says B7014 C.P4.UL A*B ********* GERMANY which is:
- B Standard OR relieved, IR symmetrical
- C 15 degrees contact angle
- P4 ISO P4 (ABEC 7)
- U Single bearing, Universal
- L Light preload
( http://www.bscindia.com/catalogue/FAGPrecBrgsCat.pdf )

Do you have notes on what order you removed stuff, including orientations? Is there any sort of tension adjustment on the bearings, such that if there's a specific preload that needs to be met you can replicate it?
Yes - everything was already marked with a reference to a key slot on the rear side of the spindle. The spindle nut was tight as hell - I needed to employ a 5 feet lever to take it off. It seems to preload is adjusted by length of the spacers in this design.

But where you are right now, it sounds like the front bearing needs replacing. Unless you're lucky and you can clean and relube the bearing and the races look good (no pitting) and the balls are OK, you'll have to figure out the bearing size and accuracy class, then find a replacement. I'd get some help removing and pressing the new one on, haphazard attempts with this will likely damage the new bearing.
I'm starting inclining to buy new bearings and replace them.
Here are Hardinge bearings: ShopHardinge - HC 265GT
The plan is to slowly replace everything with genuine Hardinge parts and sooner or later I will end up with a real Hardinge. I already got belts and way wipers so I'm half way there :-)

Part of me wants to know what happened to your childhood pets, but part of me doesn't...
Fortunately for all the parties involved, my parents have never allowed me to have pets..
 
It's sealed bearing,If it's rough, it's usually indicative of a bearing that has gone south along with accompanying issues. In the world of bearings it doesn't take much. (Had you kept your work area clean this would not have happened :popcorn:) this is a somewhat tongue and check assessment and glossing over quite a few fine points. It does appear that you are on the right track. Good luck, Hodge
 
All good advice, except you shouldn’t be surprised, a ball bearing only contacts a small area at a time. In fact, the trick is to use a good ten power loupe, not a “magnifying glass”. Then you will really get a look at what you are dealing with at the scale you need to.
Also, clean parts well first, some defects in the race in particular can have grease and steel particles packed into them and seem less serious than they really are, especially on machines that got lightly used and never oiled. Good luck!
Thanks!

It's sealed bearing,If it's rough, it's usually indicative of a bearing that has gone south along with accompanying issues. In the world of bearings it doesn't take much. (Had you kept your work area clean this would not have happened :popcorn:) this is a somewhat tongue and check assessment and glossing over quite a few fine points. It does appear that you are on the right track. Good luck, Hodge
Thanks! The bearings are not sealed. The front bearing is protected by a labyrinth seal and the rear bearing is covered with a large diameter pulley. This surprises me a lot - people say that the rear bearing should go south first because of the design. Rubber particles from the belt go straight into the rear bearing. My situation is opposite - rear bearing is decent (even though it looks ugly) but the front bearing is what I consider "shot".

I got this machine a month ago and spindle bearings is far not the only problem I have with it. Previous owner was a nice guy. Unlike me, he was a machinist and he was making money on the lathe in 24/7 mode. At least, I'm kept entertained now :-)
 
I found incredibly useful information about designing/handling/installing precision bearings here: https://feyc.eu/download/aplicacion...eToolSpindleBearingSelectionMountingGuide.pdf
Recommended locknut tightening torque for our bearings is 166 Nm which is 122 ft-lb.

It's not 166 Nm on the bearing, that's only the torque on a properly shimmed bearing stack. Actual axial load on the bearing will be much less, unless you're only using it at low speeds. This is one area you have to be very careful with, overtightening an improperly shimmed single ended angular contact bearing is a great way to instantly ruin it (@ $390/pop).

BTW, the bearing you listed is a decent price - Motion Industries version was more than twice as much.

Go ahead and order the bearing, once you've absolutely confirmed it'll work in your H-Copy. It must be installed with the correct orientation so it will accept thrust loads, and it must go on square to the spindle OD, with all pressing loads on the bearing ID race. Everything MUST be clean and burr/dirt free, with only a light coating of oil on the spindle diameter and the ID of the bearing race.

Heat the bearing in a toaster oven (buy a cheap one for this job, confirm temps are close to indicated), with it wrapped in clean Al foil to protect it. 175F or so should do, for at least 30 minutes to soak. Wear clean, thick gloves to protect your hands, the bearing should be a slip fit. Ensure it's fully seated and won't move off during installation.

You can test the process with the old bearing after you've removed it (and then remove it again). Once you can do it well with the old do the new.

Do not tighten the bearing retaining nut any more than finger tight before confirming proper shim stack thickness! Ask more questions when the time comes. Do you have a good .0001 test indicator and stand?
 
When it gets back together the nose (inner/outer) is supposed to be +-.000025. If you don't get that then you might have made a boo boo.
You won't know for sure if you made it worse.

BTW, Hardinge recommended to me that I should make little reference marks where the bearings fit. On re-assembly the reference marks make sure
that the bearing are put back in the same position.
 
It's not 166 Nm on the bearing, that's only the torque on a properly shimmed bearing stack. Actual axial load on the bearing will be much less, unless you're only using it at low speeds.
Yes - I got it. 166 Nm is a locknut torque. When I was taking it off, it felt exactly like 166 Nm. The locknut pushes against spacers and length of the spacers and bearings design are setting up the actual preload. If I get it right, in order to get correct preload for these particular bearings, spacers need to be ground to exact same length.

Heat the bearing in a toaster oven (buy a cheap one for this job, confirm temps are close to indicated), with it wrapped in clean Al foil to protect it. 175F or so should do, for at least 30 minutes to soak. Wear clean, thick gloves to protect your hands, the bearing should be a slip fit. Ensure it's fully seated and won't move off during installation.
I have a 3d printer with heated bed. I'm going to use that to heat up the bearings. It may push up to 130C with 2-3C of accuracy.
In the document they recommend to heat up an inner race only, without heating balls and an outer race. Why is it so? Is it ok if I heat up the entire bearing?

Do not tighten the bearing retaining nut any more than finger tight before confirming proper shim stack thickness! Ask more questions when the time comes. Do you have a good .0001 test indicator and stand?
I have a couple of .0001 test indicators and a surface plate. Everything was recently calibrated. How do I measure radial runout of a spindle shaft with no bearings on it? Do I need precision V-blocks or regular gage blocks will suffice?

Thank you for helping out with this!
 
When it gets back together the nose (inner/outer) is supposed to be +-.000025. If you don't get that then you might have made a boo boo.
You won't know for sure if you made it worse.
Challenge accepted! :-)
I was going to clean up inner and outer tapers of the spindle nose with a good CBN insert when everything is back together. There is a success story somewhere on this forum with turning a Hardinge spindle with a CBN insert.
BTW, Hardinge recommended to me that I should make little reference marks where the bearings fit. On re-assembly the reference marks make sure
that the bearing are put back in the same position.
I have those marks. Just in case, I will double check all the runouts and will calculate bearing positions following the procedure in the NSK document. Will see if that matches the marks.
 
Yes - I got it. 166 Nm is a locknut torque. When I was taking it off, it felt exactly like 166 Nm. The locknut pushes against spacers and length of the spacers and bearings design are setting up the actual preload. If I get it right, in order to get correct preload for these particular bearings, spacers need to be ground to exact same length.

Well, the "correct" length, not same, which could vary from using a different bearing. Correct will be the stack length that (for lack of factory values) gives the spindle zero axial play while allowing minimal preload on the two bearings. If you're lucky, Feeler has actual procedures for loading the bearings, you should contact them.

I have a 3d printer with heated bed. I'm going to use that to heat up the bearings. It may push up to 130C with 2-3C of accuracy.
In the document they recommend to heat up an inner race only without heating balls and an outer race. Why is it so? Is it ok if I heat up the entire bearing?

It's OK to heat the entire bearing, but 266F (~130C) is too much. Lower is always better, and most precision bearings are not meant for operation over ~200/250F.

I have a couple of .0001 test indicators and a surface plate. Everything was recently calibrated. How do I measure radial runout of a spindle shaft? Do I need precision V-blocks or regular gage blocks will suffice?

Buy a good magnetic stand for the indicator, it'll make life easier when setting up the preload. The V blocks are good for checking if the spindle is bent, but that doesn't seem to be the issue here.

Get back in touch when you have everything and are ready for assembly. Good luck!
 
Challenge accepted! :-)
I was going to clean up inner and outer tapers of the spindle nose with a good CBN insert when everything is back together. There is a success story somewhere on this forum with turning a Hardinge spindle with a CBN insert.

Unless you really have bad runnout, don't try cutting the OEM tapers. It's a good way to really bork the machine, as the linear travel accuracy and stiffness of the slides you have aren't as good as the ones used when the machine was made.

If you do have significant runnout, it's best to grind the spindle in place. Might cost some bucks, but better than ruining a machine.
 
Can you document what you had to do to get the spindle and bearings out and replace? I have a Victor 618EM that probably was made by Feeler. The spindle makes a whine that sounds like a bearing.

Tom
 
Last edited:








 
Back
Top