What's new
What's new

Removing and Reusing Tapered Bearing from Lathe Spindle (Clausing 5914 w/ Video)

Grizzlybagworks

Plastic
Joined
May 8, 2016
The short of it is that I purchased a lathe with a bent spindle (.0015" out). I was able to find a replacement spindle after removing the original to confirm the location of the bend.

Any ideas on how to remove the bearing without damaging it?

I was thinking of packing the spindle with dry ice, and taking a rosebud tip from on oxy/acetylene setup to heat up the inner race. Then use a bearing splitter/puller. I'm a little worried about damaging the cage though.

I would really appreciate any input you guys have. Thanks again!

 
Maybe I'm looking at the wrong thing in your picture, but the obvious bearing I see certainly isn't a "precision" bearing, it's a simple tapered roller bearing which can be had pretty reasonably at most bearing houses. There are different grades of that bearing but my guess is that yours is a run of the mill item.

Stuart
 
Sometimes they give you 2 holes through the spindle nose to knock the bearing off with a long punch. That one looks like the right bearing splitter would grab it.
 
From your video (nicely done by the way) I'm sure a bearing splitter would get down in there. You might even find that the bearing starts to walk off as you tighten up the splitter.
 
If that bearing was on the spindle when the machine was crashed hard enough to bend the spindle it is no longer a precision bearing. Your choice, but it will still probably be better than truck axle grade bearing.

If your decision is the spindle is junk and you wish to save the bearing the best you can, I would cut off the spindle next to the bearing and then bore the stub out of the bearing. Getting an acetylene flame (or any other flame) near the bearing will end any degree of precision still remaining.
 
The Clausing manual (free download at Clausing Colchester) states for an 8500 series to pit two 4x4's on the floor then hold the spindle by the collar and drop the spindle on the wood and it's supposed to come off. Have not tried this yet so YMMV.
 
Thank you all for the feedback. I tried packing the spindle with dry ice, heating the bearing with a heat gun to ~180* and using a bearing puller. Unfortunately it didn't budge and I didn't feel comfortable applying more pressure knowing the forces were being transferred through the cage. So on to plan B.

I ended up cutting the spindle in half between the bearing cover and the spindle nut, then faced the remaining material down to ~.010". Once I was close I started boring the spindle out knowing that there was a groove that sat a few thousandths below the inner race. As soon as I hit the that groove the shoulder popped off and the bearing was exposed. I turned down some pipe to matching the OD of the bearing race and ID of the spindle, welded it to a peice of 1/4" plate and used that as a pressing plate on a 3 ton arbor press. Cooled the spindle with an inverted compressed air can and then heated the bearing with the heat gun and the spindle pressed right out.

Thanks for all the feedback guys!

Screen Shot 2018-03-20 at 8.09.08 AM.jpg

Screen Shot 2018-03-17 at 8.47.57 PM.jpg

Screen Shot 2018-03-20 at 8.09.34 AM.jpg

Screen Shot 2018-03-20 at 8.10.05 AM.jpg

Screen Shot 2018-03-20 at 8.09.59 AM.jpg
 
The bearing looks pretty good. Cage is good, all the rollers are smooth and clean. I can't inspect the inner race but the outer looks undamaged.

I'm hopeful that the bearing is okay. I was measuring .0015" on the spindle nose while the spindle was in the machine. When I took it out and threw it on some v-blocks and started measuring I was getting .0015" as well so I'm hoping all of the runout was accounted for just in the spindle. I won't know though until I have it back together. Fingers crossed.
 
I just thought I'd give an update and let everyone know how everything went. A friend and I re-installed the old bearing on the new spindle earlier this week. Today we reinstalled the spindle back into the headstock. This is at a minimum a 2 person job, as there are alot of things that need to happen in a very short time. We packed the spindle with dry ice and warmed the backgear, which was a very, very tight fit.

With everything back together I brought it home and slowly started bringing the spindle/bearings under compression with the spindle nut. Tighten, then tap the spindle, tighten and then more tapping. Eventually I got the preload about right, although I'll have to do a final check when the machine is completely back together.

I took a .0005" plunge indicator as well as a .0005" test indicator to it (I need a .0001) and anxiously turned the spindle. Almost no perceivable movement. I'm getting probably a tenth to a tenth and a half runout. It's literally 10x better than it was. The inspection sheet for this machine calls for range of 0 - .0003" for a new machine.

I won't know if it's a complete success until I have the machine totally back together but the outlook is very, very good.

 
Last edited:
Did you consider turning or grinding the old spindle taper in-situ?
Or was the spindle shaft bent between front and rear bearings so badly that it was binding part of the turn?
 
You seemed to measure spindle OD, not taper.

The OD is not normally the (a) precision surface.
Only the taper.

It is a good job, looks nice, the video is great.

Hint:
Less chatter and shorter video is more engaging.

Hint, S:
The internal spindle taper is everything.

Bearings have errors, but they repeat, and finishing the taper while running in the mounted bearings, will significantly increase accuracy.
All precision bearing catalogs from Timken, SKF, NSK, et al mention this.
 
Grinding was out of my league to do on my own and the cost would have been much higher than what I paid for the replacement spindle. To regrind the OD I would have to remove the spindle, bearing and spindle nut. Then reinstall everything without the spindle nut to get complete access to the OD taper. For me the path of least resistance was a new spindle, which I thankfully was offered at an extremely affordable price. The original spindle turned smoothly. WHen I took it out and measured on v-blocks the bend was on the nose in front of the bearing, everything else was straight. The original owner put a piece of steel in the 3 jaw, wrapped a chain around the bar and hoisted it in the air. I went to get a friend to help move the machine and came back to seeing it hanging by the 3-jaw. The guy was a retiring professional though and reassured me he'd move many, many lathes like this. I let the lathe be loaded and unloaded in this manner so it's my fault for not stopping it.

Yes, sorry, I was measuring the runout of the OD of the spindle nose, not what the actual taper is. I'm getting the same numbers on the ID with my test indicator as well.

I'll have to wait to run the machine to dial in the bearing preload but given what I'm seeing it's literally 10x improvement and within spec so I'm very, very happy.

Sorry for the long video, I realized when I rewatched it that it could have been much, much shorter.
 
Last edited:
You seemed to measure spindle OD, not taper.

The OD is not normally the (a) precision surface.
Only the taper.

Hint, S:
The internal spindle taper is everything.

I don't know if you are thinking of milling machines or what but on a L0? lathe spindle the OD taper is what matters.
Internal taper is usually morse-something and used only once in a millenium..
 
Grounding was out of my league and to regrind the OD I would have to remove the spindle, bearing and spindle nut. Then reinstall everything without the spindle nut to get complete access to the OD taper. For me the path of least resistance was a new spindle, which I thankfully was offered at an extremely affordable price. The original spindle turned smoothly. WHen I took it out and measured on v-blocks the bend was on the nose in front of the bearing, everything else was straight. The original owner put a piece of steel in the 3 jaw, wrapped a chain around the bar and hoisted it in the air.

it should have been safe to do that but how far from the chuck was the chain? usually you put the chain behind the chuck, or at the chuck, and tie the chain to the end of the bed and then find the balance point.
 
I have never had to cut a spindle like that in my 50 years of rebuilding, bent spindle or not. That is something one might do with a name like Grizzly or someone who owns Grizzly machines.

For people reading this, this not a good example of "How To" It's a good example of a bad example in my opinion.

The majority of the people on this forum are professional machinists or machine rebuilders and what you did is laughable IMHO. You should have listened to the people trying to help you and not gone off in a 1/2 arshed manner. I can't imagine not replacing spindle bearings with new quality bearings. Re-using used bearing on a bent spindle is something they may do in the Grizzly forum but not in this professional forum. You may be a hero in other forums but not in this one. You can see we don't have a Grizzly forum here because of techniques you just did. Hopefully this thread will be closed as I would hate to see someone in the future read it.
 
I have never had to cut a spindle like that in my 50 years of rebuilding, bent spindle or not. That is something one might do with a name like Grizzly or someone who owns Grizzly machines.

For people reading this, this not a good example of "How To" It's a good example of a bad example in my opinion.

The majority of the people on this forum are professional machinists or machine rebuilders and what you did is laughable IMHO. You should have listened to the people trying to help you and not gone off in a 1/2 arshed manner. I can't imagine not replacing spindle bearings with new quality bearings. Re-using used bearing on a bent spindle is something they may do in the Grizzly forum but not in this professional forum. You may be a hero in other forums but not in this one. You can see we don't have a Grizzly forum here because of techniques you just did. Hopefully this thread will be closed as I would hate to see someone in the future read it.

Hi Richard, I'm not sure I understand your comment or that you understand my situation. When I purchased the machine the spindle nose had .0015" runout. I removed the spindle to see if the bearings were damaged and to measure the runout on the spindle. I confirmed the runout existed only infront of the bearing on the nose.

I was able to purchase a brand new replacement spindle for quite cheap. The original bearing appeared to be good and while I would have liked to replace the bearings I couldn't justify the cost without at least giving the existing bearings a shot. There was no good way to remove the bearing as far as I could tell and others I consulted. I cut the spindle in half to avoid damaging the cage on the bearing and then pressed the spindle out.

With the bearing off the old spindle I installed it on the new spindle and back into the machine observing the burnishing marks. The results are above ~.0001-.00015 runout, within spec.

I looked for advice here because I was looking for the most correct approach. I could not justify the cost, nor have access, to a grinder to regrind the spindle in place. I realize it's not ideal to re-use the bearings but I couldn't justify any additional expense on this machine. The price of a pair of these bearings is more than I paid for the machine. If I reused the original bearings and still had runout I would have considered it, but with the results above why do it?

Can you please advise what you would have done differently? I'm honestly curious and I take no offense by the personally insulting nature of your comments. I very much respect your opinion and would appreciate any constructive advice offered.
 








 
Back
Top