What's new
What's new

Spindle wobble on Hurco KMB-1S

_big_

Plastic
Joined
Jun 13, 2023
40+ year old Hurco/Kondia KMB-1S exhibits rather noticeable spindle wobble.

Assumed it was bearings. If I indicate on the inside of the taper and grab the Qwik Switch lock ring, I might be able to get 0.001 deflection. Not great, but even worse...

If I simply indicate the taper and spin the shaft with no load, I'll see 0.001 or so wobble. If I insert a tool with a six inch stick out, it gets much worse. Like 0.004+ wobble.

I don't know the official method for evaluating spindle wobble, but from the above I've concluded the bearings are marginal but even worse, the spindle itself is bent or otherwise crocked in some way.

A new spindle (if available) is probably not justified on such an old beast. I've heard that tapers can be reground. Perhaps this problem can be fixed by re-grinding the taper?

Any and all advice appreciated.
 
Try bluing up a tool holder that has seen minimal use with a LIGHT coat of prussian blue and attempt to see what the contact pattern on the taper looks like. Problem with this is the tangs on QS toolholders engage two notches in the spindle itself that will make wringing the taper to get a good pattern pretty difficult. Bearings are about the same size as a Bridgeport, so replacements should not be too expensive. You might contact Midwest CNC in Indianapolis- they might have an old spindle in their parts pile.
 
Thanks for the tips Dan.

I made several measurements. The first one pictured below (practical machinist won't let me upload the movie:angry:) shows the indicator riding _on_ the taper. This gives me a runout of about 0.001.

I also mounted a long tool in a tool holder, and carefully tightened the lock collar down on it. I made several measurements along the long tool. As I got further away from the taper, the indicated runout qualitatively increased as might be expected. At about the six inches from the taper edge, runout was close to 0.005 😦.

I haven't actually removed the shaft from the machine yet. I have a friend with a lathe and hopefully we can chuck the spindle in the lathe and get some idea of what kink looks like.
 

Attachments

  • Capture.JPG
    Capture.JPG
    42.6 KB · Views: 9
Have completely disassembled the spindle now. Looking to get access to a lathe to measure runout.

In the meantime, I'll probably need new spindle bearings. The dual radial bearings are New Departure 20207 and the thrust bearing is Fafnir 206PP.

New Departure is no longer around. Fafnir seems to now be part of Timken.

I think both of these are relatively standard size and bearing type. I can probably get a cross reference. But I think there is more to buying spindle bearings than just size and type. In both cases, I think these are high precision bearings.

I don't know anything about bearing grades. Perhaps somebody could point me to a bearing grade tutorial.

Thanks.
 
Just measure the runout on v blocks. Putting it in a lathe would muddy the results as you will inevitably see the lathe bearings runout as well. Put the v blocks where the bearings sit, and measure runout all critical surfaces-bearing journals, shoulders, and taper. Also measure journals for size, roundness, taper and concentricity. Good idea to check the housing bores as well.

Dont worry about getting the same mfg of bearing, you just want to get the same or better precision, and comparable other specs - contact angle, ball material, preload, etc.

If you post a couple pics of the bearings I can help, there's a lot to know when it comes to bearings.
 
Was figuring I could null out a lathe's imperfections.

I'll give vblock technique a try. One problem is the shaft is long and will be cantilevered off the end of the vblock. For an accurate measurement, it is essential that the shaft is consistently in full contact with the vblock. We'll see :).

Attached find pics of the radial bearing (N D 20207) and the thrust bearing (FAFNIR 206PP).

>>there's a lot to know when it comes to bearings.

That's why I'm looking for information. I suspect the size and type are not hard to match. My biggest concern is the accuracy of the bearings.
 

Attachments

  • IMG_7039.jpg
    IMG_7039.jpg
    1.9 MB · Views: 8
If it was impossible to hold steady on v blocks we would put it on one of our cylindrical grinders as they have much better runout than a lathe. It can be done on a lathe, but it will be hard to get granular enough to really learn much.

Those are going to be pretty standard bearings, the rear 2206 is a radial (cheapish) bearing, the other pair is a set of angular contact bearings and thats where precision matters. It looks like ********* makes a direct replacement that is almost certainly better then the new departure crap. Id call your local bearing distributor and they can cross it directly for you. This is an old design that isnt used in spindles anymore so Im not as familiar with these bearings.
 
Finally got around to trying the vblock technique. See photo attached.

My vblocks have these clamps to hold material in place. The screws are conveniently brass tipped. In order to validate my approach I put the indicator on the barrel turning in the vblock. I have only a 0.0005 indicator. When I twist the shaft, I consistently see variations that are maybe around 0.0001. Probably less than my measurement capability. I used a little bit of lubricant, and cleanliness is of course critical.

If I move the indicator to the edge of the taper as shown in the photo, I get about 0.001 runout. Consistently.

If I move the indicator to the shoulder part, I get results similar to what I see when I measure the surface turning in the vblock.

If I put a tool in a tool holder and measure the smooth part of the tool near the tool holder, I see like 0.002 runout.

If I use a long tool (about 6 inches) then I'll see about 0.006 inches runout.

This leads me to the conclusion that the taper itself is somehow tilted with respect the center axis of the shaft. I don't think the shaft itself is bent.

For sure the machine is old and there have been many insert/remove cycles on the taper. Its hard for me to see how it would wear this way? Perhaps it has more to do with forces experienced during cutting operations?

Certainly I'm nervous and "trying" some expensive repair only to find out I didn't fix the problem.

Midwest CNC sells guaranteed shafts with preloaded bearings for about $2K. There is a shaft on Ebay with the exact part number for $1800.

The debate of course... is the 40 year old beast worth that kind of investment? I'm sure the ways and ballscrews are worn too.

And finally, if somebody might have a donor KMB rusting away in their scrap heap, maybe I can find a better shaft for less $$.
 

Attachments

  • IMG_7092.jpg
    IMG_7092.jpg
    1.2 MB · Views: 10
Dont trust a tool holder for your readings, the taper itself is here you want to look.

I would have had one of the V blocks on the rear journal, or at minimum checked the runout of the rear journal while setup like this on the front. Odds are that is running out being setup this way, and therefore showing some runout in the front, but certainly you have enough runout to know you have a problem. If it is just the taper, it can be ground, if its bent elsewhere it could be a different story.

Honestly a new shaft for 2k with bearings seems reasonable, but I would think that one could be saved once you really know whats going on.
 
Thanks for the response mottrhed.

Don't really understand your comments. Its maybe not clear from the photo, but there is a single vblock about 4 inches wide with two clamps. It pretty much fully contacts the primary bearing journal which is also about 4 inches wide. There is nothing that I would call a "rear journal".

With this rig, the shaft is reasonably secure. If you grab the left end it takes a bit of carelessness to cause the taper end to wiggle.

I'm at full stop with respect to "really know what's going on". Don't really know what else to do.

I know that tapers can be reground. Although, most of the places seem to specialize in in-situ grinding. I'd prefer to ship this shaft off somewhere. The grinder would have to have a way to hold the shaft, but any lathe like chuck could do it. And hopefully, a regrind at a shop would be cheaper than in-situ.
 
There is a bearing that sits right in front of those rear threads right? That would be the rear journal where the radial bearing sits.

Grinding a bare shaft is the right way to do it, not on bearings in a machine.

Hopefully it can just get a kiss grind and doesn’t need to be built back up with plating, but I’d have to see it to know. Either way, assuming the shaft is relatively straight and just has a worn/damaged taper, it will be far cheaper to fix than replace.

If you want to send it in I can measure it all up, and quote what is the best course of action. Best case, taper grind, worst case it’s bent beyond cost effective repair, and there’s lots of in between scenarios.
 
Whats a KMB-1S? Is it a KMB1 column/table with the vari drive head or a KMB1 with the variable frequency timing belt drive?
Its actually a KMB-1m, although some of the documentation I have for it says KMB-1S. See attached pics.

I've upgraded the control with RaspberryPi/LinuxCNC/MESA so that's why the control panel looks different in the frontal pic.
 

Attachments

  • IMG_5954.jpg
    IMG_5954.jpg
    108.4 KB · Views: 11
  • IMG_5427.jpg
    IMG_5427.jpg
    132.8 KB · Views: 10
There is a bearing that sits right in front of those rear threads right? That would be the rear journal where the radial bearing sits.

Grinding a bare shaft is the right way to do it, not on bearings in a machine.

Hopefully it can just get a kiss grind and doesn’t need to be built back up with plating, but I’d have to see it to know. Either way, assuming the shaft is relatively straight and just has a worn/damaged taper, it will be far cheaper to fix than replace.

If you want to send it in I can measure it all up, and quote what is the best course of action. Best case, taper grind, worst case it’s bent beyond cost effective repair, and there’s lots of in between scenarios.
Based on the measurements I've taken, its my opinion the shaft is not bent, but the taper is somehow off center. How the taper could become off center is beyond me??

>> If you want to send it in I can measure it all up, and quote what is the best course of action.

Wow! Those are exactly the words I'm looking for. PM me and lets make arrangements to have this shaft evaluated and hopefully repaired for a moderate price :)
 








 
Back
Top