Collets: Is this normal?

[EOsteam]

After seeing some slight oblong in a bored hole after using the boring head on the milling machine, it was decided to check the 3/8" collets. The ER32 3/8" collet was showing about a thousandth run out when checked with a dial indicator against an upside down 3/8" edge locator. The edge locator is the only pin that I could find on short notice that should be straight. I had a second one that was checked as well and the results were the same. The regular R-8 3/8" collet showed about .00075" of run out.

Is this normal for these collets? It was pretty aggravating to see the difference in the bore of the part of almost .0015" when checked at the 90 degree position. If these collets are that off then how can you use them to mount indicators to check centers etc.. I have a Chinese "Indicol" knock off which drives me nuts but I'm going to recheck the perpendicularity of the head to the mill table tomorrow. A quick check tonight looked OK but again that Chinese "Indicol" seems to move a bit. Any ideas or thoughts will be most welcome!

Thanks,

Harper

[Harold_V]

I'm a little lost on your thoughts. The oblong condition you speak of shouldn't be related to collet run-out. Run-out of a collet shouldn't translate to error in work, not even out of roundness, as the spindle bearings determine that feature. That's assuming the collet is griping properly. If it is not, it may permit the part to wallow, but that's a different matter. Note that the amount of error determined will translate to the same amount of error in relationship with the portion used for gripping and will manifest itself in perpendicularity and/or concentricity, but the resulting turn/bore should be round.

Error in holding an indicator won't translate to error in a reading, assuming the indicator is firmly held. Regardless of how it is held, so long as it is rigid, it revolves on the spindle bearings. Doesn't really matter how you hold the indicator so long as it's rigid, in other words.

Might be a good idea to remove the collet and run an indicator on the taper of the spindle. You may have issues with your bearings.

Note that I do not use an Indicol type of indicator. I don't trust them, as they have mechanical action that offers the opportunity to introduce error. Might be a good idea to check with a conventional DTI.

If I've missed your point, please make mention so I can reassess the problem.

H

[GlennW]

Did you have the part held in a vise or clamped down to the table when you bored it, and then noticed the out of round when you removed it?

[EOsteam]

Ahhhhh! I see where you are going. The part was clamped in a vice when bored. It was bored a few months ago so I can’t remember if the hole was checked in all dimensions at that time when it was still in the vice.

[EOsteam]

OK, just in from the shop. Here is what has been learned so far. The head was dialed and found to be on the money as well as the machine vice. The taper of the spindle was dialed w/ a DTI and had an error of .0003" when running at 60 RPM.

The problem bore in the journal box is 1.8504" and is on the money in the first direction but at 90 degrees is .002" more. There is also a difference from top of the bore to the bottom of .0005" greater. The depth of the bore is 1.3"

The journal box was replaced in the vice where it had been machined and torqued down. The measurements stayed consistent with the other measurements out of the vice.

Is the .0003" error in the spindle causing this issue? If not, what should I check next?

Harper

[John Hasler]

As Harold explained, runout can't cause eccentricity. What kind of mill? I think that spindle wobble could do this, as could a wiggly quill. Seems unlikely, though.

First, bore another similar hole in similar material (what is the material?) and see if you can reproduce the problem. Then start checking the boring head.

[GlennW]

The material may be another clue here.

I occasionally machine Ampco 18 (aluminum bronze) and it can really move around when machined.

I rough machine and then let the parts sit on the counter over night before finish machining to final size.

I believe that Harold may have once mentioned something about roughing and finishing...

[EOsteam]

The material is cast brass and Harold's admonitions from an earlier post were taken to heart. All of the surfaces were rough finished first before any final passes were taken. The bore was the last operation.

More time was spent with the dial bore gauge in the bore and I must have missed this the first time but only the top portion of the bore is out of whack. As the bore progressed down to the inner shoulder it became more concentric. There is only a .0002" difference in the 90 degree measurements at the bottom of the bore.

Another interesting thing. I couldn't detect any spindle wobble when it was spun by hand. However, when under power the .0003" wobble was noted on the inner spindle taper. The mill is a Birmingham (Bridgeport clone) which I think has a Taiwanese head with a mainland Chinese body.

[Harold_V]

Ahhhhh! I see where you are going. The part was clamped in a vice when bored. It was bored a few months ago so I can’t remember if the hole was checked in all dimensions at that time when it was still in the vice.

While there's no way for this to be determined, it's entirely possible that time has changed the bore. Internal stresses do strange things to such items. In the "good old days", when time was worth less, castings were often rough machined, then allowed to season for a considerable amount of time before finish machining. They did that for this very reason.

One thing you might want to consider doing is to try to move the spindle (not while it's rotating) while there's an indicator in contact with some portion. I'm alluding to possible preload failure. While the .0003" you found isn't horrible, it isn't great, either. It may be because of not enough preload. Or not!

It's important, when boring with a boring head, that the tool doesn't drag. If it makes contact anywhere but the cutting edge, it will often yield a strange bore, including an out of round condition (due to the variable density of the cast material). You can general tell if that's happening by looking closely at the boring bar, for signs of a polished or abraded surface where there shouldn't be one. Because of the radius of the bore, there's often not enough clearance, so the tool drags on the heel.

H