Lets Talk Runnout

[kennyv]

Hey guys id like to know what you guys get AND deem acceptable for drill chuck runnout ?

The mill I have is a PM932 . The two drill chucks that I was dialing in were both a Jacobs super duty ball bearing and an Albriech (sp) key less chuck . I did try many other chucks and collets to see if my observations repeated and or got better worse.

First id like to note that I did many runout checks and indication fix and checked everything soup to nutz. 1st inside diam than a R8 collet it self with drill rod if I can rem .001-.0025 is what the mill produced so thats what we have to start with.
this is still bothering me and I didn't post bc been tied up with other things .. did some runout checks a few weeks ago wanted to hear from you guys . Spent some time observing why .0025-.004 TIR was what I was getting and upon mounting and re mounting chuck their was so much variance ?



After several adjustments re cleaning fixin and even tapping drill chucks I was able to get TIR within . .002 ( this check was approx 1-3” down in z from surface of jaws). .. Now id say however if ya took out chuck and re installed back it went to .0025-.004 TIR AGAIN to me .0025 may have to be acceptable for a drill chuck but not any greater what is your thoughts on this ?? still don't make sense that's like .002 too much

JMHO I never though of checking my runnout on a BP every time I used good quality stuff it never dawed on me drill wasn't running true unless it was obvious

 

[Dan_S]

you got .001-.0025 with an R8 collet and drill rod? did you have the quill all the way up and locked? how far away from the collet face were you taking this measurment?

 

[T Bredehoft]

I put a piece of 1/2" drill rod in an R8 collet in my PM25 and got less than .0002 (that's tenths) run out. TIR, needle swing,.. The indicator (Federal T! .0001) was 1/4" from the collet, spindle all the way up and locked. The inside of the spindle reads well under .0001.

 

[RJSakowski]

Measuring the TIR 1/4" from the collet face gives you a fairly good indication of radial runout. However this is only part of the problem. There is also an angular component to the runout which can be analyzed by measuring the runout at some distance from the collet face. If you have .0002" at the collet but the runout increases to say .003" at 2" from the collet, you have an angular component. That component gets worse as you move further from the collet. The spindle socket may be less than .0001" radial runout but with an angular component, it could be unacceptable at a typical working distance. Since the drill chuck jaws extend quite far from the spindle, this may be what you are observing.

It can be a problem to measure the angular component of the spindle runout. If you use a collet with test cylinder, you are measuring the combined runout of the spindle, the collet, and the cylinder and parsing it out to identify the contributions of each can be a PITA. If you had a test indicator with a long enough probe to engage the cylindrical surface at the back end of the socket, you could get a better idea. It is also possible that the clearance at that surface is enough to permit an R8 tool to take a slightly seat each time resulting in erratic runout measurements.

I was concerned with runout on my Tormach and set about trying to do just what you are doing. I used a 3/4" wrist pin as a test cylinder. It checked out fairly well for straightness, roundness, and cylindricity, as measured with a set of vee blocks. I bought a Crawford collet with specified concentricity of .0004". Still, trying to dissect the measurements was proving difficult.

The solution that I came up with (yet to be executed) was to make an R8 test bar. i.e., turn an R8 tool with and integral cylinder extending about 4" out from the R8 tool face. This tool could be independently verified for concentricity. Then, with a known good test tool, I could properly evaluate the spindle socket. Once I knew where I stood with the socket, I could move on to the collets and, in my case, the TTS tool holder system which adds an ER20 collet chuck and the ER20 collets to the mix.

Bob

 

[T Bredehoft]

Thank you, RJ, I redid my setup. Used a 4" piece of 1/2" drill rod, at the collet the needle did not move. no jiggle, anything. Two inches below, it wandered around .0008. That's .0004 per inch. I made sure there was enough drill rod in the collet to extend past the clamping length.
In order to duplicate the original set up, I installed the non-Jacobs keyed chuck that Matt supplied with the mill (R8 based). It ran out (at the chuck) .0019, (two inches down) .0049. Noting to write home about. I'll await your R8 test bar results....

 

[Dan_S]

Tom,

pick you self up a chunk of linear shaft to test with, it's hardened, ground, very strait, and very concentric.
http://www.amazon.com/Thomson-QS-12...UTF8&qid=1448477105&sr=8-2&keywords=1/2+shaft

0.0008" 2" from the face is about what is to be expected for run of the mill collets in a good spindle. It might go down a tenth or 2 if you use linear shafting for the test.

 

[Bob Korves]

Maximum spindle runout should not exceed .0005", indicating the internal taper. Get that measured accurately first.

Drill rod is not sufficiently accurate to test collet runout.

Make sure that the spindle pin that locates the slot in the R8 collet is not too long and pushing the arbor/collet sideways.

Indicate the spindle both vertically and horizontally to see if the bearing preload is loose, pushing the spindle up and down and sideways.

My 1967 Millrite mill has .0002" runout at the spindle, .0005" at the end of the old but rebuilt Albrecht chuck, and .0008" when measured 4" from the chuck with a good 1/2" test rod chucked. With my 1980 Taiwan drill press and a Chinese South Bend chuck you can multiply all those numbers by five. Both sets of results make me very happy.

 

[mksj]

An R8 collet has a TIR of around o.0005", and then you have the TIR of the rod you are using if you are measuring the machine TIR as such. You can get Tight-Tolerance Rods—Precision Ground at McMaster Carr. I use sme 3/4" and 1" to verify diameter and alignment. As far as chucks go as mentioned, Albrecht chuck itself is rated around 0.0016", the Jacobs probably worse. I have similar chucks and I have about a 0.002" runout, which is not uncommon. My default mill Chinese keyed chuck is much worse. I have an integral R8 shank chuck which has a repeatability TIR of better than 0.001". I also have a keyless integrated MT-3 shank for my lathe which also seems to hold this tolerance.

 

[kennyv]

Measuring the TIR 1/4" from the collet face gives you a fairly good indication of radial runout. However this is only part of the problem. There is also an angular component to the runout which can be analyzed by measuring the runout at some distance from the collet face. If you have .0002" at the collet but the runout increases to say .003" at 2" from the collet, you have an angular component. That component gets worse as you move further from the collet. The spindle socket may be less than .0001" radial runout but with an angular component, it could be unacceptable at a typical working distance. Since the drill chuck jaws extend quite far from the spindle, this may be what you are observing.

yea that's what I was thinking all along now finding the issues is next step
thanks

Bob

Maximum spindle runout should not exceed .0005", indicating the internal taper. Get that measured accurately first.



Make sure that the spindle pin that locates the slot in the R8 collet is not too long and pushing the arbor/collet sideways.
.

now this has me thinking cuz I had issues with collets fitting my new PM932 from get go ..I had to spend allot of time cleaning /polishing spindle R8 taper to get anything to fit ( tight jammed ect...... .. and than what I did do was remove the spindle pin altogether ( was advised by Matty at PM ) he said didn't matter ..and it sounded logical bc its a taper . hmmmmm anyone?


now what id like to know and get someone to concur is their is a good possibility in what I have observed as far as up to .0035 runout with a good quality chuck lest say 1-2 " from spindle can be from removing taper pin ?

or that should't mean squat Bc i do get good readings with lets say 2-4" down from collet face when using collet instead of drill chuck ?

 

[Dan_S]

now this has me thinking cuz I had issues with collets fitting my new PM932 from get go ..I had to spend allot of time cleaning /polishing spindle R8 taper to get anything to fit ( tight jammed ect...... .. and than what I did do was remove the spindle pin altogether ( was advised by Matty at PM ) he said didn't matter ..and it sounded logical bc its a taper . hmmmmm anyone?

What specifically did you do, to clean/polish the spindle taper? if you pulled the pin, are you sure you got the spindle bearings and everything seated properly when you put the spindle back together?