5C taper problem again

[rexcsmith]

I posted on this a few months ago. I've tried all the suggestions and have not solved the problem. An update.

I have a Monarch 10EE with the drawbar/nose-cone 5C collet system. I'm using a 3/4" Lyndex collet.
When I reduce the diameter of a 3/4" rod, it tapers down toward the collet. Over 3", the taper is about 4 thousanths; narrower near the collet. This does not happen with my Bison 3 jaw chuck.

I have borrowed a Hardinge 3/4" collet from a friend, but same problem.
The results are exactly the same whether or not using a live center. I've even drilled for the center with no work-piece stickout from the collet. Then when I pull the workpiece out, and lock the center in, there's no indication of misalignment. But, when I do the cut, I get the same tapered result.
I've tried rotating and jiggling the workpiece while closing the collet.
Bottom line, no matter what I try, the result is always exactly the same. 4 thou taper over 3 inches.

I have dial indicated everything I can think of and it's all spot on.

I'm fresh out of ideas. Any more clues out there? Right now, my beloved 5C system is unusable.

thanks, rex

[John Hasler]

What do you see when you indicate a precision 3/4" rod in the collet with and without the live center? What happens when you turn between centers? What do you see when you indicate the inside of the spindle taper? Have you checked for burrs, especialy in the back bearing?

[rexcsmith]

I'll start by saying I'm not an experienced machinist; merely a hobbyist that's about a year into it.
I'm using 3/4" drill rod, although I don't know if I can count on that to be precision. If not, where can I get some precision 3/4" rod?
I have indicated it with and without the live center. Same result. 0 runout at the collet. 2 thou runout at 3"
I have indicated the inside of the taper. Spot on.
Not sure what you mean by back bearing; is this part of the 5C bullnose?

[John Hasler]

If you're getting zero runout at the collet and inside the taper and 2 thou runout 3" out from the collet the axis of your collet is not parallel with the axis of your spindle. Mark the high spot on the spindle and on the shaft. Rotate the shaft 90 degrees and check runout again. Runout due to the shaft will move with it.

When you did the test with both the collet and the live center how long was the piece of shaft?

Are you saying that you see this problem when you turn between centers?

The back bearing is the part of the inside of the bullnose where the back part of the collet is a precision fit.

[Harold_V]

Because the condition isn't duplicated when using a chuck, you can pretty much eliminate the spindle as the problem.

One wild thought comes to mind. 5C collets have a register at the threaded end. If your collet adapter has an oversized register, it may be permitting the collet to deflect. That would account for taper.

That said, if the tailstock is offset slightly, it would cause the part to deflect, allowing the taper to develop, in spite of support. I can't help but think your problem is twofold.

Too bad I don't have a trip planned to Portland in the near future. It would be interesting to see this in person. Mean time, try to get a measurement of the small register located in the collet head/adapter. It should be VERY near the same diameter as the register you find on the collet. If it is oversized, you may now understand why the collet deflects, permitting taper to generate.

It should be noted that a burr can not cause this to happen, as the deflection would be only in one direction. It would create a problem with parallelism, but it would NOT create taper. That happens only when the rear of the collet is not held rigid and is oscillating, always deflecting due to tool pressure, as the spindle rotates.
Harold

[BadDog]

I think he's using the wrong word.

"Run out" would indicate revolving about the axis, like a bent rod would do. I think what you are describing is offset. In that case a DI zeroed at the collet reads zero while rotating the spindle. Moving down 3" to the end of a precision bar should continue to read zero, and still continue to read zero as you rotate the spindle. But, if I understand correctly, when you move down the DI shows 0.002 deflection and remains consistent while rotating the spindle (zero in an ideal world anyway). If so, I would call that "offset", and assuming no deflection it would indicate a total taper of 0.004 (+/- depending on deflection direction). If it does oscillate when indicating the end away from the spindle, then that is run out.

If it is offset, then the end of a 3/4" bar hanging out 3" might well be expected to deflect 0.002 due to cutting pressures, producing a 0.004" taper. Reducing hang out, or reducing cutting pressure will help correct. If you still get a similar results with a properly set live center, then it is possible your tail stock is not zeroed and is in fact offset by the indicated amount. But the coincidences are starting to stack up.

If it turns true with a chuck, it may be because the engagement of the length of the jaws is much longer than the engagement of the collet, thus helping reduce deflection by added lever length.

[ctwo]

These last two posts would seem to indicate taper in the opposite direction than what is being reported.

If a chip found its way into the collet and the part was held in a tail stock center, would that not produce a larger diameter at the tai lstock end and a smaller diameter at the collet?

But I think this problem is being reported as well without use of tail stock (3" stickout). The only way I can think of getting taper this way is if the spindle axis is not parallel to the ways.

What if the collet chuck axis is not parallel to the spindle axis or is offset? Can it be mounted this way?

[John Hasler]

Because the condition isn't duplicated when using a chuck, you can pretty much eliminate the spindle as the problem.

One wild thought comes to mind. 5C collets have a register at the threaded end. If your collet adapter has an oversized register, it may be permitting the collet to deflect. That would account for taper.

That said, if the tailstock is offset slightly, it would cause the part to deflect, allowing the taper to develop, in spite of support. I can't help but think your problem is twofold.

Too bad I don't have a trip planned to Portland in the near future. It would be interesting to see this in person. Mean time, try to get a measurement of the small register located in the collet head/adapter. It should be VERY near the same diameter as the register you find on the collet. If it is oversized, you may now understand why the collet deflects, permitting taper to generate.

It should be noted that a burr can not cause this to happen, as the deflection would be only in one direction. It would create a problem with parallelism, but it would NOT create taper. That happens only when the rear of the collet is not held rigid and is oscillating, always deflecting due to tool pressure, as the spindle rotates.
Harold

The register at the threaded end is what I called the "back bearing" (Hardinge's term) above. A burr there would angle the collet (which as you note excessive clearence there would do as well).

The problem description does seem contradictory. I think it would be better not to use the live center for these tests.

[John Hasler]

I think he's using the wrong word.

"Run out" would indicate revolving about the axis, like a bent rod would do. I think what you are describing is offset. In that case a DI zeroed at the collet reads zero while rotating the spindle. Moving down 3" to the end of a precision bar should continue to read zero, and still continue to read zero as you rotate the spindle. But, if I understand correctly, when you move down the DI shows 0.002 deflection and remains consistent while rotating the spindle (zero in an ideal world anyway). If so, I would call that "offset", and assuming no deflection it would indicate a total taper of 0.004 (+/- depending on deflection direction). If it does oscillate when indicating the end away from the spindle, then that is run out.

If it is offset, then the end of a 3/4" bar hanging out 3" might well be expected to deflect 0.002 due to cutting pressures, producing a 0.004" taper. Reducing hang out, or reducing cutting pressure will help correct. If you still get a similar results with a properly set live center, then it is possible your tail stock is not zeroed and is in fact offset by the indicated amount. But the coincidences are starting to stack up.

If it turns true with a chuck, it may be because the engagement of the length of the jaws is much longer than the engagement of the collet, thus helping reduce deflection by added lever length.

I've been assuming that he is not cutting the precision bar: just measuring it.

[spro]

The tail end of the collet tube may be off under tension.

[BadDog]

I've been assuming that he is not cutting the precision bar: just measuring it.

My comments at various points were regarding use of an uncut precision shaft, and also the taper he reported producing when turning (regarding cutting forces pushing the end away resulting in an over size diameter at the tail stock end).

[rexcsmith]

I don't have a precision bar yet, although I have a hard steel bar, a mild steel bar, and a drill rod bar. They all behave the same. It is runout because when you rotate the uncut bar, the dial indicator moves between 0 and .0002" When you measure the diameter after a cut, the end of the bar is .0004" larger than the collet end.
I'll make more measurements tonight, now that I now more about what I'm looking for. thanks, rex