Anyone got Experience Lapping Internal Spindle Tapers?

[awemawson]

I've spent quite a while making a rather oddball adaptor to fit an internally operated A2-5 hydraulic collet chuck to my A2-6 CNC lathe spindle. It all works rather well including complicated internal oil galleries for open and closing, and the tapers have worked out 'dead nuts on' with no measurable run out.

However this second hand collet chuck is giving me about 15 microns run out. OK it's probably 30 years old and not had an easy life but I can only think this run out is in it's taper that the collet is pushed into to close (so not the mounting taper) and 15 microns isn't too much to lap I wouldn't think. But how to go about it?

Any suggestions welcomed!

 

[JMS1972]

On a manual lathe you'd dial in the taper on the top slide and use a toolpost grinder.

 

[Brad93]

That's one for robin renzetti

 

[awemawson]

On a manual lathe you'd dial in the taper on the top slide and use a toolpost grinder.

No top slide of course but I can program the motion and I have an air grinder spindle from a jig borer. What is puzzling me is the internal taper presses the collet against the front plate (that you remove to change collets), and I'm not sure if the front plate is removed to expose the taper that the whole moving bit won't be pressed out complete with a bunch of springs and pistons !

This is the rear view showing the guts of it - springs are sitting on the moving bit and the pistons are between the springs

 

[Brad93]

Have you tried different collets? You could have a bad collet

 

[awemawson]

To be honest Brad they probably are contributing but yes I've tried a few and also had them apart and cleaned them (they are Crawford Multibore M673's)

I can find no figures for what concentricity I should expect from this range however much to my surprise I found one manufacturer quoting 30 micron for 'standard' collets and 20 micron for 'precision' in this size (30 mm that I was last testing.

Seems an awful lot of run out to me but perhaps my 15 microns isn't too bad after all !

 

[JMS1972]

What

No top slide of course but I can program the motion and I have an air grinder spindle from a jig borer. What is puzzling me is the internal taper presses the collet against the front plate (that you remove to change collets), and I'm not sure if the front plate is removed to expose the taper that the whole moving bit won't be pressed out complete with a bunch of springs and pistons !

This is the rear view showing the guts of it - springs are sitting on the moving bit and the pistons are between the springs

No top slide of course but I can program the motion and I have an air grinder spindle from a jig borer. What is puzzling me is the internal taper presses the collet against the front plate (that you remove to change collets), and I'm not sure if the front plate is removed to expose the taper that the whole moving bit won't be pressed out complete with a bunch of springs and pistons !

This is the rear view showing the guts of it - springs are sitting on the moving b

That looks quite complicated

Is the bit you want to grind effectively floating inside the clamping part? Wondering what surface keeps it all concentric? Maybe that sliding face is worn?

 

[awemawson]

It all seems pretty tight and rather nice. The female taper doesn't look to good BUT at last I've found Crawford concentricity figues and it looks like I'm chasing phantoms and my 15 microns is within specification, which I find MOST surprising!

 

[fizzy]

How do you know the ground bar is totally straight?
I was also surprised that collets aren't as accurate as I thought. I use Jacob Rubberflex on my lathe. I naively thought that they would be good enough to take the work out - put it back in and they would be spot on

 

[ronan]

Its like everything else, if you spend good money on a decent set of collets i'd reckon they would be a lot more accurate than the cheap and cheerful chinese/indian stuff

 

[fizzy]

Its like everything else, if you spend good money on a decent set of collets i'd reckon they would be a lot more accurate than the cheap and cheerful chinese/indian stuff

Oh yes.
I can only measure the runout on my Jacobs collets. I can see it on my ER32!

 

[Brad93]

Most modern ER collets are only 20 microns. Think the precision ones from Cutwel are 10 microns

 

[awemawson]

How do you know the ground bar is totally straight?
I was also surprised that collets aren't as accurate as I thought. I use Jacob Rubberflex on my lathe. I naively thought that they would be good enough to take the work out - put it back in and they would be spot on

I ground it myself on my J&S 1300 EIUR cylindrical grinder as a 'proof of parallelety' before grinding the precision tapers on the adapter that I made for this chuck. It's as straight and true as I can measure

 

[fizzy]

I ground it myself on my J&S 1300 EIUR cylindrical grinder as a 'proof of parallelety' before grinding the precision tapers on the adapter that I made for this chuck. It's as straight and true as I can measure

Lol that is more than adequate I am sure.
Jealous now