Drill Morse Taper off-centre

[Chris152]

We recently bought a nice old Denbigh pillar drill, all works fine but there's too much runout. At the bottom, outside edge of the spindle (level 1 in the photo), it's about 0.08mm off. The inside of the spindle/ widest part of the internal morse taper (MT1) is 0.12mm off. By the time we get to the bottom of the chuck (level 2 in the photo) it's 0.24mm off. As you can imagine, a drill tip will be further off still.

Using engineers blue, we could see that the bottom end of the MT was a bit damaged and we cleaned it up with a file, which improved things (the figures above are after that improvement). We've checked the spindle along its length for straightness, and it is to within about 0.04mm.

Sooo, for some reason, the MT itself seems to be off. It is not concentric with the outside of the spindle. We also tried reaming it a little with a MT1 reamer, in case there was any further damage, but it's made no difference.

I can't think what else to do, apart from pay someone to recut the MT to MT2, aligned properly with the spindle. Or is there anything else we should be trying before going down that route? And any ideas what might have caused the MT to be misaligned with the spindle?

Thanks for any suggestions/ thoughts..

Chris

 

[Brad93]

Part has snagged once upon a time in the drills life and probably bent the spindle or the quill.

Reamer won't fix that.

It needs to be ground or hard turned true to the axis of rotation which will be set by the spindle bearings.

This can be done on a Cylindrical / ID Grinder or in the lathe with a toolpost grinder, fixed steady and four jaw chuck.

 

[Morrisman]

Sounds like it was manufactured wrong, if the ID and OD of the spindle are not concentric. I’d imagine a precision grinding setup would be required, with the spindle trued up accurately to its bearing surfaces.

Can you verify, the outside of the spindle runs dead true? The Morse taper doesn’t.

 

[voipio]

If brute force and ignorance caused the runout in the first place, then an equal amount of brute force and ignorance in the opposite direction might improve the runout.

Alternatively, take the spindle out, and find out if, and where, it's bent, using an engineering reference of some sort. If it is bent, then careful application of force in the right place, using a press, might do it.

 

[Chris152]

Sounds like it was manufactured wrong, if the ID and OD of the spindle are not concentric. I’d imagine a precision grinding setup would be required, with the spindle trued up accurately to its bearing surfaces.

Can you verify, the outside of the spindle runs dead true? The Morse taper doesn’t.

The outside, bottom edge of the spindle is 0.08mm off; the inside bottom edge of the MT is 0.12mm off. So no, the spindle isn't spot on, but yes, as you say, the ID and OD are not concentric, and the MT is sending the chuck further off. ID and OD being not concentric is very odd, I'd think?!

We checked the spindle itself on our surface table, and it's straight to within 0.04mm along its length.

 

[voipio]

It seems like a classic example of how manufacturing and bending/straightness tolerances can add together.

0.04mm of spindle out of straightness might not sound like much, but it will show up as quite a bit more at the drill tip, so correcting that might be a starting point.

 

[dave h 2]

The Morse taper specs are over a longer length than the tapers supplied on tools.
I have a 2MT that has 1/2" of the taper exposed when fitted and others that are nearly flush.
Check the tongue is not interfering with a damaged taper extract slot. Try grinding the tongue to extend it a little further toward the chuck.

 

[Morrisman]

It seems like a classic example of how manufacturing and bending/straightness tolerances can add together.

0.04mm of spindle out of straightness might not sound like much, but it will show up as quite a bit more at the drill tip, so correcting that might be a starting point.

As was drummed into us in the college machine shop, first year apprentice:

Where you have movement you have clearance.
Where you have clearance you have tolerance.
Where you have tolerance you have error.

That was basically explaining why 3 jaw chucks are inaccurate, but referred to most everything else. Surprising the things that stick in your head, even nigh on 50 years later.

 

[Chris152]

I don't know if it adds anything to the equation, but the vertical movement in the spindle along the section containing the MT (using the handle) is 0.02mm.

 

[Seadog]

Are you using a Morse adaptor sleeve? The quill looks much too long to me.

 

[Chris152]

Are you using a Morse adaptor sleeve? The quill looks much too long to me.

No, that's just the spindle.

 

[pressbrake1]

That’s flung a work piece around
Ignore spindle OD
Only cure is to attempt straightening

 

[Chris152]

OK, further attempts at investigation, maybe you can help. I placed the spindle between centres on the lathe and measured left to right, noting down key shifts in depth from the indicator along the length. I did this 4 times, 90 degrees apart.

The measurements are recorded in the drawing, below (I've opened the cylinder into a plane). The numbers (X) are 0.0X mm, except where they become 0.XX mm. Also recorded (top row) are the diameters of the spindle, as there is some wear (it's a rather old drill press).

The indicator was zeroed at the left end, and as you can see, nearly all measurements are negative - I assume this is because of the wear. The top two (90 degrees apart) indicate far greater deviation away from the indicator.
If this is to be resolved by bending the spindle back (reducing the higher negative numbers - I'm not sure how to phrase this!) are we best to aim for about 0.05mm reduction of that deviation at the centre of the spindle? Ie, push the bottom two 0.05mm toward the opposite side?

Christ knows if what I've written makes sense, but please stick with me and ask for clarification, we'd love to get this sorted, in so far as we can. Cheers.

 

[Brad93]

Your lathe can't really be used as a reference because your tailstock could be out of alignment and you could have wear in the bed.

Get a flat surface, like a mill table for example, and use a pair of matched v blocks, and an indicator on a height gauge or scribing block.

Rotate the spindle looking for runout.

 

[Brad93]

Moving the v blocks you can find out where the bend is

 

[Brad93]

Personally i wouldn't worry about what the OD of the spindle is doing.

If the taper hasn't been bored concentric to the OD, no amount of bending or straightening is going to help.

You're only interested in the runout of the taper relative to the bearings.

You can use a tube micrometer or height gauge/indicator to measure the thickness of the taper wall to tell you if it's bored concentric with the OD

 

[Chris152]

Personally i wouldn't worry about what the OD of the spindle is doing.

If the taper hasn't been bored concentric to the OD, no amount of bending or straightening is going to help.

You're only interested in the runout of the taper relative to the bearings.

You can use a tube micrometer or height gauge/indicator to measure the thickness of the taper wall to tell you if it's bored concentric with the OD

Thanks Brad. I think we have to assume the taper is concentric to the spindle, I can't imagine how it'd be otherwise and anyway, as you say, it's a bit hopeless if that's not the case.

I did align the tailstock with the head a few weeks ago so that should be ok - as for wear on the bed, I've no idea. And any variable throws the whole thing out. We'll have another go on the surface plate, trying the same measurements at the same locations.

Cheers, Chris

 

[Brad93]

Thanks Brad. I think we have to assume the taper is concentric to the spindle, I can't imagine how it'd be otherwise and anyway, as you say, it's a bit hopeless if that's not the case.

I did align the tailstock with the head a few weeks ago so that should be ok - as for wear on the bed, I've no idea. And any variable throws the whole thing out. We'll have another go on the surface plate, trying the same measurements at the same locations.

Cheers, Chris

I'm not sure what you're trying to achieve by taking measurements of deviation.
All you're interested in is TIR to find where the bend is.

 

[8ob]

If brute force and ignorance caused the runout in the first place, then an equal amount of brute force and ignorance in the opposite direction might improve the runout.

Alternatively, take the spindle out, and find out if, and where, it's bent, using an engineering reference of some sort. If it is bent, then careful application of force in the right place, using a press, might do it.

I agree, I was always told when in doubt give it a clout

Bob

 

[Chris152]

I'm not sure what you're trying to achieve by taking measurements of deviation.
All you're interested in is TIR to find where the bend is.

Don't the measurements establish where the bend is - ie, deviation from flatness?

You wrote:
'Get a flat surface, like a mill table for example, and use a pair of matched v blocks, and an indicator on a height gauge or scribing block.
Rotate the spindle looking for runout.'

What am I misunderstanding?