Learning the hard way - distortion caused by chucking forces.

I faced off a batch of brass discs to 1.5mm thickness the other day holding the work by a 16mm 'spigot' on the rear of the work in the three-jaw:

When I started to polish the faced surface I was surprised to see:

The dark areas are 'low' with respect to the rest of the surface, probably by ~40µ at deepest.

I assume that the disc was distorted and 'bulged' in the areas between the chuck jaws, then relaxed and that's what caused the hollows.

I could have used a collet chuck, but being lazy I didn't want the faff of changing chucks. I have paid the price in time/elbow grease/sheets of wet'n'dry!

The point of this post is just to document one of the many things that don't work*.

Robin.

* "But the student will find that experience is the best teacher. The reason why I get along with comparative ease now is because I know from experience the enormous number of things that won’t work." 1882 June 15, The Christian Union, How to Succeed As An Inventor by Thomas A. Edison, Page 544, N.Y. and Brooklyn Publishing Co.,

Beautifully illustrated, Robin

MichaelG.

Ditto

I don't think the chuck forces caused it to bulge. More likely the 3 jaws supported the work infront of then and the rest distorted away from the tool leaving them slightly thicker as they spring back. Though you may find the hole is distorted where the chuck has closed up the thin spigot in 3 places.

Lighter cuts, sharper tool or some form of rear support would cure it, collet would just see a conical shape as the whole outer unsupported area get pushed away for the whole 360deg.

looks like bronze or GM which can do that more then other metals

Edited By JasonB on 05/09/2022 07:17:35

Just to emphasise Jason's point above, the distortion is not due to the radial gripping action or squeeze of the chuck jaws, but due to the work springing away from the tool due to the axial force when cutting. The shadow of the jaws is because there is resistance from the rear (the structure is stiffer at those points).

You could test this by using a longer spigot so there is an air gap between back of work and face of jaw. It would then be conical and concave.

Maybe a better quote would be "everything is a spring" (Thomas Hobbes, loosely translated from De Mirabilibus Pecci).

So a collett is the way to go

apparently

.

Probably not … the workpiece would likely then ‘sing’ something awful

I think what Robin needs is a backing plate

… Cue discussion about ‘variations on the wax-chuck theme’

MichaelG.

Edited By Michael Gilligan on 05/09/2022 08:22:16

Making a block that holds the part and supports the back of the face being cut is important if the flatness really does matter. You have done a very good illustration of the distortions that can happen when facing a part.

I would have bored the centre last, holding it in external jaws. You are absolutely right in decscribing the distortion process.

By the way, on my monitor, the metal looks like steel, not brass.

Edited By old mart on 05/09/2022 14:43:22

Agree totally with Jason, that the meta, being thin, l is likely to have flexed away from the tool due to cutting forces.

Think in terms of trying to face a sheet of paper !

For metal, of that thickness, the tool needs to be REALLY sharp, set dead on centre height and the cuts need to ,be very shallow to minimise cutting forces.

If that doesn't work, make up a thick backing washer to support the disc being faced.

Howard

A ring at the rear of the job would help as would filling the bore with a plug.

The pictures show almost conclusively that the chuck jaws must have had an influence on the surface profile of the part. Whilst I agree with that statement I am curious as to the actual mechanics or nature of the influence.

If the part had the proportions of a sheet of paper then support would be essential but I dont consider 1/16" at about 2" diameter to be anywhere near 'thin', brass is a rigid hard metal and it surprises me that, especially for a finishing cut, it flexes enough between each chuck jaw to leave a witness on the face.

I wonder if the part had been chucked with gap left between the jaws and the workpiece, whether the part would have ended up flatter? (as no influence from the jaws). My money is on (as speculated by the OP) that it was the radial pressure of the chuck jaws on the spigot that caused the distortion. Having said that the spigot itself looks quite substantial and I would have thought that it easily coped with the cutting forces involved.

Just food for thought

Ian P

Thanks for replies. I should have said that when I noticed the effect I checked the spigots for witness marks showing where the chuck had gripped. They were there, and the jaws had gripped the spigot at (eyeball accuracy!) 60° to the depressed areas, so the front faces of the jaws were behind the 'raised' areas. That observation got incorporated in my thinking about the cause and was implicit in my speculative explanation, but I should have mentioned it explicitly. Sorry!

If I have understood the axial deflection explanation correctly the pattern should be the other way round if it is right.

The discs were faced using a fresh 0.4mm radius CCGT insert (Arc Euro Trade), so sharp(ish). Roughing cut ~0.5mm, finishing cut ~0.05mm. So very little axial cutting force for the finishing cut I would have thought?

I take old mart's point that it might have been better to make the hole last - sequencing of machining operations is an aspect of this hobby I find particularly challenging. I guess it comes with experience - I didn't anticipate this effect so no amount of planning would have helped. The reason for the steely appearance is that I had to adjust lighting to display the contrast between the areas, and that must have mucked up the colours on my phone camera.

Anyone out there with a finite element analysis program? Well above my pension grade, but it would be interesting to see the stress field for work held in this way.

Robin

Edited By Robin Graham on 06/09/2022 00:10:28

Edited By Robin Graham on 06/09/2022 00:11:19

Edited By Robin Graham on 06/09/2022 00:12:58

Robin

Generic FEA of the part you are trying to machine. First picture shows the model with jaw forces applied. Second and third pictures both show the way the parts deflects, Values are not important it is the indicative pattern we are interested in. Blue is small, red is large relatively.

Regards

Dave

Wonderful FEA Dave, that's pretty much a perfect match.

Bernard's bung looks like a good idea.

I agree, Bernards bung would have prevented the distortion.

 There is nothing better than learning by your mistakes, I'm getting to the point where I remember the best way to do something after doing it the wrong way.

Edited By old mart on 06/09/2022 21:09:23

That's fantastic Dave, thank you! I don't fully understand the pictures though - are the deflections in the plane, normal to the plane or what? The pattern certainly seems to correspond to my observation.

Unfortunately I now have a scenario in my mind: apprentice asks Foreman to check his setup. Foreman - " That's no good lad, you need a Bernard's bung in it. Run along to the stores and get one."

Good idea for the future though, thanks Bernard.

Robin.

Robin

The part deflects along the axial plane. So the red maximum deflection zones bend backwards towards the headstock.

.

and therein lies the twist

Robin clearly stated “ The dark areas are 'low' with respect to the rest of the surface, probably by ~40µ at deepest.“

Which means the ‘mechanism’ is : the red areas move towards the headstock, thereby reducing the cut, and the depressions form when the workpiece is released.

Now … if we could use that in a controlled manner imagine what subtle surfaces could be machined !

MichaelG.

With CNC, MG's varied surfaces are probably possible.

We had some flywheel housings CNC machined

The code called for the mountimng to be milled round, by interpolation.

But when unclamped, the housing was oval by 0.004"

"No problem said the stter. He rreprogrammed the machine to mill the aperture 0.004" oval in the other plane. When unclamped, it sprang to round!

Howard