Need Help!- Turning Thin-walled Castings ???

[G0G90]

I'm trying to solve an issue with part deformation on a lathe and holding roundness.

Has anyone ever used or is there a commercially available "expanding ring" to support the part from the ID?

Part is; 20" OD, 19" ID, by 4.5" long Gray Cast Iron. Tolerance on OD/ID is +/- .0015", roundness is .003"
Yes, I realize size controls form however, we cannot hold this part without deforming it. Typically, around .005 - .01" even at low clamping pressures.

I've designed an adjustable "spider" to install pre-machine however, lead times aren't going to be pleasant.

????

[RICHARD ZASTROW]

Try filling it with low temp melting metal or plastic or wax.

We filled 60" dia. turbine blade assemblies and completed the machining after filling the assembly with a hard wax.

After machining, the wax was melted and removed. This method prevented distortion caused by machining the thin cross sections of the turbine blades.

Dick Z

[Geof]

Even if you solve the problem of holding it without introducing distortion that tolerance on that diameter is tight for gray cast iron. Ductile or malleable maybe, but gray will probably move when the stresses are altered by machining.

[G0G90]

Originally Posted by Geof Even if you solve the problem of holding it without introducing distortion that tolerance on that diameter is tight for gray cast iron. Ductile or malleable maybe, but gray will probably move when the stresses are altered by machining.

Yea, that's going to be my NEXT concern. I figured so however, they've been holding these tolerances on a machining center with boring tools. Of course, they need to hand-load the tools so, that's not an option I like to use.
They're also bolting this onto a plate and probing the center so, throwing this into a lathe is going to be a trick.

I'd love to use the Wax / soft metal idea however, production volumes, even as low as they are, won't justify.

I'm going to attempt the "ID-spider" concept by using a plate with jack-screws to prove the concept. (and get some production accomplished)

... I'll post my results next week - THANKS for the suggestions!!!

[Geof]

Originally Posted by G0G90 .....They're also bolting this onto a plate and probing the center so, throwing this into a lathe is going to be a trick...

There is your answer, maybe.

What is your length tolerance, probably more generous than the ID/OD tolerance.

Make a crude spider so you can grip the OD not worrying about a little bit of diameter distortion, and simply face the ends to length.

Now make a plate with two rings of four tapped holes for clamps, one inside the ID the other outside the OD, put the holes 45 degrees out of alignment. Clamp it onto the plate with the clamps only on the ID, chuck the plate in a four jaw and true it up. Rough the OD to within 0.05" or thereabouts. Now put on clamps from the OD and remove the ID clamps. The part will not move out of alignment overall but there may be some localised movement due to stress relief. Now rough the ID to within about 0.05", change the clamps again and dial the OD to see just how much things moved. If the movement is minimal finish the OD, change the clamps and finish the ID. If there is a lot of movement creep up on the final sizes in two, or more, steps.

As well as avoiding chucking distortion and minimising stress relief distortion, this should be faster than doing one diameter and then having to dial in again for the other diameter. Especially if there is a lot of stress distortion and it is necessary to creep up on the final sizes.

[G0G90]

Originally Posted by Geof What is your length tolerance, probably more generous than the ID/OD tolerance.

You're correct - quite generous.
I did fail to mention that there is drilling on the ends and milling done to the periphery hence, this will be put into an Hitachi HiCell - live tooling lathe.
I need to keep this simple. Low volume with production tooling parameters really limits me to what I want the operators to be in control of. (dialing-in is not a valued option)
The whole trick to this part is from raw. If we can establish a diameter (either ID or OD) that's round, I can maintain this association on secondary operations using a self-centering, plate style setup.
I do not have a four-jaw chuck, as much as that would be perfect, I cannot find a 24" one readily available. Oh, nor a tailstock for that matter.

I'm also not a huge fan of using clamps on a lathe. But, looks like I might not have a choice on this flimsy thing.

Thanks again Geof!

[TechGuy5002]

Originally Posted by G0G90 I do not have a four-jaw chuck, as much as that would be perfect, I cannot find a 24" one readily available. Oh, nor a tailstock for that matter.

I may not have a good understanding how your machining this part. But what about cutting a jig out of wood that is bolted on the lathe spindle. Perhaps using a metal backplate that screwed or bolted to the wood jig that the part fits into. You could laminate pieces of wood to get the oversided dimensions, and machine out the wood to fit the shape of the part. The wood should be strong enough for a short run production. If you feel wood isn't sufficient they perhaps Alum. Jig plate disks would work.

[uli12us]

If you first turn the inner diametre to a fine tolerance, you can make simple an aluminium bolt that has a little bit bigger diametre at room temperature. If you cool the alu part down to 0F it will shrink and you can put it in your cast part. After maching you give your complete part in the fridge and the alupart falls out.

[G0G90]

That's GOOD and I have been looking at something like this but, in a production environment AND the part is around 20" ID, not a very feasible solution. - but, THANKS!

[Joe NH]

If you had a tailstock.
I have had good luck turning delicate castings round by chucking very lightly on either the OD or ID, usually just enough to center the part without distorting the out of round shape.
Then using a plug fitting the other end fairly closely, but again not to distort. The plug has a hole allowing the tailstock to seat in. Using pressure from the tailstock to drive the part, it can carefully be turned round.

This is only a suggestion of something I have done successfully. If you try it proceed with caution.

Joe

[Joe NH]

I just noticed that you have no tailstock.

Joe