HELP! machining both faces at same time and still not parallel

i have a 2.75 dia OD shallow cup looking part. .350 oal with a few steps(turned dias on one side ) and a 2.0 dia flat face on the other side that tapers 2 deg. the OD also tapers 2 deg. the cup side is 2.5 ID and about .200 deep.

On my Mori i finish the ID and rough the steps and OD cutoff

then on my 16c lathe with an expanding id collet i finish the face and large od and 2deg angle with a OD Turn Tool, then i reach behind the od with a top notch and finish the steps and OAL(back edge)


I cant seem to get better than .0015 parallel, needs to be .001

All dia tolerances are +/-.001 with exception of the ID

I will say they have been made on these machines before but i just cant seem to get them this time. ive even switched to different heats from different mills and still no luck. Im hoping the Gods on here can lend a hand.

I cant post a drawing for aerospace reasons

You've put an indicator on the collet chuck and checked runout?

Check your chucking pressures, gripping over hard in either direction will cup / dish ''washer like'' parts.

Are you parting them off ?
Maybe they are not flat after parting or sawing.
What Material ?

It od true to chuck on? Some times aligning on a mark helps you understand what happens as when you orient the part very close to the same. It you use two different machines the clamping is different also.

Hi tay2daizzo8:
Your chucking is obviously the problem, what's not clear is which detail is the culprit.
When you ID chuck on the expanding mandrel, are you registering on the bottom face of the pocket you turned in OP#1?
Is there a little corner radius on the inside of that pocket?
Could you be fouling on the radius with the corner of the mandrel?

Related to that, is the corner of the pocket as nicely finished as the rest of the pocket?
Are there burs, tears or other imperfections in the corner?
Did you look under a microscope?

Next question: is this a dead length mandrel?
Are the registration pads for the length stop running perfectly true when the mandrel is unclamped?
Does it run true when it's clamped.
Did you face it to run true in the machine?
If you did, did you debur the slots perfectly?
Does the mandrel have to expand more than a few thou in order to clamp?
Are you squeezing it hard when you clamp it?
How big are the pads compared to the registration surface on the part?

I assume you turned and faced the mandrel when it was in the clamped condition...what did you expand it against and where relative to the business end?

How are you loading the parts for OP#2?
Hand load and hit the foot switch?
Robot?
Does it get a belt with a rubber hammer or is it just load, finger push, and trip the foot switch?
Have you tried blueing the ends of the mandrel jaws? Do they touch the registration surface of the part evenly?

Somewhere in all that verbiage, is the answer to your complaint.
Now you just gotta find it.

99 times out of a hundred when I've had this sort of problem it's been fouling in the corner so I'd look there first.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Whoops...I've been bullshitting you...I re-read your post and realized you're turning it to final length in the second OP by back facing it.
It's not out of parallel with respect to the bottom of the pocket...it's not parallel front face to back face even though you turn both in one operation.

You gotta be chucking it too hard.
Your wall is only 1/8" thick and the part is 2.75" diameter.
It's likely expanding unevenly as you clamp.

Cheers

MC

Trying to turn the faces at the same time using a collet is not a stable solution.

The right approach is to (1) face one side, (2) turn the taper, (3) face the other side.

Doing things in this order will allow you to indicate to 0.0002" to make sure your cut will be parallel.

Instead of holding the work with a collet, a bushing with the inverse taper should be constructed to hold the work against the chuck or face plate.

In general, whenever you have an awkward work piece, the solution is to make the right fixture.

the other factor not discussed here is the material, if it is cold worked or continuous cast or anything but fully annealed you can have some residual stresses etc. "different heats from different mills" doesn't really say much...

I had a little trouble following your description of your part. Maybe a little trouble shooting can help. From your description you must be chucking on the ID with the bottom of the ID against the end of the ID collet. That would allow facing both ends. After finishing the part but before opening the collet indicate both faces. If both ends run zero TIR as expected you have a distortion problem caused by the collet being too tight, not perfectly round, radii interference, or something. If not zero TIR the part is shifting between the 2 facing operations. Identifying the cause will lead to a solution. A test of collet repeatability would be to mark the collet and part and rechuck. Readings the same? Rotate the part 180 degrees and recheck.

If you have an OD that is parallel enough to chuck on you might change the order of operations and finish the OD, machine both faces and cutoff from the ID side a little long. Second OP would be to drill and bore the ID. I do this on one of my parts.

HI guys, the material is 304. Im glad to report the machining gods have changed there mind and we were able to run off 20 parts in a row to complete the order. I am convinced it was the material. all 20 good pcs came from a new bar. the collet pressure was not distorting the parts or running out. i verified this by indicating both face and od...also, we were getting the same results using aluminum ID jaws. My thought is since the material was bigger (3.5 dia instead of 2.75) there wasnt as much heat and stress introduced in the cutoff operation. the next run of these i will be using sawed off slugs to eliminate the cutoff.

Thank you all for your help and ideas, greatly appreciated.

cyanidekid....different heats from different mills says exactly that, not sure where the confusion is. one bar was outokumpo, one bar was NAS and the bar that finally worked was valbruna

Thanks for the feedback something so often missing on this forum.

tay2daizzo8 said:

HI guys, the material is 304. Im glad to report the machining gods have changed there mind and we were able to run off 20 parts in a row to complete the order. I am convinced it was the material. all 20 good pcs came from a new bar. the collet pressure was not distorting the parts or running out. i verified this by indicating both face and od...also, we were getting the same results using aluminum ID jaws. My thought is since the material was bigger (3.5 dia instead of 2.75) there wasnt as much heat and stress introduced in the cutoff operation. the next run of these i will be using sawed off slugs to eliminate the cutoff.

Thank you all for your help and ideas, greatly appreciated.

cyanidekid....different heats from different mills says exactly that, not sure where the confusion is. one bar was outokumpo, one bar was NAS and the bar that finally worked was valbruna

Click to expand...

there was no confusion at all. I was just saying that it might be the material, and you hadn’t included any info on that, such as wat it was, what size/shape you were cutting it out of,what state of work it was in, etc. happy you got the parts out.