What's new
What's new

Weird problem turning a shaft

Sam L

Cast Iron
Joined
Oct 12, 2007
Location
NJ
I think i have finally lost it.
Time to get a hot dog cart.

We have been machining inconel shafts.
Side 1 finished no problem.
Side 2, machined to length and checked from a shoulder to within .001"
Machined rest of shaft.
The original dimension that we checked to within .001" is now .017" short.

I cannot come up with a reasonable explanation.

In the picture the yellow arrow points to the shoulder on the left of the steady rest and the other arrow points to the face that we checked the length to before machining the rest of the shaft.

this has happened twice on the first 2 shafts we are finishing. The first one i assumed the operator made a mistake.
The second one i checked myself with very similar results.Shaft.jpg

If anyone has heard of this, i would love to know what the hell is going on.

We made one out of steel first and it came out perfect.

Thanks
Sam
 
I wouldn't scrap another shaft until I got it figured out.

Based on experience, it's one of a few different things.
1. Tailstock pushing the part into the Chuck.
2. Steady/Follow catching on a feature and pushing the part.
3. One of the Tools pushing the part. Usually a HSS Drill, a Turning Tool working very hard, or a teeny tiny bump that isn't even noticable.

I cannot see whether your Chuck jaws are through or stepped. Either way, usually that's where the error happens.

R
 
Pushing the part should result in the second side being long, Rob?

Both parts coming up exactly -.017", sounds like either an offset issue to me, or a negative W finish allowance on a G71 cycle type problem.

Or- depending on how the tools are set and TNRC is handled, possible TNR error.
 
Check all your cutting tools, look for damaged edges. Adjust process to either replace tools when dull (establish cut life on bad shaft), or come up with arbitrary max time in cut and replace (accepting excess tooling cost rather than bad parts).

Do another test part from a hardened steel, trying to match hardness/toughness with 718 values. See if new process parameters gives a good part.

And talk to the OP in this thread: Having difficulty machining Inconel maybe his stock is the right size (or close enough) for your project.

[I kid, but hell, if the certs and numbers are right at least you could use it for test parts]

['Nother edit - you don't mention the type of Inconel or hardness. Anything lesser than hard 718 shouldn't be an issue, but shift happens, if you get my drift. And that looks like an expensive part to scrap, you have my condolences]
 
Is the part heating up?...Phil

I dont think its getting hot. The part is flooded with coolant during the machining cycle.

Once the length is set, the front face is not cut again.

How can it possibly shrink .017" over 18.625" length?

I was thinking about residual stress in the bar from the manufacturing process, but I have been making shaft for over 25 years in all sorts of metals. This is the first time I have seen this.
 
I'm going in tomorrow and trying another one.
This time i will leave it .050" long, then rough the OD.
I will then measure again.
Adjust the length as necessary. then finish machine.

Hopefully this will take care of the problem.
 
Once the length is set, the front face is not cut again.

Do any tools come close to that shoulder after you first measure the distance from it to the face?

Are you making the measurement in the manner both times? Or like is the first one taken in the machine and the second one taken out the machine?

After your 25 years I’m sure it’s something else but the number of times I’ve fixed broken devices by checking to make sure they’re plugged in....mot that that’s ever happened to me haha!
 
Check your cutting edges as you go! Clean in the machine, take pics after various cut times so you can judge wear and when to replace inserts.

Checking part temps after cuts might be helpful too - more data is good.
 
Pushing the part should result in the second side being long, Rob?.

Depends on how the operations are ordered and processed. If it's just about done on the 1st Op. then get pushed .017" on the Finish pass (as an example). Then when you flip the part for the 2nd Op. it's gonna end up short.

The coefficient of Inco 718 is 13 µm/m-°C. So doubtful it shrinking .017". But that's based on a picture....

My guess is push back, or inspection error at some point.

R
 
Went in on Sunday and tried another shaft of the same Heat.
Faced it to length again but I left it .050" longer than needed.
Drilled center hole for thread and left good center for tailstock.
Measured again and got the same dimension, 18.680" from machined face to shoulder on other side of shaft.
Roughed using ceramic 3/8" button, used flood coolant. The material was warm to the touch after roughing.
Measured again and the dimension changed to 18.664".
The face was not machined again.

I got the cert from my customer for the material.
I have a call into the manufacturer to see if they have an explanation.
The only thing I could come up with is some kind of linear stress in the bar, once we cut away material from the OD the center shrinks a bit. It sounds strange to me too, and i wouldn't believe it myself if i didn't see it with my own eyes.

By the way, its Inconel 625

Any thought would be appreciated.

Thanks
 
In my experience any smooth shaft will try to slide in a chuck. Get a dimension from the chuck face to a machined feature both before and after roughing as well as after a cool-down period.

That will tell you if it's moving in the chuck or not.

BTW, TG&P shaft (turned, ground, and polished) has the same tendency. I have to take a lighter DOC and feedrate to stay in tolerance.

Just my $0.02.
 
if you measured with a mic or calipers your mesuring the same dia. on both ends. stick the flange part on 123 blocks and run a indicator on the face of the shaft. see if you have any taper and its square.
measuring the length with a mic or even calipers wont tell you if you have a dished in face.
you could even do it on the machine, run an indicator down both faces and see if you have any taper(index it 3 times to see if its quare in there also.

I have had inco shafts shrink when running ceramics taking off alot of material on the O.D. but not that much on a solid bar. on parts with ids in them yes the faces will change a ton in some instances.
 
Went in on Sunday and tried another shaft of the same Heat.
Faced it to length again but I left it .050" longer than needed.
Drilled center hole for thread and left good center for tailstock.
Measured again and got the same dimension, 18.680" from machined face to shoulder on other side of shaft.
Roughed using ceramic 3/8" button, used flood coolant. The material was warm to the touch after roughing.
Measured again and the dimension changed to 18.664".
The face was not machined again.
What are your starting and finishing diameters?

Think about when you are making a large bushing. Turn the OD, measure. drill out and bore the ID, measure the OD again, it's shrunk. You've relieved all that internal stress, and it moves.

The same thing happens when you turn the OD on a shaft, you just don't have a reference to measure against. You relieve the stresses, it wants to "grow" radially. That has to come from somewhere so it gets shorter. .017" is a lot for 18" though!

I'd be curious if you took it down in steps. Take half the material, measure your length. Take the remainder, check again. See if the length is shrinking in a predictable way.

Can you change up the order? Rough it out, then establish the length, then just a finish pass?
 
Are you making your finish cuts with sharp carbide? Ceramics by their nature can put a lot of stress into the workpiece, so using carbide for finishing should give a more stable final part (presuming ceramic-induced stress isn't crazy-high).
 
In my experience any smooth shaft will try to slide in a chuck. Get a dimension from the chuck face to a machined feature both before and after roughing as well as after a cool-down period.

That will tell you if it's moving in the chuck or not.

BTW, TG&P shaft (turned, ground, and polished) has the same tendency. I have to take a lighter DOC and feedrate to stay in tolerance.

Just my $0.02.

Its in a collet chuck.
Even if it slides, I'm not measuring from the face of the chuck.
I'm measuring from a feature that is machined into the other side of the bar.
 
if you measured with a mic or calipers your mesuring the same dia. on both ends. stick the flange part on 123 blocks and run a indicator on the face of the shaft. see if you have any taper and its square.
measuring the length with a mic or even calipers wont tell you if you have a dished in face.
you could even do it on the machine, run an indicator down both faces and see if you have any taper(index it 3 times to see if its quare in there also.

I have had inco shafts shrink when running ceramics taking off alot of material on the O.D. but not that much on a solid bar. on parts with ids in them yes the faces will change a ton in some instances.

We checked the face with a ground bar.
No dishing or cupping. Besides, .017" would be quite dramatic on the face.
 
What are your starting and finishing diameters?

Think about when you are making a large bushing. Turn the OD, measure. drill out and bore the ID, measure the OD again, it's shrunk. You've relieved all that internal stress, and it moves.

The same thing happens when you turn the OD on a shaft, you just don't have a reference to measure against. You relieve the stresses, it wants to "grow" radially. That has to come from somewhere so it gets shorter. .017" is a lot for 18" though!

I'd be curious if you took it down in steps. Take half the material, measure your length. Take the remainder, check again. See if the length is shrinking in a predictable way.

Can you change up the order? Rough it out, then establish the length, then just a finish pass?

Our starting diameter is 4", Finished is 3.5" down to 1.75"
We started roughing it out leaving .050" on the face.
It is still shrinking predictably around .017"
We then just finish to length, and then finish all the OD's and the remaining features without incident.
I'm curious to see what happens with the next heat of material.
 
Are you making your finish cuts with sharp carbide? Ceramics by their nature can put a lot of stress into the workpiece, so using carbide for finishing should give a more stable final part (presuming ceramic-induced stress isn't crazy-high).

Yes, we are using carbide to finish.
The roughing is whats causing the "shrinkage"

Hopefully i hear from the mill today regarding the material.
 
Its in a collet chuck.
Even if it slides, I'm not measuring from the face of the chuck.
I'm measuring from a feature that is machined into the other side of the bar.

Right, but that will tell you if it's moving.

Making sure it's not goin' anywhere would be the first order of business IMO.
 








 
Back
Top