Thermal expansion. Permanent? 1

[imcjoek]

I have an 8" ID ring made from 8620. The ring was case hardened to approximately Rc55 and ground to final dimensions. We made note of these dimensions at inspection of the part.

The ring fits over a shaft which is made from 4130 and was also case hardened and ground to size. There is approximately .0005 diametral clearance between the ring and shaft.

The shaft is heated internally with an electric heater to a working temperature of 400F. This was done about 5 times in the testing of our machine. We also measured the temperature of the ring during heat-up to assure that it heated along with the shaft (which it did within about 15-20F).

After dis-assembly we are finding that the internal diameter of the ring has grown .008" resulting in a loose wobbly fit on the shaft.

I don't see any evidence of wear (and there is no relative motion of ring and shaft, anyways). Can a steel part not return to original dimensions when heated and cooled? I am aware of some warp occurring after heat treat operations, but I don't see such a drastic dimensional change at only 400F as making any sense.

What mechanism could explain this?

 

[swall]

When you heat the ring, you are re-tempering it each time. This would normally cause the ring to shrink, as tempering causes a decrease in the volume of the martensite. But, it can't shrink like it wants to, since it is constrained by the shaft. The stresses imposed by the constraint of the shaft have acted on the ring, causing it to expand. You should also be observing a decrease in ring width, as the expansion has to come from somewhere.

 

[imcjoek]

I have just checked, and the ring did shrink in length. Quite significantly as well. .030 on a 3 inch width.

At least I know why it's happening. If I follow... the hardening heat treatment created a larger percentage of martensite. So would the dimensional change be less pronounced if the part were not case-hardened and also tempered prior to finished grinding?

Thank you for the answer.

 

[desertfox]

Hi imcjoek

I have just calculated the free expansion of the shaft and compared it with the free expansion of the ring, based on the ring being 20F below that of the shaft at the final temperature, what I find is that the shaft wants to expand by 0.0006389" more than the ring so you are physically forcing the ring to expand more than it wants to, this will also generate stresses within the ring and only makes the situation which swall as described.

desertfox

 

[mfgenggear]

wow never seen this much movement, but never the less it did

Swall
Good explanation.

To make sure the parts where stabilized & complete martensitic transformation.
would it have helped at the carburizing & hardening to have
stressed relieve,Then carburize, subcrittical annealed, austinitize, quenched, snap tempered, suzero stabilize, double or triple tempered.


imcjoek
was there any heavy machining after heat treat?
what is the temper for 55 HRc case hardness for both details.

Desert Fox
that is a good point
thermal expansion due to the heating at the machine cycle.

I would have expected the ring to expand larger than the shaft.

 

[desertfox]

Hi mfgenggear

The would have expanded more than the shaft (theorectally anyway) if the temperature of the ring wasn't lagging by 15F-20F, when there is only 0.0005" clearence it doesn't take much in terms of temp difference to eliminate clearence.
My ten pennies worth anyway can't comment on whats going on a metallurgical stage as I don't enough about it.

regards

desertfox

 

[imcjoek]

Thanks for all the comments.

The idea was for the clearance to be removed as the machine reached operating temperature. Hence the .0006 difference in expansion desertfox mentions was (somewhat) planned. The idea was for the ring to *just* snug up on the shaft when at operating temperature. This worked fine (the first time!)

Unfortunately I expected that the ring would shrink back down as it cooled, and obviously this did not happen.

mfgenggear: There was no major machining after the case hardening. Only finish grinding to final diameters. I do not know the details of the heat treat process as we leave that up to our vendor (we wouldn't know how to specify any better than they do...)

Perhaps the bloody ring should have been ceramic...

 

[arunmrao]

Just a crazy thought, did you consider freezing the shaft ,so that it contracts and then insert,followed by removing the refrigerant (dry ice). This might have avoided the phase transformation related dimensional instability.

Learn the rules,so you know how to break them properly.
Dalai Lama

_____________________________________

 

[mfgenggear]

imcjoek

It has been my experience the a sub zero treatment -100 Deg F or better yet -300 Deg F really helps with stability,
with the issue such as yours.

the double & triple temper also helps.

Ask your supplier for the process sequence, as it is important in the manufacturing game.
It is also important to be informed of the heat treat process. this where most failures happen.
discuss the options with the heat treater.

obtain material data sheets for the recommended process procedure, & if a specification is to be followed,
it is wise to really study the procedure specified.

good luck

 

[HgTX]

The ring didn't shrink back down as it cooled because the shaft was in the way. It still sounds like you considered the heating part of your process but not the cooling part. This is not "permanent thermal expansion"--this is heat upset, where the expansion is constrained to go in one particular direction but then the contraction is constrained to go in a different direction.

Hg

Eng-Tips policies: faq731-376