thermal expansion of a ring 1

[sydneyjongleur]

Hi,

I am wondering how to work out the thermal expansion calc of a ring with OD 1000mm and ID 800mm and L20mm. I
I am heat treating a nickel alloy ring at 1000°C and I am wondering how much the ring will grow at the OD, ID and L at this temperature. I am only looking for a rough estimation.

The linear thermal expansion from 20-1000°C of the alloy is 18.2 X 10-6.

Also, if I was to heat treat another ring with an OD 1000mm but with a bore of 300mm and L50 would the rate of expansion be the same. ie would the OD in both rings end up the same size at 1000°C?

 

[desertfox]

Hi
This link might help

http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/thexp2.html#c1

 

[dhengr]

Desertfox:
I think there have been other threads on almost this same topic, in the Mech. Forum, and you and I participated in them. Maybe a year or so ago. They had to do with shrink fits and the like, large gears over a shaft and what kind of growth or shrinkage do you get in all directions on a pipe shape. Did it grow inward and outward, and in length too, when heated. You seem to be better at searching than I am, and then placing the links in your posts. Would you do a search if you can find the time, and/or care to?

 

[IDS]

Also, if I was to heat treat another ring with an OD 1000mm but with a bore of 300mm and L50 would the rate of expansion be the same. ie would the OD in both rings end up the same size at 1000°C?

If you are looking for a rough answer, assuming uniform temperature increase, then surely you just multiply the dimension by temp rise * coefficient of expansion.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[dvd]

I would be curious about the constant linear expansion coefficient over that wide of a temperature range. Maybe it was bred to exhibit those traits?

Another point to ponder is what is the Young's modulus at 1000C, and is it low enough that there will be deformation of the ring due to its self-weight. Probably not, but may be some local deformation at fixturing. Also curious about creep deformation at this temperature, and whether fixturing solves that potential problem.

 

[racookpe1978]

time and temperature and and shape and how much force (the mass being heated) how the gear is supported all are going to determine how much deformation the gear is going to deform when it is heated.

A shrink fit "should be" on the order of 0.003-0.010 "overlap" between the shaft and the component, or - as I've done it - between the outside valve casing and the inner valve seat. Up to, perhaps, 0.012 overlap as I recall. (In our case, we had to cool the inside valve seat (the sleeve) with dry ice bits, and heat the outer part (the valve body) with electric pad heaters that had been previously placed to do the preheat weld heat treat and post weld heat treatment.)

If you start trying to get the outer (hot) part up over 1000 you'd better be dam*m sure you DON'T foul up the heat treatment and the metallurgy of the part. Support it soundly, and move it so it NEVER has a chance to "sag" or droop as you heat it and move it. Better to only heat it to 700 or so degrees, then cool the inner part. If you leave it over 1000 for several hours? Sure, expect some deformation and loss of machining control of the thinner (hotter) areas of the part.

 

[desertfox]

Hi all

There is one link


thread404-61378

 

[sydneyjongleur]

Thanks guys, I think I have the calcs now.

Some of the guys have been talking about supporting the part through fixturisation. That is something that I am looking at for a thin ring that I am trying to make round again.

I have a ring that is OD 1000mm ID 950mm and L 5mm. It is out of round and I was hoping to use heat treatment to make it round again. I am planning to support it circumferentailly and on the face using a fixture.

The material is still a nickel alloy and the HT would be at 1000 deg C again. I am tying to acertain the material of the fixture- I am thinking the same nickel alloy so that it has the same expansion rates. However, I am thinking because the fixture will be of a larger bulk I believe it will heat and cool at a different rates. I am concerned that the part wont be the correct dimensions never mind roundness when it comes out the oven.

 

[desertfox]

Hi

Yes you are right to be concerned about the mass of the fixture and the expansion rate due to differential heating of the component and fixture.
Also if you sandwich the thin ring between two plates any force exerted on the ring would have to be overcome during the ring expansion (i.e. friction), if your friction is to high it will leave the ring under stress.

desertfox

 

[Tmoose]

Is the heating just to correct out-of-roundness on the 1000 x 950 ring?
Or, does it need dimensional stability when heated during a later process or in service.

The reason I ask, depending on the roundness you require, surface finish and profile specs, and the future processes,
I think cold straightening ( bending and possibly including some peening) can be pretty effective. I speculate roundness within 0.005"/0.13 mm might be achievable, but there will be marks or even shallow dents on the ID and OD with peening.

 

[sydneyjongleur]

The rings are essentially finished components therefore dimensionally they will have to be good when the come out of the oven. The error is in the roundness.

I was thinking that as the part will be located onto the diameter of the fixture it will be essentially be one body, and the heating and cooling rates would in fact be the same, I may be wrong with this.

Another suggestion was to use mild steel plate and pins to locate the part(as there are some holes in it). Once the pins have rounded the part up, as its flexible, clamp it down and put in the oven at 1000°. I personally dont think that this would be a good solution due to the different rates of expansion etc. I am not sure what result you would get. Also this is a very high temp for mild steel( well into the austenitising phase).