Stress Relief Article

[swall]

Daniel Herring, aka "The Heat Treat Doctor" had an article in the latest issue of Industrial Heating about stress relief. One of his points was that for a stress relief heat treatment to be successful, the parts must not be rapidly cooled. "Rapid cooling will only serve to re-introduce stess, and this is the most common mistake made in stress-relief operations." Having never heard anything like this, I am skeptical. I am of the impression that once the part is at a temperature sufficient to reduce the material's yield strength, stress relief happens via a creep mechanism and subsequent cooling rate has no effect on the stress outcome. What do you guys think?

 

[metman]

Rapid cooling will undoubtedly induce new residual stress.

 

[metengr]

Based on data I have seen and my years of involvement with various Codes and Standards, for a subcritical stress relief once the metal temperature is below 800 deg F, cooling rate can be increased from a specified cooling rate.

Basically, the concern is to avoid harmful thermal gradients to the component. The 800 deg F limitation has been established by experience and means that once the part falls below 800 deg F, one can cool the part in air or moving without a specified cooling rate.

Would I use a quenching medium to cool the part from 799 deg F to ambient temperature, no, because you still have thermal gradients upon cooling. Part geometry will influence how quickly one can cool from 800 deg F without introducing harmful thermal gradients.

 

[CoryPad]

While creep may be operative, thermal expansion and contraction are present and plastic deformation via dislocation motion may be present too. The rapid cooling, leading to thermal gradients, can result in residual stresses. Certainly these stresses are smaller than those induced by cold work or phase changes, but they may be significant to the product.

 

[Guest102023]

Rapid cooling will introduce stresses, but not as highly localized as the initial stresses concentrated at weld joints. The new stresses will be more a function of the self-restraint of the component. This could still be deleterious under the wrong set of circumstances.

 

[emxhps06]

Please we could consult the TTT diagram for the specific material of interest?

 

[swall]

TTT diagram describes cooling rates vs transformation microstructure. It would be of no use in steels stress relieved at temperatures below Ac1.