304/304L went through PWHT. Question about sensitization 1

[timsch]

Due to several mistakes, I have some components made from 304/304L (welded with ER308L) went through PWHT at 1325F for 2 hours, cooled slowly to 800F and then air cooled. I understand that at this temperature, austinitic stainless steel can become sensitized. From what I've read, the amount of sensitization is somewhat dependent on the amount of carbon in the material, with higher carbon content resulting in higher sensitization. With this dual grade material, the maximum carbon content should be 0.030%, limited by the 304L.

How much of a concern would you have with sensitization of these parts?

 

[EdStainless]

There are two other factors.
How clean was the material to start with? If there were still traces of segregation (doesn't even take carbides) then the material will resensitize more rapidly.
Likewise if some sensitization had started in the welding (HAZ) then this would progress immediately without an incubation period.
The only way to know is to take samples of the original materials, use the same weld procedure and PWHT.
Then cut them up and look at the microstructure and do some corrosion tests.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[timsch]

Thanks Ed.

I'm not sure if I understand "clean" as you intend it. We are making thermowells, and the stem is turned down from bar and then TIG welded to a flange. Any machining oils will have been removed. For us, clean means the weld won't be contaminated. We don't perform material analysis beforehand to check for traces of segregation.

Regarding the HAZ, how can sensitization be avoided even without PWHT? With metal being melted, there must be areas that see 500-800C every time.

 

[EdStainless]

What I was talking about is a clean microstructure.
I have seen material that was sensitized in production and then barely annealed.
It was enough to dissolve the carbides but not even out the Cr and C in the chemistry.
This material resesitizes faster.
If the material is well annealed and your weld has well controlled heat input you should be fine.
This is where the standard curves come into play.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[EdStainless]

You could recover existing parts by solution annealing them.
1950-2000F in a protective atmosphere (usually hydrogen, but in your case Ar might be acceptable).
The hold time will depend on thickness, but I would be surprised if you needed longer than 20min.
Then rapidly cool (Vac furnace forced gas cooling is fine).
If you have a way to remove surface oxidation you could anneal in air and water or fan cool them.
But this would require acid pickling for cleaning.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[timsch]

Thank you very much for your assistance, Ed, particularly regarding recovery.

An example flanged thermowell has a 1" OD stem, 15" long from the face of the flange. Should I expect noticeable deformation from the solution annealing? I can arrange for them to be annealed with the stems facing down.

 

You could recover existing parts by solution annealing them.

It's the only solution (sorry) that I could have high confidence in.

Don't forget to requalify the welding procedure with this heat treatment.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

If you used an open rack with the flanges supported and the stems hanging the only source of distortion will be uneven cooling.
I wouldn't go crazy with the cooling rate.
Getting them under 800F in 120 sec is plenty good.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed