B31.3 table A-1, note (30) - 321H heat treatment

[XL83NL]

We're designing a piping system in 321H. Design temp. of the piping system is in the range 650-700°C. Wall thicknesses are less than 10 mm. Pipe spools are not PWHT-ed.
B31.3 allows this material for use above 538 °C only when note 30 is applied (1095°C min. heat treatment).

Sourcing of A403 fittings that meet specs turns out quite difficult. We have found a batch that was heat-treated at 1050 °C. These fittings, therefore, require re-solution annealing to 1095°C to meet note 30. Upon review of A403, there's a warning under 6.2 that a solution anneal above 1065°C may damage the resistance to IC for 321H (after subsequent exposure to sensitizing conditions). Anyone who was practical experience on this, and/or recommendations on procedures to follow? Supplementary requirements S2 would be one.

PS: A403 fittings, for the range we require, are made of A312 TP321H. A312 requires 1100°C heat treatment for 321H cold fin., as does A182 for F321H. Where's the catch, why aren't they bothered with that issue?


Huub
- You never get what you expect, you only get what you inspect.

 

[EdStainless]

The same issue applies to all 321 (H or not and also to 347).
If you anneal hotter you will at least partially dissolve the stabilizing carbides.
If this material is then used in the lower end of the sensitizing range it is likely that carbides will re-form along grain boundaries and leave your susceptible to IGA.
However, if the material is used in the upper portion of the sensitizing range, then there is little worry.
When carbides re-form at higher temps, even if they are along grain boundaries the Cr can diffuse well enough to minimize the local Cr depletion that leads to IGA.

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

 

[XL83NL]

Its used in the 500-600 deg C range (operating) so that’s quite right in the middle of the sensitisation range, if I read literature correctly.
So, the hotter you anneal, the more aggressive the IGC is. As a remedy, you need a thermal stabilising treatment then at say 870 deg C? Would that then be after welding?

PS: would such a re-solution anneal require a re-certification acc. A403?
And, how detrimental would it be (from a mechanical perspective that is, i.e. Creep) if we accept the fittings that were heat treated at 1050 deg C, thus don’t meeting note 30? Somehow feels like you have to choose either of the bad scenarios.

Huub
- You never get what you expect, you only get what you inspect.

 

[EdStainless]

Hold on, what is the use temp? In the original post you said 650-700C, or is it 500-600C?
At the lower temp range, I would use material that had been annealed at lower temps.
This material will still have scattered carbides in the structure and when it is later heated these carbides will grow first rather than ones along the grain boundaries.
If I were going to stabilize the material I would do it before any welding is done.
How serious your IGA risk is is a factor of both the anneal temperature (how much C is there in solution) and the exposure temperature (how fast is carbide growth vs Cr diffusion).
Or is the reason for the note the unstated desire for larger grain size for high temp use?


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

 

[XL83NL]

The reason for the note is unknown to me. Code requirement, trying to find the rationale behind it.
Design is 650-700, operating 500-600 range.
Thanks Ed.

Huub
- You never get what you expect, you only get what you inspect.

 

[XL83NL]

If I read your comment correctly Ed, would a hotter anneal per note (30) perhaps be related to "partially dissolving the stabilizing carbides", as you mentioned?
As such, is note (30) then be more of metallurgical nature, than from a mechanical/strength perspective?

Huub
- You never get what you expect, you only get what you inspect.

 

[weldstan]

The higher annealing temps were a result of a number of premature failures of 321/321H in the early 1970s. I believe in Creep. Unfortunately, I do not recall the actual metallurgical reasoning but did sit in on meetings with one of the prominent stainless steel metallurgists in about 1975 but it did involve fine carbide dispersion. When it was introduced in the ASTM specs and B31.3 Code, I do not recall.

 

[XL83NL]

Thanks Stan. I think we can close this one.

Huub
- You never get what you expect, you only get what you inspect.

 

[XL83NL]

Just found this parallel in ASME BPV IID, where the catch is the following. When using 321, incl additional requirements from IID, like the increased carbon content, the allowable stress is lower vs 321H, but the annealing temperature requirement isn’t there anymore.
Does that make sense from a material point of view? Is it a logical trade off?

Huub
- You never get what you expect, you only get what you inspect.