Knife-line Attack prevention when welding 321 SS 1

[metaleng123]

Dear Gents,

I've been trying to understand the welding metallurgy of stainless steels. I'm working mainly with 321 material(347Si filler) and one particular problem with this can be KLA. I've understood that during re-heats of an existing joint the Niobium carbide / Titanium carbide can dissolve and during cooling chromium carbide will be formed instead, near the fusion line.

In my research I found that this can be mitigated by use of Ni based filler or by limitation of heat input ( max. 1.5 Kj/mm), but i haven't found anything detailed that describes how to prevent this type of failure. Can anyone point me in the right dierction? Is the limitation of heat input enough?

 

[mrfailure]

An aside: Why are you assuming only "Gents" are involved in this forum? I really hope women are an active part of our community as well as a large number make up the metallurgical engineering population.

 

[MFAMET]

Inconel alloy 112 or others is an optional metal for 321 SS welding filler due to its less galvanic corrosion and is metallurgical compatible with SS.
KLA is caused by high heat which dissolves both Ti and Cr carbides. If you re-heat the welded material at 1850-1950F, this will allow Ti to pick up most of carbons again to form TiC and leave along the dissolved Cr carbides. This ends up most carbons are caught by Ti instead of Cr. This will alleviate "sensitization" along the grain boundaries and intergranular corrosion.

 

[EdStainless]

How thick of material are you welding?
Is it being stabilized after welding?
What is the service temperature?

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

 

[isoca]

Stabilizing phases are formed at higher temperature than chromium carbide. If you reach the dissolution temperature of the stabilized carbides and then cool to a rate that does not allow their formation, when you heat up again the first thing that will form is chromium carbide (in the case that you hold or slowly pass through its formation temperature range).

 

[EdStainless]

I have seen this in cases involving thicker material and where it was presumed that it would stabilize in service.
However, if the service temp is near the lower end of the temperature range for stabilization, then you get competition between the various carbides forming. And when Cr carbides form the temp is too low for Cr diffusion to even out so you are left with Cr depleted regions along the carbides.
The lower temp will also favor carbide formation in grain boundaries.

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

 

[metaleng123]

Thank you for all the answers. Your input in very appreciated. I think i have a better understanding now of what happens when you re-heat a stabilised SS.

@EdStainless. We are welding thicknesses ranging from 3 mm to 25mm( design temps. ranging from 490C to 660C), most of the products are not pressure retaining. My concern is with pressure containing welds. For example, a 4 mm thk. 321 rolled plate butt welded in a steam vent silencer. Design temp. is 490C. These longitudial seams are welded either by GTAW or GMAW-Pulsed with heat inputs usually bellow 1 Kj/mm and interpass temp. 150C.

 

[MFAMET]

As a stabilized steel, such as 321SS, a low heat input could avoid the dissolution of TiC (1900F or above) which is designed to tie up free carbons. However, if the input heat is too low, the weld quality could be compromised, such as the reduction of toughness with others

 

[EdStainless]

stabilization should be done at 850-875C.
If the service temp is below this then the parts should be furnace stabilized prior to service if they will ever see temps above 425C.

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

 

[MFAMET]

The starting dissolution temperatures of TiC and CrC are 1900F and 1450F, respectively. Therefore, if the welding temperature could be maintained at 1550F-1750F, this will prevent both of the TiC dissolution and the formation of CrC. This result will greatly decrease the deletion of Cr along the adjacence of grain boundaries. A maximum intergranular corrosion resistance could be obtained.

 

[EdStainless]

MF, this isn't possible.
When welding the pool is molten and the HAZ will be well above 1900F.
All that you can do is minimize the time that it is hot.
It isn't the Cr carbide formation itself that leads to this, it is the local Cr depletion around those carbides.
I have sensitized 304H at a high temp and it does not fail sensitization tests as the Cr could diffuse enough to heal the areas adjacent to the carbides.

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

 

[MFAMET]

I should make it more clear for my previous common.
KLF is a very thin film perhaps, a few mms wide. It lies at the very vicinity of the molten weld.
For some cooling regimes, TiC (or NbC) precipitates are dissolved, but instead, CrC precipitates are formed. The steel becomes unstsbilized in these areas.. That will generate the risk of sensibility along boundaries.
KLA could be fixed with stabilization anneal at 1550F-1850F for hours. The cooling rate won't be critical. Some article also states low heat input at final weld pass could help reduce this risk.
.