Chromium carbide of 321 stabilized stainless steel

[kowngs]

I've had a exprimental test for 321 stainless steel plate and welds if chromium carbides of 321 stainless steel can be removed by solution annealing.
I chose specimens that thickness is 52t plates and welded as per ASME code.
And the specimens were exposed 450 deg for 4 hours to create chromium carbides.
After exposing sensitized temperature (450 deg), microstructure showed chromium carbides in grain boundary.
Then, the specimens were solution heat treated in 1,060 deg. for 2 hours (according to ASM recommandation).

As I know from ASM code, heat treamtnet above 1,000 deg. will removes chromium carbides.
But, result shows that chromium carbides in welds are not removed.
Yes, chromium carbide of base metal and some of chromium carbides in welds are gone, but lots of them in welds is still not removed. Finally, it fails A262 test practice E.

Here is my question comes up. If 321 stainless steel with welds are sensitized (with chromium carbides), chromium carbides can be removed by solution annealing? Is there other ways to remove?
Or just it can not be removed?

 

[moltenmetal]

I thought the whole idea of 321 was that the titanium in the alloy would bind the carbon as titanium carbide rather than depleting chromium as chromium carbide?

 

[metengr]

What was the filler metal and welding process used for the 321 ss weld coupons? Weld deposits can have microsegragation which results in the features you have seen regarding incomplete carbide dissolution for solution annealing temperatures stated for wrought materials.

 

[EdStainless]

I suspect that you are seeing Ti C N compounds and not Cr carbides.
Most 321 that I have seen is recent years has ultra low C levels, like <0.015%. What is your C level?
The other thing is that the carbides themselves do not hurt the material, it is the associated Cr depletion around them that is bad.
If you sensitize at 850-900C you will form secondary phases, but that is hot enough to allow Cr diffusion and prevent any Cr depleted zones from forming.

I find it very hard to believe that your 1050C reanneal samples failed 262E.
You may have actually redissolved some of the stabilization and re-sensitized the material.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[kowngs]

Dear metengr

Thank you for your reply.
Composition of base material is C=0.047%, Cr=17.29%, Ti=0.358%.
Weld material is AWS A5.4 E347.16, and C composition of weld material is 0.06%.

So you think solution temperature is not appropriate?

 

[kowngs]

Dear EdStainless

Thank you for your reply.
As I posted for Metengr, Composition of base material is C=0.047%, Cr=17.29%, Ti=0.358%.
Weld material is AWS A5.4 E347.16, and C composition of weld material is 0.06%.
And sensitized temperature is 450C, not 850-900C.
I'm sure what I saw in base material was chromium carbide. But I'm not sure it was in welds material.
What I saw in welds material was empty round shape and there was lots of them.

Also, the specimens were heat treated in 1060C for 2 hours. And it fails A262E. These are fact.
I really want to know why. At least, I want to know chromium carbide can be removed.

I don't understand the one of your reply "You may have actually redissolved some of the stabilization and re-sensitized the material."

Thank you for in advance.

 

[kowngs]

Dear metengr

Sorry, I didn't say what was welding method. Welding was SMAW.

 

[EdStainless]

The exposure at 450C should have created both CrC and TiCN in the structure.
I don't know why you are sure that it is CrC, because you must have some Ti compounds in the structure.
At 1060C you will dissolve all CrC, but not all of the TiCN.
Then if you don't cool fast enough you will re-form CrC upon cooling.
If you take it from the furnace at 1060C and put in another on at 875C and hold for a few hours you can then slow cool. There will be no CrC, only TiCN.
That is how the alloy is designed.

If your service temp is over 350C but under 800C you really need to give it a stabilizing treatment to force the Ti to soak up all of the C and N.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[kowngs]

Dear EdStainless & Metengr

Sorry for cross post it, but I don't know hot to delet it.

Actually, I don't have much exprience in microstructure area.
If you give me your email address, I would like to show you microstructres of my test.
I know it is really hard to tell which one is CrC or TiCN.
But, I really hope you share your opinions on my test results.

Please share your wisdom.

 

[kowngs]

Dear EdStainless & Metengr

And one more thing, base metal plate was ok. But, weld metal was problem.
Weld metal was fail to A262E.

 

[weldstan]

Is weld metal ER/E347?

 

[EdStainless]

Did the weld metal fail or the HAZ?
Why the 450C exposure before re-annealing? You may have grown secondary phases so large (in the non-homogeneous weld metal) that you could not fix it with your anneal.
Two options; either go from weld directly to anneal; or first use a high temp homogenization step (1200C) then cool, then anneal.

You can link to your micros here by posting to engineering.com, see the box below the one that you type your reply in.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Martinos]

Dear all,
from my experiences it is relatively easy to recognize TiCN contra chromium rich M23C6 carbide in titanium stabilised SS. Optical microscopy with suitable magnification is good enough to do so. TiCN should be angular, cubic with relatively sharped edges.

If it is possible, please, link here the photomicrograph.

What about the weld metal microstructure, I suppose there must be a lot of delta ferrite which is typical for austenitic SS weld metal (and as-cast state most of austenitic SS). Carbide recognition in weld metal can be influenced with the presence of delta ferrite.

 

[weldstan]

347 weld metal does not respond the same as 321 to solution treating parameters tailored to 321. Also see Martinos re delta ferrite.

 

[EdStainless]

In these alloys you never want to fully solution anneal and remove all of the Ti (or NB) carbo-nitrides. That is when you get into trouble with them reforming along grain boundaries.
Even when fully annealed there should be a nice uniform distribution of small carbo-nitrides throughout the structure.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[kowngs]

Dear all

I really appriciate your replies. Thank you for that.
And I attach microstucture solution annealing(1060C) for 2 hours after exposing 450C with A262E test picture in ZIP file.


Thank you in advance.

 

[Martinos]

Yes, what about the weld metal some delta ferrite isles can be seen. For me it seems to be like: austenitic matrix, delta ferrite and TiCN particles.

What about carbides the magnification and photomicrograph resultion is not enough to give you 100% clear statement but I can see some particles with sharped edges which reminds TiCN, but as I said, the resolution is not enough to recognize all these particles for sure.

 

[EdStainless]

The residual delta ferrite in the welds is very wrong.
If these parts really saw 2 hours at 1060C then there should be none.
You need to look at the as welded structure and do some ferrite measurements.

And I am not sure that he A262 is a fail. I see a lot of texture but I don't see fissures.

Remember, when you anneal this alloy you are removing the CrC (shorthand) and dFerrite, and not the TiCN.
If you anneal so hot and long that you remove everything then you need to go back and deliberately form the Ti rich particles again or it wont be stabilized.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[kowngs]

Dear Martinos & EdStainless

Thank you for your reply.

I attach microstructure of other fail case (same condition as before, but solution annealing is for 3hours).
And welds is not 321, but it is 347.
There can not be TiCN as you said.
I think round shape particles in welds are reason for fail A262E.
The crack in A262E was in welds, not base metal.
I guess 347 welds is quite difficult to heat treat.

Again, the problem is not base metal(321), it is welds metal(347).
TiCN must not be seen in welds metal.

And annealing temperature 1060C is not too hot temperature, I think.

Please share your opinion on this one.

 

[Martinos]

After I saw the last pictures carefully I have to say I am confused.

500x magnification showing a lot of globular particles, but this particles are quite coarse regarding guaranteed low carbon content. Ok, you have carbide creating elements Cr+Nb(Ta) in 347 but I think there is a lack of carbon to reach coarse carbides like these (and so much charbides). Even after long-time exposure on high temperatures are the carbides usually finer.

EDS analysis can be helpful. Nitrogen and carbon content analysis expecially.

Maybe Ed will have some explanations, but after correct solution annealing of weld metal 347 in my opinion you can't reach carbides like these.