I'm stumped with Super Duplex!!!! 3

[iam42]

Hi Guys,

I have just tried to qualify a Super Duplex test using 3/4" Sch 160 pipe and it is failing on the corrosion test!

The test was welded in the 2G position with the following parameters:

Wire ESAB ER2594
Root 75A, 11V, 1.8 in.min TS ((27,500 J.in)
2nd Pass 95A, 11.6V, 2.9 in/min (22,541 J/in)
Fill & Cap 95A, 11.8V, 3.4 in/min (20,018 J/in

Shielding Gas 100% Ar
Backing Gas Ar + 2.5% N2 (Maintained at < 50ppm O2)


Testing carried out was as follow:

Tensile Test 122,000 psi & 127,000 psi
Bends Acceptable
Micro Exam - No continuous precipitates in the grain boundries.
Ferrite Count - Weld Cap 39.9%, Weld Root 32.7%
The microstructure was examined for intermetallic phases, nitrides and carbides. None were observed.
The microstructure was observed for sigma phase. No evidence of sigma phase observed.
Corrosion Test G48 (40°C for 24hrs) - No Pitting
Weight Loss 12.7g/m2

As you can see the test for pitting was good but the weight loss was over the required 4g/m2

I'm really not sure where to go from here!

Any advice on changes that I could make to improve the weight loss would be greatly appreciated.

Thanks

 

[EdStainless]

Are you doing a G48 or a A923?
The difference is that an A923 uses the acidified solution and requires that all surfaces be freshly ground. This isn't a corrosion test but an exam for detrimental phases. (and 40C sounds high for this). With sch160 I would look at using impacts to qualify this instead.
A G48 may or may not use the acidified solution, it is a 24hr test, and you use the as produced surfaces (and cleaned as they will be in production, even if that includes pickling) with the cut faces usually being highly polished. This is a pitting resistance test.

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

 

[kingnero]

How did you monitor interpass temperature?

 

I don't see much wrong with what you've done here. The hot pass has an appropriate reduction in heat input relative to the root pass.
The only thing that stands out is the very small diameter pipe. The ergonomics of that tends to make the energy-inefficient manual GTAW process even less efficient.
Corrosion tests on duplex SS don't tend to yield borderline results; they either pass comfortably or fail badly.
I'm a little surprised you don't see anything in the microstructure. Have you done Charpy testing?
Is Inconel filler metal permitted?

IMOA, Practical Guidelines for the Welding of Duplex Stainless Steels, 3rd Ed., 2014


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

 

[iam42]

Ed,

We are doing a G48 Method A test per the customers specification. It is also pickled per the customer requirements. The customer also stated that impacts were not required on this material thickness.

Kingnero,

The interpass was monitored using a digital touch pyrometer.

Thanks

 

[iam42]

Ironic,

Charpy testing was not a requirement. I would have much preferred to do charpy testing!

I have actually submitted the results to our customer for review. If they come back unacceptable then my next approach would be to talk to them about using an Inconel filler.

Its a crazy situation where we are fabricating a 12" Sch 160 pipe spool with two 3/4" branches. The qualification for the 12" Sch 160 weld was great with no issues with the corrosion test. The customer wanted a separate qualification for the small bore and I am now starting to lose sleep over it!!

 

[Mr168]

What was your interpass temperature?

 

The overall thermal cycle seems to be the only unanswered question here. It is especially important for a small diameter pipe that represents a very small heat sink. Where duplex SS is concerned, this small pipe is a world away from the 12" pipe.

It's not simply peak interpass temperature, it is time at temperature that matters. So I echo Mr168's question.

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

 

[EdStainless]

Are you sure that they are testing with non-acidified FeCl?
Is the temp control +/-0.5C or tighter?
Have you polished the cut faces? I would wet polish through 800 grit or better.
Can you see pitting on the ID surface?
If you can see attack other places then we can talk about how to handle that.

I am guessing that this is staying too hot during welding or taking too long to cool down.

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

 

[CWEng]

It's been a while since I've had to pass one of those corrosion tests, but if I'm remebering correctly, that seems like a lot of weight loss to not have any pitting. Probably review your lab's testing procedure/experince. See if another experience lab comes up with the same results. If the surface finish was bad, it can lead to accelerated weight loss and mask pitting. So it's possible you had pitting that wasn't seen due to surface conditions.

Also, on the filler metal, what was the PREN?

Heat input/time at temperature could be your issue, but I'd look into these two items as well.

 

[iam42]

Thanks for all the input guys.

As far as Heat Input is concerned, I maintained a Max Interpass temp of 160F and I used forced cooling between passes so it was not at high temps for very long.
The details of the test are below:

 

[EdStainless]

So if this is a pitting test the only surfaces that should be preped are the cut faces, and 600 isn't fine enough.
The sample should be full thickness with OD and ID intact.
If you section out a piece for testing you are exposing a lot of internal material, and you have a lot of metal being tested that won't be exposed in service.

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

 

[iam42]

Thanks Ed,

That is something I cannot get my head around when it comes to this test.

I may be missing something but why are all surfaces exposed to the acid when it is only the pipe ID that will see any exposure to the service fluid?

Very confusing to me!

 

[EdStainless]

It would be preferred to test a full section of pipe with the weld at one end, but often these get too large to handle.
You do want to minimize the cut surfaces exposed.
I would never split the thickness for example.


This may be a dumb question, but are you sure that a section of the pipe with no welds will pass this test?

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

 

[iam42]

Ed,

I don’t think it’s a dumb question. I did not test the pipe alone but did ensure that the pipe was tested at the mill and the results were shown on the MTR. I don’t have it in front of me but I think the weight loss was 0.006g/m2 (50C for 24hrs)

 

[EdStainless]

We always re-test the base materials because it lets us see testing differences.
The G48 tests are not very precise at all.

We don't use pract A, we use C or E.
We like the acidified solution because it is more stable and repeatable from batch to batch.
Of course it is more aggressive so the temps need to be a little lower.
And I am a bit surprised that with a superduplex you aren't using the 72hr test.

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

 

[iam42]

Ed,

If I was to propose a Nickel based consumable I am pretty sure ferrite testing would not be required as the weld would be fully austenitic.

Would there be any need to check for intermetallic phases, nitrides and carbides or sigma phase. I’m also not sure if checking grain boundaries for continuous precipitates would be relevant.

Any insight on these tests would be appreciated.

Thanks

 

[EdStainless]

Even with a Ni filler GB phases and detrimental phases are still a concern.
I don't think that you need to go that route.
One thing that I would do is run the G48 at 35C and see if you have zero wt loss.
I suspect that it a minor issue that is giving you problems.

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