2507 super duplex pipe not passing G48A 3

[OZWelder]

Hello all,

We have been trying to complete a PQR for super duplex pipe. As per client specification this PQR must pass G48-A 40degC. We have passed all other testing requirements, except for the G48.

Pipe: ASTM A790 USN32750
Size: 200NB sch 40 and 200NB sch 20
Position 5G vertical up
backing gas: Ar + 2%N
shielding gas: 100% Ar
process: GTAW
Filler wire: 2.4mm Er2594

We have passed hardness testing, macro and ferrite count good, no sigma phases some surface level martensite detected along with precipitates.

My thoughts from hear are the following:
1. change baking(purge) gas to 100% N
2. change brand of wire from Exton to Bohler Welding – Avesta 2507/P100 GTAW as this is batch tested with G48 testing

could some please help in regards to the Formier gas ( I am not sure if I can get it in my contrary) but if I can is 95%N+5%H better for super duplex will it help much with reducing the heat tint coloring, is there a problem with hydrogen cracking using the gas? Should I use these gas for backing purge?

Thanks very much.

 

[SJones]

2 questions:

1. What are the test pieces failing on: pitting, or weight loss, or both?
2. What are the requirements for test piece preparation: completely as welded or pickled?

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/in/drstevejones

All answers are personal opinions only and are in no way connected with any employer.

 

[OZWelder]

Thank you very much for your reply Sjones,

The specimens have all failed on weight loss

the acc criteria is, no pitting at 20X magnification and less then 4g/m2 weight loss

 

[SJones]

……….and the answer to question 2:………

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/in/drstevejones

All answers are personal opinions only and are in no way connected with any employer.

 

[OZWelder]

Sorry, pickled.

I do not know what the level to which they are polishing the sample face, I would have to ask this.

 

[EdStainless]

So did you do A923 testing?
Sorry to inform your customer but G48A is crap.
the reproducability is very poor, as in different labs will get +/-10C commonly.
Did you pickle with Nitric/HF?
If so you can't fail a G48 because you already ate out all of the 'weak' material.
This is why A923 is done on freshly ground surfaces.
Don't go full N2, you will not have enough ferrite.
I have run Ar+4%H2 for both baking and weld gas.
It is always worth trying different filler.
I have seen it matter in some cases and not in others.


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

 

[OZWelder]

Thank you for your reply Edstainless,

I am looking at proposing a new WPS to the client. Where I will change the filler wire as stated above to the Boholer brand, due to the fact this wire specifications state the manufacturer has tested with G48.

I would like to try a hydrogen gas mix as I believe the hydrogen has a high affinity to oxygen and may therefore reduce colouring on the root side of the pipe. I am not sure if the client will accept this as I suspect them to bring up hydrogen induced cracking… and these gas mixed do not seam readily available in my state.

I will also be asking if they will accept A923 instead of G48.

This spooling is for a sea water pumping skid for offshore fire water system.

I have run PQR’s in the past on duplex UNS31803 and not had any issues with G48. Running Ar+2%N on backing and shielding. Heat inputs were definitely in the upper limits of around 1.5-1.7Kj/mm for root and 0.7-1.0 fill and capping, still passed G48
With the DSS.

Move onto the super duplex and all testing passes but can not get the G48 over the line.

 

[OZWelder]

thank you very much for all you help, it it greatly appreciate.

The outcome of this PQR was as followed:

1. I added an extra purge line to the coupon totaling 3 lines each running at 20Lpm.

this was to have a good volume of gas with out having excessive pressure / to high a flow rate from less hoses. Less flow per hose but more Lpm​

2. Ar+2%N for purging and changed from 100%Ar for shielding to Ar+2%N for shielding

3. Welded the coupons my self getting better heat inputs root= 0.92Kj/mm cold pass=0.65Kj/mm fill/capping=0.8Kj/mm

All testing including G48 were compliant, low heat input section had a loss of 0.2g/mm2 high heat input are had loss of 0.1g/mm2

thanks all once again

 

[EdStainless]

Keeping the heat up without getting excessive helps a lot.
Glad to hear that it is working for you.
Thanks for letting us know.

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