duplex PQR failed tensiles

[adaz27]

I've ran 2 PQR tests on duplex stainless steels.
1 X PQR on 22%Cr duplex UNS S31803 - 76mm OD X 19.1mm THK machined from forged bar
1 X PQR on 25%Cr super duplex UNS S32760 - 76mm OD X 19.1mm THK machined from forged bar
GTAW, 98Ar 2N shield & purge

Both PQRs have failed on tensiles in the parent material at values below the min req in ASME IX. 22Cr - 559MPa. 25Cr - 599MPa
Done tensiles on the parent material and both results are well above the min requirement. 22Cr - 772MPa. 25Cr - 844MPa
impacts, hardness, macro, bends all acceptable on PQR

Has anyone experienced loss of tensile strength in the parent due to weldng of duplex? Or can offer any advise on an explanation for this?

 

[metengr]

So you performed a Section IX WPS qualification and failed the tensile test and also ran only tensile on the base material and passed. First, I would check the math on the lab results because on occasion the area for determination of tensile strength was calculated incorrectly.

If the math is correct, you have a more serious problem with the duplex unrelated to welding. If embrittlement was issue from welding, the tensile properties would have increased and not decreased. I would send the failed tensile samples to a metallurgical lab for examination of the microstructure.

 

[EdStainless]

You need to look at composition and microstructure of the welds.
What is the ferrite/austenite ratio?

If the failures are in the base metal, and the hardness looks right, then someone made a mistake in testing.

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

 

[adaz27]

The calculations on the report check out OK.
There is no evidence of embrittlement in the weld, HAZ or Parent as the hardness values are well below the max requirement. 22Cr: 238-302 HV10 / 25Cr: 258-309 HV10

The material for the 22Cr & 25Cr was sourced from different mills so the chance of getting the exact same issues with tensiles for both materials is unlikely.
We have qualified previous duplex PQRs using the exact same processes and variables and never had this issue before.

I have been hesitant about sending the samples for micro examination due to the large cost and the samples already being deemed a failure due to the failed tensiles. What could the micro prove?

 

[EdStainless]

If the phase balance is way off then the heat input or weld speed was wrong.
How are you going to fix this so that you pass next time if you don't look at these?

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

 

[metengr]

I have been hesitant about sending the samples for micro examination due to the large cost and the samples already being deemed a failure due to the failed tensiles. What could the micro prove?

Large cost? What could the micro prove? Sounds fishy to me. You either do not care to get to the root cause of the problem or simply are in over your head.

 

[adaz27]

I agree that examination of the microstructure will shed some light on whats went wrong. I've sent the test pieces to the lab, waiting on the results.

metengr, apologies for the badly worded question. I did not mean what could the micro prove in terms of the microstructure but what possible root causes could it point towards. Each run was monitored on our arc logger so I'm confident that heat input was not the issue here.