Failed Impact Tests

[iam42]

Guys,

I just had the test results from a PQR that I recently ran and the impact testing failed miserably.

The test coupon was 12" Sch 80, SA106 Gr.C pipe. The test temp was -46°C

The results were as follows:

Weld Metal - 100,121,27
Fusion Line - 14,20,13
FL +2mm - 11,8,10
FL +5 - 2,8,8

The coupon was welded with GTAW all the way using ER80S-Ni1.

The coupon was preheated to +70°F and the average Heat Input for the fill was 1300 j/mm (32,700 j/in)

I have a feeling that the preheat was too low and that I need to increase the heat input?

Any help with this would be greatly appreciated.

Thanks

 

[david339933]

I would definitely not increase preheat or heat input....for impacts you want both low. Interpass temp. should also be low. I would think the problem is the pipe, not the weld....looking at the values....assuming weld metal is 127 not 27.

Why not use SA-333 Gr. 6, which is designed for low temp service rather than one designed for high temp?

 

[iam42]

Hi David,

Thanks for the help. My problem with using A333 Gr.6 is that it is P1 Group 1 and the material being welded in production is A350 LF2 which is P1 Group 2. Since impact properties are required I do not believe I can use P1 Gr.1 material to qualify the P1 Gr.2 material.

Please correct me if I am wrong.

 

[david339933]

Correct....group number is essential in this case. What about using plate? I have often used dual certified plate (SA-516-65 P1G1 to SA-516-70 P1G2), which then qualifies P1G1-P1G1, P1G2-P1G2, P1G1-P1G2. Done in Vertical Up position will qualify all.
Should be noted that some construction codes require the PQR coupon to be the same heat-treated condition as the production material....Sec.VIII Div.1 - UG-84(h)(2)

 

[iam42]

Thanks again David for the confirmation.

I am contemplating using 516-70 and maybe even PWHT but if there is another option without the PWHT possible I would love to hear it.

I know impact testing can be quite erratic but i have never experienced such a bad test results in the past!

 

[Mr168]

They're bad, but consistent. You're cold enough that you'll need to be selective with your starting base material for acceptable values.

 

Mr168 is correct, the problem is the base metal. Your heat input is fine.
Start by looking at SA-106, which has no requirement for a normalizing heat treatment, always beneficial for low temperature impacts.
Second, impurities, particularly sulphur, have a deleterious effect on impact toughness. You want sulphur to be low, less than 0.010%.
°
For procedure qualification with low temperature impact testing, I will shop for a base metal MTR showing normalizing was done, has good impact test results, and low sulphur. In the case of 516-70 or 106C I will also look for carbon < 0.20%. But even with all of these good characteristics, at -46°C you are around the lower temperature limit for plain carbon steels as far as impact toughness is concerned. You can make a PQR pass, but not all steels you weld in production will have properties as good.
°
Note that 'HAZ', 'fusion line', and 'fusion line + 2mm' locations are all fairly meaningless unless (i) the fusion line is parallel to the thickness direction AND (ii) the notch is accurately situated by the testing lab. In reality these conditions almost never occur unless the coupon is deliberately so designed and great care taken. This is effectively a built-in loophole in ASME IX.
°
david's suggestion about testing P-1, Gr-1 to P-1, Gr-2 is a good one which I have used myself, welding 516-60 to 516-70. Some additional testing is required compared to a single qualification.


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

 

[iam42]

Hi ironic, thanks for taking the time to consider my problem. You have highlighted a few things I did not really consider.

I think I will try and source some a350 LF2 as that is what we will be using in production.

I am also going to PWHT the remainder of the material that was not used for the test and rerun the impacts. This is just for my own education. I am interested to see what improvement the PWHT will do for impact properties.

Thanks again.

 

[CWEng]

Do some unaffected base metal impact tests, and you can proove it's a base metal problem instead of a welding issue, which is probably the case. As suggested, normalized or quenched and tempered base metal is going to improve your impact values. Do you know the original material heat treat condition? However, depending on your code of construction, getting a specific heat treatment on the base metal means that's equivalent to an essential varaible on your PQR. ASME Section VIII Div 1 says that the weld coupon has to be in the same heat treat condition as the base metal to be used for production. Make sure you know the limitations of your code of construction.

To get group 1 and 2, we typically use dual certified SA-516 Gr. 60/70. Then you don't have to take impact tests from both sides of the joint.

 

[weldstan]

Your choice of 106C was doomed to fail. Purchase a material with impact testing at -46C with a minimum absorbed energy of 30 J.

 

[iam42]

Thanks Weldstan and CWEng, yes it certainly looks like it was doomed to fail from the start.

Thanks for all the help.

 

I am also going to PWHT the remainder of the material that was not used for the test and rerun the impacts.

That will not turn low toughness steel into acceptable toughness steel.

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

 

To get group 1 and 2, we typically use dual certified SA-516 Gr. 60/70. Then you don't have to take impact tests from both sides of the joint.

You're right, CWEng, that is how to run the PQR so that no extra testing is required.
Another way to add value on a 1½" thick plate PQR is to deposit ¾" each of two weld processes, which qualifies a full 8" deposit thickness for both processes (Ref. QW-451.1).

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

 

[david339933]

True....but then you are stuck with 5/8" as the min. thickness qualified.

 

True...unless I specify 'impacts not required' on the WPS.


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