Impact testing for all welds, no matter how small? 1

[JCrash]

Lets say I have a pressure vessel being built to ASME Section VIII Div 1. The main portion of the vessel is a rolled cylinder (SA-516 Gr 70) that requires impact testing due to an excessively large weld (>6" governing thickness). Therefore, per UCS-66 and UCS-67, i need impact testing on my weld procedures and i need to create production impact test plates.

However, there are also some small holes in the casing for assembly of some internals. These holes have plugs that are getting welded to the casing with small .5" bevel welds. This weld is small enough that the materials (plug and rolled ring) are exempted from impact testing. Since the rolled ring has the larger weld, it's still getting impact tested.

My question is - should impact testing be performed for the weld procedure that puts the plugs on? UCS-67 seems to say that yes, since the rolled ring requires impact tests, the procedure needs to be impact tested. However, the governing thickness of this weld is so small I find it hard to believe that this is required for this procedure.

 

[metengr]

JCrash;
The bevel welds are called fillet welds, and should fall under of a groove weld procedure qualification for the specific base materials, that would be impact tested.

 

[stanweld]

JCrash,
Note that ASME IX, QW-403.6 refers to base metal thickness - not deposited weld metal thickness.

 

[JCrash]

metengr - Thanks for the correction. Can you provide me a rough reference to where your information is from?

stanweld - I am not certain how your answer relates to my question. The governing thickness of the weld i am talking about is from UCS-66 of Section VIII div 1. Could you further clarify?

 

[stanweld]

ASME IX QW-403.6 governs the procedure qualification requirement with regard to thickness. Per the conditions you defined, a WPS impact test qualified on 1.5 inch wall minimum thickness will qualify the small welds described as well as the full penetration groove welds. You would not need to additionally qualify the procedure on 0.5" thick material.

 

[JCrash]

As a further clarification - the > 6" weld is performed by Vendor A, and the small groove weld is performed by Vendor B. Vendor B (which is performing assembly of some internals) doesn't currently have a procedure qualified with the impact testing. Unfortunately, this weld had to be performed after the assembly of some components and therefore could not be performed at Vendor A.

This is why i am investigating whether or not the impact testing would be required for just this particular very small weld. If so, Vendor B can qualify a procedure, albeit with some hassle. If not, we can proceed with current Section IX welding procedures.

 

[metengr]

JCrash;
If I understand your additional information, vendor B must have the welding procedure impact qualified if you are welding to the base material that requires impact testing because of MDMT and base metal (ring) thickness.

 

[JCrash]

metengr - Thank you for your clarification. We will have to proceed in this fashion.

Can you (or anyone else) shed any light as to the reasons why the code requires this? If I were to do the SAME small groove welds on the SAME material but I didn't have the large > 6" weld, I would not have to impact test any of this material. How does the larger weld force additional requirements on the small weld, regardless to its proximity to the small weld?

Is there any reason? Is this just a conservative requirement?

 

[metengr]

JCrash;
The reason is that any fusion welding that is performed on base materials, which are required to be impact tested for service, will effect mechanical properties, namely, notch toughness.

 

[stanweld]

JCrash,
As base material thickness increases, impact toughness properties are adversely affected. Even small welds may initiate cracks undiscovered by Code required examinations. Small welds in thick materials can be quite detrimental as witnessed by some recent hydrotest failures in China and Saudi Arabia.

UCS-66 and UCS-67 were installed in the Code rather recently to reflect the deterioration of toughness with base metal thickness and the effects of welding there upon.

 

[JCrash]

metengr and stanweld -

Thank you for your insight on these matters. If it's not clear, I'm not a weld engineer nor do i have experience with these particular code requirements...

I just would have thought that if small welds can initiate failures, they should require the impact tests in and of themselves. If all I had were small welds on this thick vessel, the code (whether right or wrong) would not impose impact test requirements on the materials or the weld procedures.

If there is truly a technical issue here, than thank goodness we made this from a rolled ring with the large welds and not a forging! And if next time we do make it out of a forging, assuming we stay even slightly below 6" thick, I'm no longer required to impact test. Would you still suggest I do, above and beyond code requirements?

 

[stanweld]

Please refer to UF-32 when welding is used for fabrication. The requirements of UW & UCS & UG-84 still apply. Pay close attention to UCS-67(a).