Saturday, January 3, 2015

Fillet Welds on Corner Joints

Hello,
I have a question regarding D1.1 code.
My question is regarding the corner joints in the plates encompassing the columns. They are calling them Fillet welds, yet there is not any faying surface. D1.1 says a fillet can have up to 3/16ths misalignment (with certain stipulations) which is effectively is making the joint an open root CJP.
They are using a ceramic backing that (desired because of the tight fit around the square column plus the misalignment) is reducing the theoretical throat greatly. Now they are having substantial issues with cracking threw the throat, which isnt surprising.
My issue though is with the joint design; Im inclined to say it is not a fillet weld, but I cannot find a code reference to support that claim. And I cannot find a prequalified wps and joint config that in any way resembles this joint PJP or CJP groove.
Am I missing something in the code that resolves this? Specifically any denominational requirement for the length of faying surface on fillet welds? What course of action would you advise?
Thank you for any assistance,
- Caleb

Caleb,
I'm not sure of the plate thickness, it looks (from the photo) to be about 3/8”. If that's the case these 1/4" welds are undersized. That said:
This is a common Fillet weld on a Corner joint. If you have a copy of AWS D1.3 Fig 3.2a you'll see a picture of it.
Questions I would ask…
Does the shop/contractor have a WPS for welding this?  The fillet looks to have been done vertically down. This would require testing. What is the process used?  It looks to be GMAW-S. If so, this would require testing also.
I'm sure your cracking is due to insufficient throat. You can fix that by requiring multiple passes.
Here's where you should go with this…
Ask about the WPS and the process. If they are not in compliance, take out a rubber hose and beat them until they are (that’s a metaphor for fix that first 😊).
Ask the engineer to change the weld call out to:
Weld size = T (remember, T is thickness)
Require a Convex contour (that will insure the throat size)

That should do it.
PWC

1 comment:

Arthur Greene said...

Welding is a question of technology, but it is also a problem of mathematics. Glad you've laid it out as that, making the information you have disseminated something truly worth noting. Thanks for sharing such a good read, Paul! This will truly help a lot of people. All the best to you!

Arthur Greene @ Central MM