Analysis of an over-weld ??

[Structures33]

There are welding criteria that state the max thickness of a weld. I assume that if the weld is thicker than the material it is welding, this will alter the failure of the member. Is this correct?
Is there a way to analyze something that has been over-welded? I have a structure I'm analyzing and the truss web members have been welded with a filet that is thicker than they are.
Thanks in advance!

 

[nutte]

I don't think there are maximum fillet size requirements. At some point, the weld strength will have exceeded the member strength, but I don't think you've changed any failure mechanisms. Do the same checks you normally would.

 

[kelowna]

Maximum weld size requirements exist for fillet welds, If I remember right (somebody please correct me if I am mistake) for fillet welds on a right angle configuration it is 0.7 times the thickness of the thinner plate connected, lap plates, again, I think it is 0.7 again. It will also depend which code is used.

Maximum weld size is specified because of heat imput. Large welds relative to thin plates might just 'melt' the plate or distort it too much.
If you are analyzing an existing weld, and the parent material is OK, I do not see much of an issue.

 

[frv]

See section J.2.2.

The max fillet weld size for a plate > 1/4" is 1/16" less than the plate size.

The reason for this has more to do with ensuring that the proper weld size is achieved than on any strength considerations (beyond the obvious fact that a fillet weld extending beyond the edge of a plate is essentially welding air).

The reasoning behind this is that a welder welding a thick plate could conceivably melt away the top edge of the plate, making it appear as though there is more weld than is actually deposited. By essentially requiring that the edge of the plate be visible, the size of the weld can be accurately determined.

 

[nutte]

frv, the requirement you speak of refers to welds placed at the edge of a plate. Being a welded truss connection, I imagined a Tee connection, where you're not welding to the edge of the plate. Thus, no maximum weld size requirement exists.

 

[frv]

I guess I was visualizing a web angle coming into a bottom chord T.

But you're absolutely right. What I wrote above only applies to edge conditions.

 

[Structures33]

Nothing about the truss is "normal" so I won't begin to go into that. The weld is a filet along the edge of the members so a maximum weld requirement does apply. However, that's not what I was questioning. I know that the weld is too big for the members.
What I was after was more along the lines of what kelowna and frv commented on w/ how it can affect the members being welded.
It sounds like there really isn't a way to determine how this may affect the structure except to look for visible deformations that may have occurred.
Thanks to all of your for your input.

 

[HgTX]

Heat input determines *minimum* weld sizes. The 0.7 factor has to do with the maximum you can consider in your calculations, not the maximum size you can make the weld.

There are still weld profile considerations. Some codes address excess concavity/"reinforcement".

Hg

Eng-Tips policies: faq731-376

 

[conceng]

The American Welding Society Structural Welding Code AWS D1.1 2.3.2.9 has a maximum fillet weld size detail for lap joints. Frv has it correct. The max fillet weld size for a plate > 1/4" is 1/16" less than the plate size. For less than 1/4" plate it is the plate thickness. This is a design detail requirement, not construction.

The as-welded condition may encroach on the 1/16" provided the weld size is clearly verifiable. Meaning that one still needs a positive reference point in which to place the fillet gauge to determine the weld size.

The typical overwelding I see comes in the form of excess convexity, creating excess tension in the weld face and possible toe cracking.

Gregory A. Johnson, P.E.

http://www.midwesttestinglabs.com

 

[HgTX]

Whoops, I meant convexity. I know better than that.

Hg

Eng-Tips policies: faq731-376