Intermittent Fillet Welds As Rigid Links

Don't know exactly what you are doing, but stitch welding is done a lot to join members together and is fine for the proper application. If you are trying to design a moment connection, then it doesn't sound like "intermittent fillet welds" is going to qualify for that, so you shouldn't model it as a rigid connection.

It is not a moment connection. The application is a outer frame comprised of a 1/2" X 3" plate and the inner frame comprised of a 1/4"x1" plate. These two plates are laid flat against each other and welded together using fillet welds every 12". I modeled the outer frame connecting to the inner frame using rigid links where the fillet welds occur. I could also have analyzed the outer frame and inner frame as a built-up member and then designed the welds based on the shear flow.

Val Courtney, PE
Valstone Engineering, Inc.

Hard to imagine what's going on from your description.

Let me guess that you have modelled two plates at thier centerline, and you want to connect, say, every fourth node from plate to plate. I think you could use a rigid link for this, but I would pin on end of the link so you don't transfer rotation (bending moment) from one plate to the other.

I don't understand why you are trying to model a weld in RISA. If you know the load between the plates, just design it for the load. If you are trying to make the sections act compositely, I think do it as a built up member like you said.

The issue I have is that the fabricator is required to weld the back side of the plates only. See diagram below:

------
| |\ stitch fillet weld this side only
------ \
---------------------------
| |
| |
---------------------------

This structure is an exterior vertical trellis at a mall entrance in Palm Beach, FL which will be exposed to hurricane wind forces. You would get the trellis' perimeter outline if you were to draw a rectangle and then replace the bottom portion of the rectangle with an arc. My concern is getting the 1/2" x 3" outer frame plate and the 1/4" x 1" inner frame plate to act together using a one-sided fillet weld. The vertical trellis is only supported continuously at the top and sides.


Val Courtney, PE
Valstone Engineering, Inc.

I would be more worried about cracking where you have a fillet weld on just one side.

I think providing rigid links could be a way to model the two members being connected as long as you can verify that your model is behaving as it should with another program or some hand calcs. The links may try to add unwanted stiffness or other effects if you're not careful with applying them and checking the behavior.

I went ahead and modeled the vertical trellis using a built up member and am satisfied with the deflection and so forth. My concern is the one sided fillet weld which the fabricator insists using because the members are stainless steel and providing a weld on each side will warp the members. My maximum shear in the built up section is 38.4 pounds and my maximum torsion is 214 lb-in. I know this doesn't sound like much, but I don't see how weak axis bending on a single line of weld satisfies recent codes and principles. For shear flow, my one sided weld will work, but for torsion I don't think it will. I have a feeling under a wind event the welds will crack. The fabricator tells me that they have taken sledge hammers against the mock up of this trellis and it doesn't budge. It's hard to argue by the book when they say from experience it is hellacious to remove stainless steeel apart from each other after it is welded using heliarc welding. Any thoughts would be appreciated.


Val Courtney, PE
Valstone Engineering, Inc.

From doing some welding myself, if you are able to induce bending perpendicular to a single linear weld, the weld breaks very easily (I'm not a certified welder, but I can run a bead!) If you only transferring direct shear between the two, it sounds promising. I think it depends on what forces you are trying to transfer.

If you stresses are high and you are tying to make a composite section, I think you need to have your connection balanced about the centroid as much as possible, which would require two lines of welds. Or, can you make one of the sections big enough to take the entire load, and then not worry about it?

Well I suggested to the fabricator to provide a weld on each side of the 1" plate. I think that is the way I am going to go. Thank you all for helping me on this discussion.

Val Courtney, PE
Valstone Engineering, Inc.