Shear tab welded to supporting member AND supported member

[Lion06]

Does anyone see an issue with welding a shear tab to both the supporting AND supported member? AISC only provides a procedure for bolting to the supported member. Is there a reason for that?

 

[ltlspf]

It makes an overly stiff pinned connection.

 

[nutte]

There is no mechanism to develop a pin with this connection. Bolted shear tabs have bolt holes to allow some rotation. Clip angles and end plates will flex when loaded. A welded-welded shear tab has none of these.

The fact that AISC is silent on this topic should also raise some red flags. I'd recommend avoiding this one like the plague.

 

[Lion06]

They don't rule out slip-critical bolts which is a little curious. That would make for an equally stiff pin, right? Especially if you design for slip at strength levels.

 

[Lion06]

That's another question. If the bolts are the pin mechanism in this connection, then why is the moment on the supporting member ignored? It seems like the shear tab is a small cant off of the supporting member, but that's not true for the conventional configuration.

 

[DaveAtkins]

If the supporting member can rotate (torsionally) a sufficient amount to create a pinned end situation, I am OK with this type of connection.

DaveAtkins

 

[ishvaaag]

You can model the behaviour with a 3d model. ANSYS, Visual Nastran etc cand do it. Even with simpler -to this purpose- programs it can be studied.

Redundancy is always a plus to final strength. There is also the code question, sometimes merely trying to keep the things in a particular practice, but sure there must be others that also work.

 

[Lion06]

Why in the world would I want to model this with software?!

 

[KootK]

I have a similar problem. My shear tab is concurrently a shear connection and an axial load transfer mechanism for a diaphragm drag strut. LSL bolt holes were provided which kinda screws up my whole drag strut arrangement. I'm contemplating having the tab welded to the supported member so that the connection can transfer axial force.

Surely the fact that the beam flanges aren't connected to the shear tab can be used to justify some pinned behavior, right? I'm going to concentrate my welds around the center of the tab to try to minimize the moment transfered.

 

[Lion06]

Kootenya-
I would be careful with welding like that. You don't want to initiate a crack and have it unzip. Can you provide slip-critical bolts in the LSL?

 

[KootK]

That's a neat idea SEIT. I might just do that. Are we not concerned about the moment resistance of the bolts once they are slip critical?

Also, do welds really unzip? Don't they have some minor ductility themselves?

 

[hokie66]

A welded-welded shear tab connection is just as good as a bolted one. The plate can flex, giving you the hinge behaviour you want.

 

[slickdeals]

Hokie66,
I know they allow welded-welded single angle connections because the angles can flex, but I am not sure if a welded-welded shear tab is recommended. I am not saying it won't work, but would you want to use it everywhere every time?

 

[Lion06]

Kootenay-

You still have to look at moment strength, but as you know the LSL doesn't allow for that because the bolts can move in one direction (This isn't quite true if you have more than one vertical column of bolts, but let's leave that aside for now). If you make it slip-critical and use the slip at strength level loading, then you should be good for counting on no slip right up to ultimate loading.

 

[Lion06]

I had to do a yield line analysis for a column cap plate on an existing building once. It was terribly overstressed using elastic analysis (450%), but using yield line analysis knocked it down to around 125%. I was quite impressed with that. I guess that's the extra capacity you get out of 4 yeild lines.

Anyway, the point was that I used the procedure outlined in my concrete text (McGregor/Wight), because I couldn't find any easy to follow references for steel. The concrete example was easy to follow and implement.

 

[hokie66]

slickdeals,

No, I would use welded-bolted shear plates. Some use bolted-bolted. But if I had one where the bolts didn't line up for some reason, an acceptable fix would be to weld the plate to the supported member, provided of course the misalignment is not very much. You would then just have a similar situation as if you cut out both flanges of an otherwise intact section. Still a pin.

 

[nutte]

Hokie, are you talking about the same connection? How can you have a bolted-bolted shear tab?

 

[ash060]

I don't see a problem with using a welded-welded connection. I would make sure that the moment capacity of the one weld group is more than the moment capacity of the plate to ensure the plate yields before the weld ruptures.

I wouldn't want to do it because you are probably going to have field erection bolts anyway to get it plumbed up.

 

[nutte]

Usually this comes up with a welded-bolted shear tab, where the bolts can't be made up in the field due to a fabrication error. The obvious fix is to weld the shear tab to the beam web. However, if this is such a good fix, why hasn't AISC come out and said so?

 

[hokie66]

nutte,

Shear tab is not actually in my vocabulary, but I thought I knew what it meant. I was talking about a shear plate connection between two beams, like in a cantilever and drop-in system. If the term only applies to a beam connecting to a column, no, bolted-bolted doesn't make sense.