Fixed-Pinned Beams in Steel Special Moment Frame

[Charred]

Originally posted this in the Moment Frame forum and didn't get too many replies so I'm reposting here...


I've got a building in a high seismic area where I am using steel special moment frames. The building layout is somewhat forcing me to have my columns oriented differently in each bay. Can I have a fixed connection at one end and the other end not fixed? This would make the 'moment frame' span a bay or two until the next rigid connection.

Next question is can the 'non-moment connection' be a simple shear tab or would it need to be a CJP weld through the column web to make the beam continuous? I haven't found anything in AISC 341/358 that specifically allows or prohibits either.

I doubt I could use anything but hand calcs for the design. I know Risa won't run the SMF checks for a pinned-fixed beam. Also, I don't think the basic design/analysis formulas from AISC 341/358 would apply so I would have to be somewhat creative. Have any of you ever came across this scenario or have any suggestions on the design? Any references you know come to mind?

 

[JoshPlumSE]

Charred -

I think this can be done by code. But, it will be a little trickier. My immediate thoughts:

1) You calculate your Probable plastic moment at the hinge location is calculated the same way. But, the probable plastic moment projected to the face of the connection is different because it relies on the shear.

2) The Vg (shear due to gravity at the hinge location) is the same. But, the shear due to the hinge will be different. I would assume this would be approximately half of what it would be if you had hinges on both sides of the beam, right.

Mpr / (Lh) instead of 2*Mpr / Lh

In this case Lh would be the distance from the one hinge to the other end of the beam. Instead of distance between the two hinges.

3) At that point, the moment connection design should be pretty straight forward.

4) I would think the shear connection on the other side of the beam would be pretty important. I would follow the provisions for the web of whatever type of moment connection your using. So, probably CJP. Or, at least design the shear connection to resist the shear strength of the beam with strain hardening considered.



Note: I thought about this situation when I led RISA's efforts at adding the AISC seismic provisions. But, decided that it would be best to handle only the common cases. Maybe add additional options if users later requested them. Ultimately, that's why RISA detects whether your framing situation matches the program assumptions.

I've seen some cases (older / existing building) where there where a lot of moment connections, but then the end bay would have a moment connection from the beam to only one of its columns. Designed long before these seismic rules were in place. So, these situations exist..... While I can't immediately think of a reason why it wouldn't be allowed by code, you might have a tough time convincing the city / plan check engineer that it's allowed.