Braced frames - connection type for analysis purposes

[formi]

Hello guys!

I'd like to find out your opinion about modelling issues of braced frames. As horizontal members of braced frames (girders) are subject to high axial loads, should it be a moment connection between horizontal member and column or not really? I mean, theoritically if we design braced frame we don't need moment connections in it, but when there are huge axial forces in bracing members, how to transfer them onto columns without moment connection? Summarizing: what is a better approach in modelling braced frames- treat all members as axial members only (diagonals and girders) and design connections accordingly or take advantage of massive axial connection due to large axial force and do it as moment connection--> bending moments in horizontal member even out?
Please see the attachement.

 

[KootK]

Is your connection in a high seismic application? Is your brace concentric or eccentric? US?

Normally, I would pin the far ends of the beams. Axial transfer capacity alone usually doesn't necessitate fixed connections. That being said, your modelling should reflect the real world detailing. The connection in the photo certainly looks fixed to me.

In an eccentric, seismic application, fixity provides a redundant lateral mechanism (moment frame). If I remember correctly, I think that US codes reward that in the force reduction factor.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[formi]

Low seismic risk, CBF's. In the US application I've seen mainly fixed connections but it's probably due to seismic provisions I think. But on the other side, if assign pin condition for beams how to transfer an axial force of around 1000 kips? Maybe sth like in the attached pic?

 

[KootK]

I'd investigate the connection a bit and let that be your guide. If you end up having to attach the entire section for axial transfer then sure, call it fixed.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[paddingtongreen]

It looks as though the load that would cause the moment is horizontal. If so, therte will be points of contraflexure in from each end. By modelling the joints as pinned, you force points of contraflexure at those pins. See where they would occur in the fixed model and see if it is very different.

Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin