Moment connection in steel w/composite deck

[bookowski]

When using slab on composite deck is it typical to ignore moment connections when designing floor beams for gravity loads? I guess there are two parts - is this acceptable for the beam design and is this acceptable for the connection design.

If considering the moment connection then the composite action would be lost in negative moment regions but you get the benefit of the end restraint. If you consider it simply supported you get composite behavior but neglect the contribution of the end restraint.

From what I see so far it seems like people design for gravity as a simply supported beam and then design for lateral with moment connections.... does that make sense? I've seen discussions on this topic on here before but never had to deal with it (I believe that blodgett is referenced for this discussion). It also seems like people always refer to these moment connections as 'wind' connections... when would you have wind but not have to satisfy seismic also? Is this considered valid for seismic?

thanks

 

[WillisV]

It is common to treat the moment connected beams as non-composite for gravity loads (i.e. you do take into account the moment connection fixities and design the beam as non composite).

"Wind", "Type 2" whatever you want to call them is just a prescriptive type of PR connection and does not have to do with if the floor is composite or not (though you can use composite PR connections, and there is an AISC Design Guide on them). They cannot be used in R>3 seismic regions.

 

[Lion06]

It is a valid design approach and applies
to seismic, as well, they're just called wind moment connections. They are flexible moment connections and need to be distinguished from full moment connection. The connections are designed to yield at a specific moment, this is usually exceeded by the fixed end gravity moment.

When lateral loads are applied (for talking purposes I'm assuming a frame with two columns and one beam), the leeward connection tries to load further, but can't because it's already yielded. The windward connection tries to unload and that is where the frame stiffness comes from.

There are a couple hurdles to cross with this approach. You can't use the DAM with this lateral system and you need to use modified k factors if using the old-time approach, because only one lateral beam framing into a column can be considered fixed because of te loading/unloading behavior noted above. Additionally, that one connection that is considered fixed has a far end that is pinned - that requires modifications to Gbeam when using the nomographs.

Also, you need to make sure not only that the connection has only enough capacity for the wind moment, but also that the connection is controlled by yielding of te angle/plate and NOT by the bolts/welds. This is a rotational ductility requirement in The 13th edition manual.

 

[a2mfk]

when would you have wind but not have to satisfy seismic also?

Where I live (Florida) we have no seismic design requirements...