Diaphragm Tie Plate

[palk7 EIT]

Hi,

In the roof diaphragm, as shown in the below detail, what is the need for this tie 3/8" plate? if there is a beam on both sides of the grid along this diaphragm line (J) and this line has a vertical brace frame positioned, could this 3/8" tie plate detail eliminated?

 

[KootK]

I would expect that the purpose of the tie plate is to axially splice beams forming part of the collector / chord system on either side of the column.

 

[palk7 EIT]

But then, the beams are connected to the column, and then they maintain the continuity for the collector/ chord system right? via shear tabs or however, the beam column connection is

 

[JAE]

The beam connections to the column would have to be designed for both the vertical shear load as well as the horizontal tie load - which the original designer perhaps didn't do - instead separating the function of both entities to "do their own thing".

The vertical shear beam connections probably have some inherent "give" with short horizontal slotted holes perhaps letting the tie plate do all the horizontal work without any load sharing.

 

[palk7 EIT]

Thats where I was going to, when the beam-column connection is designed for shear as well as axial loads, it should be fine.

However, I was playing with Ram connections and in the program for this particular case, where there are beam on both sides of the column the program doesn't consider both the beams when designing for the axial load, and the connection fails because of the web of the column (Beam to wide flange column web). The other design checks passed except that the column web failed. In this case, if there is a beam on the other side of the column, then this failure on the column web shouldn't occur right?

 

[KootK]

In this case, if there is a beam on the other side of the column, then this failure on the column web shouldn't occur right?

Sounds about right. I assume that we're speaking of a wide flange column and not an HSS column.

 

[palk7 EIT]

Yes its a wide flange column, the detail shows HSS (from other project)

 

[KootK]

You should be good then. The only EOR hang-up that I can envision, potentially, is concern for rotational ductility in the beam to column connection. You will often loose that when that connection is designed for axial load transfer. But, then, you often lose it regardless to some degree so I've not been inclined to lose much sleep over it.

 

[SWComposites]

The tie plate is presumably there to transfer the load in the beam flange across the connection, rather than forcing it down into the web. Much more efficient if there is significant bending moment in the beam at the connection.
Why would you want to delete the plate?

 

[palk7 EIT]

SWComposites: it becomes an additional site work and if it can be avoidable trying to ensure it is ok to avoid