seismic connection between buildings

[tmr4791]

I am working on an expansion to a steel framed freezer building. The freezer building is a rectangular box, approxmately 40' tall, in seismic design category D. Because of thermal issues, I am investigating tying the buildings together (instead of providing a seismic separation per code). Since the column lines will match, my current thinking is to bolt or weld the new and existing columns together at the roof. The existing freezer has concentrically braced frames on all sides, and the new expansion will also have concentrically braced frames on all sides. So the buildings have roughly the same rigidities. So, for the combined building, I will have two separate diaphragms connected at intervals. Does this sound reasonable? Does anybody know where I can find some information about about how to determine the design forces in the connectors between diaphragms? My first thought was to consider them as "collectors" to be designed seismic overstrength. But working through some quick numbers, I think that this large force will overwhelm the joist and beam connections to columns on the existing side of the expansion.

 

[ishvaaag]

Since you will be having a new structural behaviour, it is from the combined model that you can extract the proper forces.

You might start thinking of the joined parts as if being one, say from two columns one, and then see that in more than the correct overall behaviour the connectors meet the requirements of union.

For joists of different pattern this may not work but one always can get forces for the discrete link connectors to be put in place, once put in the model. This also would work for the columns, always exercising engineering judgement.