Wood diaphragm in tension/compression

[JIMEY]

I come from a region of high seismicity, so I am typically not worried about wind loads (seismic loads are always higher). Since the weight of a building tends to be more or less uniformly distributed throughout the floors/roofs, we usually assume that the seismic load in the diaphragm gets transfered via shear to the edges of the diaphragm, and we design collectors at those edges to drag loads to the shear walls.

However, I am doing a job now where we have concrete walls supporting a wood diaphragm. This means that you've got mass concentrated at one spot - where the concrete wall is. I suppose you could think of it as a line load on the diaphragm rather than a typical area load. My question is: how do we reconcile this? The only way I can see this load getting distributed throughout the diaphragm is if it is transfered via tension or compression (depending in which direction the building is shaking) through the plywood. But none of the literature I've read ever mentions designing diaphragms for tension/compression. This leads me to believe that I must be thinking about this incorrectly.

By the way, I suppose this scenario would be analogous to a typical wind load situation. You've got wind hitting the walls around the perimeter of the building. These walls transfer wind load into edges of the supporting diaphragm. Now how does that load get distributed throughout the depth of the diaphragm? Again, I can't see any way other than tension/compression in the plywood, but I've never heard of anyone designing plywood for that. I suppose it's kind of funny how seismic analysis actually makes more sense to me than wind analysis, but I guess that's because it's the soup I swim in. Thanks for any help.

 

[sticksandtriangles]

Can you provide some sketches to aid in this discussion?
I think you are talking about walls that run perpendicular to the applied loading? The walls shown in blue below:

I suppose you could think of it as a line load on the diaphragm rather than a typical area load. My question is: how do we reconcile this?

In general, we will add discrete braces to the concrete wall with hardware/members to resist tension/compression and then drag that load back into your diaphragm through diaphragm nailing. See detail below:

This will get the load into the diaphragm and into the traditional shear you design your diaphragms for. Read up on sub-diaphragms as well in ASCE 7 for a little more guidance on transferring local loads into your main diaphragm.





S&T -

http://www.re-tug.com

 

[driftLimiter]

@JIMEY you have identified an important aspect to flexible diaphragm design that should be considered carefully.

This part of ASCE 7- 16 relates to how the mass is loaded into the diaphragm...

12.8.4.1 Inherent Torsion. For diaphragms that are not
flexible, the distribution of lateral forces at each level shall
consider the effect of the inherent torsional moment, Mt ,
resulting from eccentricity between the locations of the center
of mass and the center of rigidity. For flexible diaphragms, the
distribution of forces to the vertical elements shall account for the
position and distribution of the masses supported.

Now given that method of loading the flexible diaphragm (i.e. increased line loads at the concentrated masses).
We also need to consider the local anchorage of the heavy walls and develop the anchorage forces out into the greater diaphragm.
like S&T said this is accomplished using the sub-diaphragm approach. This particular failure mode has been observed and should be closely examined to ensure proper behavior.

Read ASCE 7 - 16 section 12.11 for more detail.

 

[JIMEY]

Thank you all for your helpful responses.

@sticksandtriangles, your plan diagram is exactly what I am talking about. Your section you drew makes perfect sense to me if the framing runs perpendicular to the wall in question. If the framing runs parallel, how would you normally deal with this issue? I suppose you could place full-depth blocking between the joists, and connect them all with a strap.

@driftLimiter, I took a read through ASCE 7 section 12.11. It basically confirms what @sticksandtriangles suggested regarding sub-diaphragms. I think that basically answers my question. It's interesting because I don't often see designs like this where I'm from. Just so you know, I'm from Vancouver Island, Canada - so ASCE 7 is not a standard we use, and our Canadian standards don't get into that level of detail. But I do believe that what you're suggesting is the right way to do it.

 

[driftLimiter]

Its important to remember that the sections of code I have cited are each from chapter 12 in asce which is for seismic applications. There have been notable failures in this arrangement due to earthquakes over the past 50 years that has progressed this part of the code quite a bit.

 

[sticksandtriangles]

If the framing runs parallel, how would you normally deal with this issue? I suppose you could place full-depth blocking between the joists, and connect them all with a strap.

That's right. Strap for tension, blocking for compression. These connections are also notorious for cross grain bending and the failures that driftLimiter mentions, make sure you do not induce any cross grain bending in your connection.


S&T -

http://www.re-tug.com