] [Thread] [Date] [Author]

["Rand Holtham, Sigma Engineers" <rand(--nospam--at)sigmaengineers.com>]

 

----- Original Message -----

From: Rand Holtham, Sigma Engineers

To: njineer(--nospam--at)att.net

Sent: Monday, January 27, 2003 5:11 PM

The classic roof diaphragm model comes from a box structure with a flat roof. This is simple to grasp because it simple to see a vertical shear wall turned horizontal. Its also easy to understand how the chord members work in relation to the web (plywood diaphragm). It seems to me that engineers somewhat blindly said hey it'll still work if we bow the center of the diaphragm upward to form a ridge. In my mind this can be pictured as the web of an I beam buckled laterally. Obviously this weakens the I beam, but if we have a horizontal tie from flange to flange (like a roof rafter in the Roof Diaphragm situation) that prevented the web from moving perpendicular from the plane of action the web would continue to perform it function. Now to further complicate the model of the flat roof diaphragm there is a slit in the diaphragm at the ridge. The shear stresses must transfer out of the diaphragm into the framing (at the point where the theoretical shear is at it highest (near the shear wall)) and then back into the diaphragm on the other side of the ridge. If you split an I-beam down the middle along its length and then nominally tacked it back together  it would work for very small loads but under high loads the tack welds would break loose and you would be left with two tee sections not one I section. It's my belief that all the misc. framing used in roof construction that allow the single diaphragm concept to live on. And if you could test load a whole roof structure it would probable behave nicely, but in a cathedral type ceiling you lose all that misc framing and you (I think) have to separate the roof into two diaphragms and resolve the loading accordingly. When the wind hits the gable end wall you really have a problem because now the split at ridge is parallel with the load that's like an Ibeam with the web completely removed. But if the chords are somehow made continuous across the ridge you may be alright because the shear theoretically is zero here and the tension/comp in the chords is highest. As far as the hip situation I don't think it can behave as a diaphragm since you have discontinuities at all the critical locations. Here I believe you have to rely on the ceiling for your diaphragm.

 

Rand Holtham, P.E.

 

 

Rand,

 

Thanks for your comments.  I think that they would be helpful to the seaint list, if you would forward them there also.

 

I've never seen a hip-roof diaphragm that displayed seismic damage -- perhaps wood-framed residential structures are light-weight enough that diaphragm deficiencies are accommodated "somehow".  However, in my work on old structures that place significant demand on diaphragms [walls of stone, adobe and other types of unreinforced masonry]  I install a horizontal diaphragm whenever possible [however, sometimes historic authenticity will not allow adding a diaphragm under an exposed vaulted roof structure, for example].

 

I have seen evidence of a roof diaphragm edge being lifted at the eaves off the top of a brick wall that the steeply sloped diaphragm tended to brace -- a condition that I interpreted as indicating that the diaphragm was deforming in its plane -- a deformation that had a vertical component.

 

Nels

 

Nels Roselund
Structural Engineer
South San Gabriel, CA
njineer(--nospam--at)att.net