Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

Plwd: Rigid Diaphragm Analysis - Opinions Wanted

[Subject Prev][Subject Next][Thread Prev][Thread Next]
I am in the process of working the new code, as many of you are. I would like 
to formulate a methodology that deviates from conventional known rigid 
diaprhagms (ie., concrete diaphragms). I would also like to figure a 
consensus of engineers who either agree and disagree and would like to 
understand why it may not be a valid engineering assumptions.

Problem: The determination for diaphragm torsion as applied to shearwalls in 
a non-rectangular (specifically "U" and "L") shaped structure. The building 
would be multistory wood framed with a plywood diaphragm.

Method: Rather than analysis the entire structure to determine the overall 
Centers of Mass and Rigidity, I contend that plywood diaphragms can be 
reduced to their basic "blocks" and each block designed for shearwall loads 
based upon the torsion contributed from that block.  The common line of shear 
where the blocks join must still be accumulated and the shearwalls in that 
line be balanced  by drift (deflection) criteria.

1. Plywood Diaphragms are admittedly (by most members of Seismology 
committee) neither rigid or flexible.
2. There is no way to create a rigid connection between wood members short of 
embeddeding them into another rigid material (ie, concrete, steel or earth). 
For example, a nailed connection can not be considered rigid however, posts 
embeded into a concrete foundation (although not particularly wise) will 
create a fixed support. Wood posts embedded into Tube steel may also be 
considered rigid by nature of the embedment.
3. The joint where two rectangular diaprhagms meet can not be made entirely 
rigid so as to eliminate rotation at the joint - by nature of the imperfect 

4. In the common direction (parallel to the long legs of a "U" shaped 
structure) - especially if the legs are symetrical the distance between the 
CM and CR will be governed by the 5% requirment of the code. 

Therefore, my opinion would be to allow the design engineer the latitute to 
simplify the design by isolating the separate blocks and design shearwalls at 
the common lines by summation of shears and wall rigidity analysis.
I am, however, unsure as to the contribution of torion in the long direction 
when both sides (legs) are symetrical. If the code required distance between 
CM and CR are neglected, will there be a sufficient eccentricity to cause a 
major increase in torsion induced loads?

Please discuss these with me (and other members). It may be an interesting 
discussion that could help work out a shortcut method without reduction in 
the performance intention of the code.

Thank you,
Dennis S. Wish PE