shear transfer with ship sheathing

huangyhg

shear transfer with ship sheathing
in the northwestern u.s., and presumably elsewhere, where it is popular to use cedar shake roofing, it is common to install this over skip sheathing.  rather than lay down a full plywood diaphragm on top of the rafters, 1x4 or similar boards are run perpendicular to the rafters and are spaced apart approximately 9 to 12 inches to allow the shakes to be nailed to them.  my question is how is the lateral shear transferred to the walls?  is the nailing of the skip sheathing boards to the rafters enough, or is there typically some sort of additional diagonal bracing installed?
thanks in advance,
tom w
check out our whitepaper library.
i do not live in the nw but i would presume that the higher rigidity would be in the gypsum board ceiling.  i have never seen any "diaphragm" criteria given for shake shingles.
skip sheathing is not considered a diaphragm, it considered structural roof sheathing. the uniform building code was changed in 1997, and "part ii-requirements applicable to engineered design of wind and earthquake load-resisting systems" and a new section (2315) were added. if the roof sheathing was to be considered a diaphragm it had to be installed at a 45 degree angle to the framing   
thanks for the feedback, i appreciate your comments.
tom w
i've used diagonal tension bracing as well as let-in bracing with spaced cleat roofing systems.
i saw an engineer do an analysis on spaced cleats with variable nailing patterns similar to the analysis of metal diaphrams. i suspect it was for the purpose only of  justifying a stable model rather than anything more than that. i think the practice defies rational analysis.
old houses used spaced sheathing different from the nailers used for shakes. that old spaced sheathing was maybe 4 inches wide with 1 inch gaps. i think the code provides a diaphram shear value for this.
shear transfer is handled by the moment couple in the two nails used to attach the skip sheathing boards to the rafters.
it isn't a very big moment capacity per connection, but there are a lot of them, and it adds up.  aitc has an example calculation in their handbook.