i have always calculated the chord

huangyhg

i have always calculated the chord
i have always calculated the chord forces for a simple diaphragm building and placed an appropriate sized continuous member around the perimeter.
  i know someone who has been designing buildings since the 70's who has never done this. i believe he has been successful in part because most of what he has done has been simple one story box warehouses (cmu wall with steel jst and deck) located in a lower wind/seismic zone. he basicly uses the same angle no matter what the building size or location is.
  i guess you could figure in the section of the entire roof, then check the deck's capacity (including attachments) to help carry the chord forces itself. is anyone else doing this?
haynewp:
there are lots of engineers out there designing the sme way as your friend, but i suspect in most cases it is out of ignorance.
another common error, i find, is in one story wood buildings using engineered metal plate trusses. no connection is provided between the plywood sheathing and the sill plates. so even if a continuous bond beam is provided as a chord member, there is no way to tranfer the diaphragm forces into it.
on the same subject, it is my contention that in a gable roof, there are two distinct diaphragms since the shear forces on the windward side are "out of plane" with the forces on the leeward side. it is my practice to provide a chord   
i have also worried about the 2 distinct diaphragms as you mentioned. especially with higher pitched gables and no lower (horiz) ceiling below to help take the wind load from the wall.
haynewp,
you are designing diaphragms properly.
in a pinch (say, justifying an existing diaphragm), i have utilized the roof deck at the edge of the building as a chord.  when i do this, i check the welds (or stitch screws) at the ends of the sheet for the tension (or compression).  then i reduce the depth of the diaphragm for calculating shear in the diaphragm by the depth of deck required for the two chords.
redhead,
i have had numerous arguments with contractors, architects, and even other engineers about blocking between roof trusses.  blocking is the only surefire method for transferring shear out of the edge of the diaphragm into the chord at the top of the wall.  unfortunately, i am a little ashamed to admit that i have backed down on being so rigorous, and when the heel height of the roof truss is small (say, less than about 12"), i do not insist on blocking.  all the trusses, nailed to the roof sheathing, can act sort of like a rigid frame parallel to the shear wall.
dalese (visitor)1 jul 02 10:46
daveatkins,
i have commonly seen truss blocks left out, particularly in areas that are low seismic zones.  my question is when you are counting on the trusses to act as a rigid frame over the shear wall, what mechanism are you using to develop the moments at the top and/or bottom of the trusses.  at the bottom of the truss, since there isn't a block for shear transfer, the toe-nails from the truss to the top plate are resisting the shear.  at the top of the truss, the nails from the sheathing into the top chord of the truss are being used to transfer the shear from the roof sheathing.  do you count on the sheathing nails to develop a sort of fixity at the top of the truss?  where does the boundary nailing go?
daveatkins
i agree with the rigid frame analogy, but this is not a substitution for the chord   
i use the simpson holddown straps (required at each truss here in wisconsin) to resist shear (simpson publishes values for shear in their straps perpendicular to the trusses). and yes, i assume some fixity develops at the sheathing.  you end up with a rigid frame with pinned bases.
i still think blocking is the best solution.