It has been my opinion for some time that a drag strut need only be as long
as required to develop the capacity of the diaphgram equal to the demand of
the shear wall. Therefore if a 10 foot wall lies in the middle of a room and
has a demand of 4800 pounds, then the connection to the diaphragm if at one
edge (and with a diaphragm capacity of 240 plf) need only be 20 feet long.
However, if the wall lies inside the room, the drag needs to be only 10 feet
or the length of the wall provided the nailing of the diaphragm to the
shearwall is equivalent to 240 plf at each side of the wall or 480 plf in two
rows of nails.
I've had a discussion with another engineer who believes the drag "strut"
needs to develope the walls demand over the entire length of the diaphragm. I
disagree with him with one exception. If the drag is to be developed in a
roof rafter nailed typically with the equivalence of field spacing (say 12"
o/c) verses boundary or edge spacing (6" o.c or closer) then the capacity of
the strut will depend on the value of the actual nailing and additional nails
may need to be added to balance the demand of the wall to the capacity of the
The other engineer believes that if the drag were not the length of the
diaphragm, the drag could pull away from the remainder of the diaphragm at
the point where the drag stops.
I understand his opinon as it is typical at the perimeter of a structure with
openings where the drag across the opening must be sufficiently straped to
transfer shear through the strut above an opening from one shearwall to the
next. However, I believe that if the wall occurs only on one side of the
opening the strut above the opening must be sufficiently connected (strapped)
to develope the equivalent difference in shear from the length of diaphragm
connected to the wall and the amount of shear needed to satisfy the rest of
the demand. Generally the shear in the diaphragm is sufficiently low and the
capacity of the walls high enough to effectivly transfer the shear through
the plates and only consideration is made when the plates are discontinuous
by discontinuity caused by a header or beam.
Finally, I use the example of a sub-diaphragm analysis where the capacity of
the strut connection normal to the wall only needs to be straped together for
the length of the diaphragm necessary to develope the capacity of 30% of the
wall weight between struts.
I would appreciate any comments you might have?
Dennis Wish PE