Need a book? Engineering books recommendations...

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

[SEAOC] Re: Basic Roof Rafter Analysis Question

[Subject Prev][Subject Next][Thread Prev][Thread Next]

When wood shingles/shake were "outlawed" a few years ago, I was doing 
a lot of this work for various roofers. I found a lot of different 
conditions and very poor connections in the roof structure. But to 
your point...

I checked the 2x purlins as rafter supports, which resulted in 
multiple span design of the rafters and found many times, the rafters 
would work, but not the purlins.  The span of the purlins is 
typically reduced with struts that may go parallel or perpendicular 
to the purlin (and sometimes angles in between).  Many times, these 
struts are supported by interior walls with no footing or double 
joist.... (another possible problem)

For struts parallel to the purlin, I just analysed the purlin as a 
pinned frame. This quite often required the addition of a 2x member 
at the center span to handle bending or deflection problems. Again 
the connections at the pins need investigated.  Many of these were 
just angle cut struts with one nail.  I would use plywood gussets or 
nailed metal plates or scab a 2x on the side of the strut that would 
overlap the purlin.

For struts perpendicular to the purlins, the problem is a bit more 
difficult as quite often there are not oposing members.  Therefore 
I'd take the closest rafter to the strut and use that as a component 
of a "truss".  The purlin would impose a vertical load that is 
resolved by the strut, and the opposing horizontal thrust caused by 
the slope of the strut is resisted by the closest rafter.  This 
requires investigation of the purlin in shear as the the rafter may 
be 12 inches away from the strut.  This also requires investigation 
of the connections.  Many times the connection of purlin to the strut 
and the rafter needs improved not just to resist the forces, but also 
to address the cross grain bending as the strut is attached to the 
bottom of the purling and the rafter to the top.

In addition, the rafter ties typically need checked.  If they were 
face nailed to the rafters, quite often the nailing in inadequate to 
resist the horizontal thrust of the rafter. Sometime the addition of 
ties just above the purlins work and are easier to install.

Additional loads to headers are not usually a problem when the change 
is from shake to lt wt, but to full mission tile it could be.  The 
installation of the mission tile makes a big difference.  Just the 
tile may not impose much but if installed with a lot of cement for 
"asthetic" reasons... it could kill you..  (Another thing to remember 
with any tile, is the addition of plywood shtg -- 5/8" in LA city I 

Another thing to consider is the additional load to the lateral 
bracing system... While I believe the "Blue Book" used to have 
something in it about increases less than 3 percent could be 
overlooked, I use the change in roofing to "look" at the rest of the 
structure for inherent weaknesses and point these out to the owners. 
I believe UBC actually says ANY increase in the loads or change in 
the capacity requires the upgrade to current code.  Most of the time 
I can convince the owner of the practicality of "upgrading" the 
system. Not only does the address the walls, but many times, the 
shear transfer from roof diaphragm to the walls is inadeqate. Drag 
struts, collectors and ties at re-entrant corners can then also be 

On complex roofs, the "engineered look" of an existing systems quite 
often reveals problems.  Especially when the building was "just 
built" per general construction provisions... But that's a completely 
different thread <G>

My apologies to any who think I get long winded and wordy....

Steve Privett  CE