the validity of adding straps to the bottom. We talk about deformation
compatibility often when we discuss seismic, what about the straps? How
much will the GLB need to deform before the nails or screws engage the
strap? However, that doesn't solve your problem. I also second
Bill's comment. Make sure you are using the correct allowable
stresses. If necessary, you may even be able to regrade the dimensional
lumber to improve your capacity.
solution I used recently was to add kickers to the columns and reinforce the
columns for bending. You will have unbalanced loads, so you get moments
in the columns. The kickers can shorten the span of the GLB's. Be
careful, you create tension stress in the top and most simple-span GLB's have
only half the tension capacity in the top as compared to the bottom. You
might also try creating moment continuity with the short GLB along with adding
kickers to the columns. But by the time you do all that, Patrick's idea
of a new frame may make sense.
From: Gerard Madden, SE
October 28, 2003 4:48 PM
Subject: Glulam Roof Girders in a
I’m working on everyone’s favorite type of
project – New Rooftop Mechanical Units on an existing panelized wood roof
on a 1 story tilt-up building. The system is comprised of 2x4 subpurlins @
24” o.c., supported by 4x14 purlins @ 8’ o.c., supported by
25.5” deep glulams beams (simple span of 40 feet) at 24’ o.c.,
supported by glulam girders (a series of single cantilevered continuous beams).
It creates 40’x45’ bays to columns.
I’ve checked the 4x purlins and glulams girders and
have determined both are overstressed under just DL + LL without the new units.
Accordingly, I am reinforcing all of the existing 4x purlins that have new
units sitting on top of them.
For the glulam girders, it is a series (13 bays) of single
cantilevered glulam beams with the cantilever of 10 feet and a backspan of 35
feet. Under pattern loading, the original design has some overstresses on the
order of 25-30%. When I walked through the building, I noticed that someone
along the line also came to the same conclusion I did and tried to reinforce
the glulams. They placed simpson coil straps (or just a really long strap) on
either side of the bottom of the glulams at the last lam. The strap is probably
20 feet long at about middle of the back span. The new units add about 4-8%
more demand load to the glulams so the situation is getting worse.
Another odd thing they did (or maybe not so odd) was that
the glulam beams spanning 40 feet frame in about 1.5 feet away from the column
to the girders. This creates a huge shear stress in the glulam, but by taking
advantage of the D from face of support clause, the glulam would not be
Adding the straps (as the previous engineer did) helps the
tension at the bottom, but does not increase the compression capacity at the
top. It is difficult to strengthen the top due to the deep glulam beams that
hang off the girders and would this interrupt any new side mounted reinforcing
steel plates. Some of the overstress, I suspect, is that the glulams were design
without the volume factor (Cv) for glulams – the original was designed
under the 1976 UBC…. There’s a question here coming…
What are some solutions you would recommend or care to
comment on. I was thinking adding more straps on the underside of the glulam
for the tension increase, and at the top lam on either side. Another option
would be to use a channel or lt. gage steel track lagged/screwed up to the
bottom, and side mount channels to the top on either side just under the
sheathing. How would you tackle the problem of an intersecting beam at the top
every 24 feet and located near the highest stress points (over the support and