Subject: Re: Truss Diagonals in Compression (AASHTO)
Date: Mon, 24 Apr 2000 12:11:20 -0700
I'm not sure who is "right" in this case, but my practice is to go from panel
point to panel point when considering AISC slenderness ratios. I then use this
slenderness to determine buckling loads for both the member and the gusset
plates. note that the L in the KL/r ratio is defined as the laterally unbraced
length of the member. Unless the gusset is REALLY stiff, I don't see how it can
provide enough lateral support to be considered a lateral brace.
Rick Drake, SE
Fluor Daniel, Aliso Viejo, CA
"Polhemus, Bill" <wlpolhemus(--nospam--at)sbinfra.com> on 04/24/2000 10:59:29 AM
Please respond to seaint(--nospam--at)seaint.org
To: "'seaint(--nospam--at)seaint.org'" <seaint(--nospam--at)seaint.org>, "'structx(--nospam--at)topica.com'"
Subject: Truss Diagonals in Compression (AASHTO)
I'm in a sort of "dispute" with a "peer review" engineer. They're checking
our design calculations for retrofitting some overhead sign supports on a
freeway project where the existing signs are begin replaced with very heavy
electronic message boards.
In reviewing the existing structure, a truss supported on large concrete
columns, it was noted that while the compression stresses in MOST of the
diagonal truss members are fairly low, these members are comprised of single
angles, only 2.5" x 1.5" (a non-standard size, leading me to wonder if
they're cold-formed). Since the l/r value I calculate for these members is
greater--in some cases much greater--than 200, I called for them to be
replaced with larger "standard" size angles.
My rationale is that, even though such members have "worked" for lo, these
many years, and since we're putting much heavier signs on these supports,
and since my "okaying" smaller members means I think they'll work fine under
the new loading, AND since a plaintiff's attorney would have a FIELD DAY
with me in the witness box should one of these signs for whatever reason
decide to part company with its support and attempt to occupy the same
space-time as a human being passing unsuspecting in an automobile below....
Well, I think you get my drift.
Problem is, I typically calculate the "effective length" of truss members
from panel point to panel point. The peer review guys claim that "the code"
allows you to go from actual end of member to actual end of member. I
disagree strenuously with this, since I happen to know that gusset plates
aren't infinitely rigid.
So, what do you think? Who's right?