We did this stuff at school. But the way it is
reiterated here (no pun intended) I wonder if I could get through school if I
had to do it again.
Subject: Re: Standard
Portal Frame Analysis
Date: Mon, 31 Oct 2011 23:48:22 -0600
I am assuming you want
to do this manually.
For a one storey multi
bay plane frame structure where the axial shortening of the second storey
"beams" permits only very small joint translation relative to the magnitude of
joint translation permitted by bending of the "columns" I would proceed as
1.) Pin a convenient upper storey joint (call it Joint
"A" for future reference) to provide the location for an
artificial horizontal reaction. Give Joint "A" a unit deflection
keeping the beams rigid.
2.) Calculate the fixed end moments (F.E.M.) for the columns
(note: the beams have zero F.E.M. because you kept them rigid in step
3.) Now release the formerly rigid beams and do a moment
distribution analysis to determine all of the beam forces, moments, and
reactions you may need later, including the horizontal reaction at Joint
4.) The horizontal reaction at Joint "A" divided by the unit
deflection from step 1.) will give you a spring constant (lets call it
"k" for all horizontal (sidesway) movements.
5.) Analyze the structure from 1.) including the horizontal
reaction from Joint "A" for each of your actual load cases assuming no
initial deflection for Joint "A" to determine all beam forces,
moments, and reactions, including the horizontal reaction at Joint "A"
(call it Ra).
6.) The amount of sidesway will be Ra/k.
7.) The forces on your real structure (including the effects
of sidesway) will be the sum of (the forces and moments from 5.) + (Ra/k)*(the
forces and moments from 3.)
I have assumed a
moment distribution analysis because that is what I am most familiar
with; but any method of analysis should do.
H. Daryl Richardson
----- Original Message -----
Monday, October 31, 2011 12:28 PM
Standard Portal Frame Analysis
I’m doing stuff I should done in school J
I am reviewing how to quickly arrive at sway in a rectangular
(or any, for that matter) portal using simple portal frame equations.
I calculate the moments from the std equations and then
release the top corners to arrive at a flagpole concept tied at tops by a
The resulting base moments are approx by iterative moment
If I use a partial, or offset, load on the beam I expect
My question is, “Is it too simplistic to take the resulting
moment difference at the bases and apply slope-deflection arithmetic to
arrive at an estimate of the sway?”
Thor A. Tandy
P.Eng, C.Eng, Struct.Eng, MIStructE
Victoria, BC, V8T 1Z1
consider the environment before printing out this