Need a book? Engineering books recommendations...

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

Fw: lateral drift

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

> From: Mark & Jessica Pemberton <jmpember(--nospam--at)>
> To: seaoc(--nospam--at)
> Subject: lateral drift
> Date: Saturday, March 21, 1998 3:41 PM
> I recently designed a trellis using a cantilevered
> column lateral resisting system.  I seems that
> deflection always controls on these type structures
> when trying to abide by allowable lateral drift.  (0.005h)
> First, is this what should be used for wind as well
> as seismic lateral drift?  Secondly, because the low stresses
> and the type of structure involved is it acceptable to
> allow much more than this?  At what point does the amount
> of deflection render the predicted stresses in substantial
> error (with formulas using the assumption of small deflections)?
> I have seen steel building manufacturers use allowable drifts
> as much as L/100.  Where do they get that?  (by the way the
> trellis involved utilized tube steel - OK HSS to be precise -
> and the wall thickness would be 1/4" by stress/column analysis
> and nearly 1/2" to meet the 0.005h requirement)
> Mark Pemberton, P.E.

I figure my first question would be whether there are any adjacent
structures that could be pounded if you had excessive deflections under
seismic loading.

If not, then are there any attached items that would be damaged under
excessive deflections.  Is the owner concerned about serviceability with
regard to this structure.

If the answers to the above are both no, then my next concern would be the
ability of the structure to endure the required deformations.  What type of
reduction factor are you using to obtain the design seismic base shear?
(Rw=3 would seem appropriate, given the lack of redundancy).

One approach might be to determine an expected target displacement using a
displacement spectra (you can use a standard acceleration spectra and work
backwards, although with the smoothing of the code design spectras this
approach may yield a conservative result, especially for long period
structures).  If your column members are compact, and can reach Mp, then
you can assume an appropriate hinge length, and determine the strains in
the columns at the target displacement (this displacement would be the
actual expected displacement during the design, unreduced, earthquake, so
Rw=1).  Since you should basically have an SDOF, this shouldn't be too

This way, you could look at the actual seismic situation, instead of
dealing with reduced forces and smaller than expected deflections.  Even
though this may not be in strict accordance with the wording of the code,
it is more rational IMHO.

Anyway, if you determine the strains in your columns and they look OK, then
everything should be copacetic.

Actually, for this particular case I would probably be inclined to go with
a stouter member unless there are just a whole lot of them.  In these types
of small systems, I figure that a response as near to elastic is best,
primarily because of the lack of reduncdancy and the small nature of the

Hope you figure this one out!

T. Eric Gillham