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

RE: Quick/Easy Question for Seismic List

[Subject Prev][Subject Next][Thread Prev][Thread Next]
I'm with you Ben.  This may help, but then again ... consider steel SMRF w/
13' story heights-

1994 UBC (T<0.7 sec.):
max. drift = 0.005(13')(12"/') = 0.78"
           = (0.04/12)(13')(12"/') = 0.52" ... governs (at ASD "elastic"
level drift)

1997 UBC (T<0.7 sec.):
max. drift = 0.025 (13')(12"/') = 3.9" ... (at maximum "inelastic" level
drift)

To compare 1997 UBC drift to 1994 UBC drift, you need to divide by 0.7R to
get to elastic level and also divide by 1.4 to get to ASD level, therefore:

(3.9")/(0.7)(8.5)(1.4) = 0.47" ... approximately equal to 0.52" per 1994 UBC
(unless I blew a digit some where, which wouldn't be the first time).

Bottom line, drift should be in whatever units you multiply by the maximum
allowable "drift ratio" (i.e. 0.025, etc.).  Whew!

Steven T. Hiner, SE
Folsom, CA

-----Original Message-----
From: Yousefi, Ben [mailto:Ben.Yousefi(--nospam--at)ci.sj.ca.us]
Sent: Thursday, March 23, 2000 1:42 PM
To: 'seaint(--nospam--at)seaint.org'
Subject: RE: Quick/Easy Question for Seismic List


It's amazing how such a simple issue, which was clarified earlier, has
turned into a huge discussion item. The drift is always a percentage of the
story height, hence unitless. There are no changes from 94 to 97 UBC in the
definition of building drift. The only change is the allowable limit, which
by first inspection seems excessive. But that is only because the calculated
deflection (at strength level) gets multiplied by 0.7R.

I sure hope that I am not missing something here, otherwise we have been
checking drift calculations incorrectly for almost 9 months now! 

Ben Yousefi, SE
San Jose, CA

	-----Original Message-----
	From:	SEConsultant [SMTP:seconsultant(--nospam--at)earthlink.net]
	Sent:	Thursday, March 23, 2000 1:12 PM
	To:	seaint(--nospam--at)seaint.org
	Subject:	RE: Quick/Easy Question for Seismic List

	Paul,
	It's a shame that the evolution of the code equations have not been
	documented so that these issues are simply not questionable. I
personally
	think this is a failing of the code writers (again).

	There is a general opinion that empirical formulas often use
arbitrary
	constants and that the final units are assumed. For example, the
first term
	of the diaphragm deflection formula (for blocked diaphragms)
represents the
	chord deflection. We had a number of debates that this was an
empirical
	formula and that it was assumed in inches to match the other three
	components of the formula. However, one or two engineers responded
to the
	list and recreated the derivations of the formula from the basic
deflection
	formula to the one used in the code. The constants assumed to be
unitless
	were, in fact, conversion factors.

	I am not convinced that at first glance we should make assumptions
as what
	the final units should be. It is the responsibility of the code
writers to
	insure that the formulas and glossary of terms is written so that
the
	engineer understand what units of measurement is expected as input.
This
	way, the engineer knows whether or not he or she needs to convert
the
	formula.

	This one short thread has had very strong opinions from each side.
When we
	thought that it did not matter if the we were to consider the height
in feet
	or inches, Shafat's opinion is that the formula is unit dependent
and
	therefore it matters.

	I'm still confused and think the answer should come from the "horses
mouth".

	Regards,
	Dennis S. Wish, PE


	-----Original Message-----
	From: Paul Crocker [mailto:PaulC(--nospam--at)ckcps.com]
	Sent: Thursday, March 23, 2000 10:44 AM
	To: seaint(--nospam--at)seaint.org
	Subject: Re: Quick/Easy Question for Seismic List

	Why do you assume that if you put feet into the equation that it
yields
	something
	other than feet as its result?  It is also acceptable to use meters
for
	story
	height.  Surely this could not be so if the 0.02 was calibrated to
make a
	feet to
	inches conversion.  If the EERI web site still has the field report
from the
	earthquake in Columbia last year (?) posted, it might be worthwhile
to read
	this,
	as it briefly discussed observed correlations between drift and
	damage/failure,
	which largely agreed with the pre-existing body of research.
Depending on
	your
	anticipated loads and building use, getting too close to the 1.48'
drift may
	not
	be appropriate, but I do not see it as a code violation.  Also, if
you get
	too
	close to the drift limit, drift will exceed the threshold beyond
which
	P-Delta
	analysis cannot be ignored, which may cause problems of its own.
Overall,
	though, if a maximum inelastic drift limit approach is taken, a
limit of
	1.48'
	for a 74' building is not surprising.  Surely moment frames would
never be
	possible if a 1.48" level of stiffness was code required.  In fact,
only the
	longest shear walls could possibly comply with that.  Try to
visualize 1.48"
	in a
	74' building under extreme loading and imagine if that would be
possible.

	Paul Crocker