From: Charles Greenlaw <cgreenlaw(--nospam--at)speedlink.com>
Date: Wed, 26 Apr 2000 20:28:50 -0700
Mike, you have succinctly set forth a very good case for LRFD in seismic
codes. I was not in doubt that such a case exists. Neither did I argue
against LRFD in either concept or in adequacy for its purpose. I did argue
against the glaring circularity in Rick's assertion per se (along with
noting that for me, ASD remains convenient and trustworthy in minor and
occasional steel designs.)
My "inside" knowledge is what we all know: recasting seismic codes overtly
into Strength format awaited development and acceptance (as valid) of
strength design methods. But behind the older, Allowable Stress seismic code
format has always been a strength awareness. Christopher Wright has offered
many postings that relate the commonality of the two approaches, which in
presentation are at heart inverses of each other. Just because it finally
became possible for code adoptors to have chosen strength-formatted seismic
code is not a Hobson's "reason" to embrace the means that made it possible,
or to celebrate their decision. It's like the boy in Hans Christian Andersen
who failed to approve in a circular way the King's choice of new clothes,
just because they were the King's choice for reasons the King found compelling.
As for the Strength-based seismic code format necessarily being incompatible
with stress methods, let alone obsoleting them, I submit the current
treatment of woodframe seismic systems: The strength seismic code had to
contain a translation of itself back into stress design to accommodate wood,
for which a settled strength design code is not yet available. If that
reversion can be done for wood, it can still be done, as before, for all the
other materials. Hence two design methods exist for them, and the choice of
which method to have in code remains an arbitrary one.
Methinks maybe those who chose Strength methods for adoption into seismic
code are a little touchy that not every "subject" in the "Codedom" shares
their enthusiasm for it in every particular.
But alas, where choices exist, differing preferences will also exist.
Thankfully we have in this this e-mail listserver a forum for these
differences to be openly exposed and shared.
BTW, thanks for your recent excellent report and candid commentary on the
deliberations at ICC on revisions to the 2000 IBC.
At 05:32 PM 04/26/2000 -0700, you wrote:
>Since I'm generally tired of the ASD/LRFD debate, I'll be reasonably
>quiet. However, I believe that I can elaborate on a point Rick made
>so that it becomes apparent that the reasoning was not circular.
>Rick and you wrote:
>> >Another reason for structural engineers to support LRFD
>> >design is that all U.S. seismic codes are strength design
>> >based, including 1997 UBC, 1997 NEHRP Provisions, 1999 BOCA,
>> >ASCE 7-98, and 2000 IBC.
>> This is no more than a circular argument. Here, the
>> egg is chasing the chicken. The code basis is as arbitrary as
>> is the steel design choice. Either could just as well be the
>> other way around.
>A significant portion of seismic design is based on calculating some
>limiting forces. In such cases the primary design parameter is the
>expected "real" capacity. This "real" capacity is slightly increased
>(for conservatism) and used as a demand on other parts of the
>structure. (Using ASD terms) if we want to have a uniform "factor of
>safety" against being wrong in our calculation of such "real"
>capacities we need to use an LRFD-style approach. ASD hides the
>factors of safety (which also differ for various limit states). Once
>we decide that we want to use "an LRFD-style approach" (as is done in
>a lot of ASD seismic design and detailing for the reason described
>above), I believe it is a reasonable step to simply use LRFD.
>- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>Michael Valley, P.E., S.E. E-mail: mtv(--nospam--at)skilling.com
>Skilling Ward Magnusson Barkshire Inc. Tel:(206)292-1200
>1301 Fifth Ave, #3200, Seattle WA 98101-2699 Fax: -1201