Need a book? Engineering books recommendations...

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

Re: asd vs lrfd (not lfrd)

[Subject Prev][Subject Next][Thread Prev][Thread Next]
-----Original Message-----
From: Lynn Howard <lhoward(--nospam--at)>
To: seaoc(--nospam--at) <seaoc(--nospam--at)>
Date: Sunday, October 12, 1997 4:39 PM
Subject: Re: asd vs lfrd

>Jon turner wrote:
>> I was wondering why allowable stress design is still being used in
>> the United States.  I am a practising structural engineer in Canada and
>> been using limit states design, which i believe is called lfrd in the US,
>> since I graduated from university.  I do not have much knowledge about
>> so can somebody clue me in on the advantages and disadvantages between
>> two systems? Also I am wondering why you have so many exams to become
>> registered as a professional engineer.  Here in Canada, after you
>> you are registered as an engineer in training.  After a period of two to
>> four years, depending on your province of residence, as an engineer in
>> training you become a professional engineer after passing a civil law
>> related to engineering.  Many people have told me the reason for the
>> differences in our systems is that your engineering programs are not
>> regulated from university to university, so the quality of students is
>> judged more from whether they can pass the exams after they graduate.
>> engineering programs are regulated by a central body insuring that all
>> students are taught the same courses and that all of our engineers are on
>> equal footing.
>It has nothing to do with the ability of the graduating engineering
>All recent engineers can do lfrd design.
>The problem is with the Boss.  In our case, I am the boss, and I don't want
>learn lfrd.  It is just a waste of my time.  I have been using working
>for almost 25 years now, and it works just fine thank you.  If I am going
>supervise engineers doing structural calculations for me and put my stamp
>signature on their calc's, I need them using a design method that I am
>I know that it seems like we may never switch over to lfrd, but with time
>will happen as it did with concrete.  When I went to school we were taught
>working and ultimate methods for concrete.  My first job at Bechtel we used
>ultimate strength.  At my second job, the boss had no idea now to do
>strength method, and would not allow me to use it, so I used working
stress.  At
>the present time in our office, either method is acceptable.  I tend to use
>working stress the most because I know it the best.  However, I am equally
>familiar with both methods.
>I suspect that as time goes on, lfrd will become more common, as the "boss"
>becomes more familiar with it.

I too learned in school to design steel in allowable stress design (asd) and
concrete in working stress design (wsd). I soon learned (in practice) to
design concrete in ultimate strength design (usd). However, my mentors
cautioned me early on not to forget design for service loads and the
relationship between the applied loads and states of strain. For example,
footings that have to be certain sizes for soil bearing
pressure and other practical considerations would still be in an elastic
state of strain even when the design loads are factored. I must admit that I
do not do very many (if any) structures where the economy of usd would be
warranted. Younger graduates who only learn usd do not have a feel for crack
control and other serviceability requirements that wsd addresses.

With regards to lrfd (it is lrfd, not lfrd - load and resistance factor
design), I never learned this method in school and I have never used it in
practice. Outside of the "old dog" syndrome, I always found it odd to apply
"ultimate" forces to a frame in a linear elastic distribution rather than
using plastic hinging, etc. BTW, many years ago when I served on the
seismology subcommittee of SEAOSC, a senior member of the committee reported
that, with high rise steel framed structures, only the bottom six stories
ever go inelastic. I don't remember his name, so don't ask me for
documentation on this conversation. The point is, one should correlate the
design loads with the state of strain of the design member before deciding
which analysis method is appropriate. On a concrete section, this concept is
fairly easy to visualize. Upon first loading, the section is uncracked and a
transformed section is the most appropriate method of analysis. Once the
section cracks, a cracked section elastic analysis is appropriate. When the
strain hits approximately 0.003, then plastic deformation begins and usd is

Regardless of the structure, I caution all to use the method of analysis
appropriate for the state of strain in the structural section.

Bill Allen