I heard a speaker on LRFD once suggest that using that method for steel
design, it is best to size all members based on serviceability requirements
and then "check" the structure for stress compliance using LRFD. This
certainly implies that lighter weight steel members are to be expected using
LRFD than using ASD. One concern I have is that the serviceability
requirements are not as clearly defined as I would like, and thus I am more
likely to have unexpected problems due to flexibility of the structure when
using LRFD in lieu of ASD. Thus I prefer the "safety" of sticking with ASD.
I mostly design structures which house or support equipment, and in general
I prefer a stiffer structure. Thus I prefer ASD because I know it will
result in a stiffer and safe structure. (And yes, it is true that I learned
ASD and prefer not to "re-learn" steel design. However, I learned "ultimate
strength design" of concrete in school and learned "working stress design"
of concrete on the job, so I would learn to use LRFD if I felt it would be
I also prefer ASD for steel vs "ultimate strength" for concrete because of
material differences. Reinforced concrete does not behave in as linear a
fashion as structural steel and thus ultimate strength design does a better
job of accounting for material properties. With monolithic reinforced
concrete, my "connections" are designed consistently with my members (i.e.
using embedded reinforcing steel) vs an anomaly in steel connections (bolted
or welded). I feel comfortable that concrete can redistribute moments
fairly readily but not so comfortable that a steel braced or pinned frame
can adjust to yielding members. Thus I don't have the comfort level with
LRFD in steel that I have with concrete.