aisc lrfd and deflection calculation

huangyhg

aisc lrfd and deflection calculation
does anyone know the rationale behind the lrfd method using service loads for deflection calculations. what i have found is that lrfd typically is the same size or a slightly smaller beam selection than asd from flexural calculations. however when checking deflection, the lrfd beam deflects more due to the smaller moment of inertia and so i go back to the asd size anyway. has anyone else come across this issue. any tips will be greatly aprreciated.
depends on a lot of things, but yes.  as far as rationale...... why would you use lrfd values to check deflections?  then you would need even bigger beams than you are getting with asd.
quote:
what i have found is that lrfd typically is the same size or a slightly smaller beam selection than asd from flexural calculations
this isn't necessarily true.  lrfd placed a higher level of load factor on loads that have higher variability, such as live loads.
so if you have a high live load condition (relative to dead load) your lrfd beam would tend to be larger than the beam determined using asd.
conversely, a low live load condition presents a smaller beam under lrfd than with asd.
deflections are checked as "actual" deflections.  in other words, you use the true loads to determine the true deflections.  lrfd load factors are a means of applying a safety factor against failure, so deflections should never be based on factored loads.

deflection limits are based on how people perceive the deflection of a floor they are walking on etc. in other waords, if a floor deflects too much people won't feel comfortable on it. the deflection limits are set so that people feel comfortable using a facility during every day use. there are no failure or life safety issues associated with failing to meet deflection limits.
the factored load cases represent extreme cases. as such, using factored loads to come up with deflections is very conservatve as they do not represent the every day use condition that deflection limits based on.
i think the main point is that for most simple spans beams, deflection will govern, so there is no point to even bother with lrfd for checking stress.
abusementpark,
respectfully, i disagree with that.  there are all sorts of conditions where strength does govern over deflection.  both strength and serviceability should be checked.

i am not saying you shouldn't check stress.  you should always check stress.  i am just saying majority of the simple span steel beams i design are governed by deflection, so whether i check stress with asd or lrfd isn't an issue.
abumesmentpark..
i agree with jae.  it is dangerous to dismiss checking strength provisions.
on a side note, i think some people get way overworked about deflection limits. if the code states a total deflection limit of l/240 governs, and you're coming up with l/230 or l/220, you are well within engineering judgment to allow that deflection. conversely, depending on your situation, the code deflection limit may not be suitable; you may require much more stringent limits for your particular purpose. i emphasize judgment and experience.
i think that amusementparks piont is that if you need to check deflections, be it for floor stiffness, facade integrity, ponding, control of secondary stresses, you are probably better using asd than lrfd which is essentially the same code, except lrfd requires you to compute stresses twice rateher than once as with asd.
the aisc spec writers in 1924 somehow knew to use 1.67 as a factor of safety against failure for an allowable bending stress. according to j.c. smith the author of an lrfd steel book he indicates up until 1989 the aisc spec writers never gave a reason for selecting 1.67. bottom line asd was and still is a great simple to use format but lrfd simply confirms it. even the connection designs are almost identical when comparing them in asd to lrfd. thanks for all the tips.
to my understanding, the lrfd is an "ultimate strength design method". the design loads are factored to reflect potential overload/underload situations, and the design philosophy is at such stage, the structure may undergo significant deformation, but will not fail. thus, using the factored loads to check deflection is an act of unnecessary waste (unrealistic and overly conservative).