sectional area sizing with fu per asd

huangyhg

sectional area sizing with fu per asd
anyone know how tension member sectional areas came to be sized with the ultimate strength, fu, with a factor of safety = 2, while the middle of the member's area is sized with fy with a fs = 1.5?  
area net req'd at connections   = tensile load / (fu/2)
area gross req'd btwn. conn's   = tensile load / (fy/1.5)
my guess is that the rupture stress, fu, the stress at which the connection bolts will tear thru the member ends, was just divided by 2 to get the allowable stress.  but why is the middle section based on yield stress and not ultimate as well?  (or vice-versa.) for a36, fu/2 = 29ksi; fy/1.5 = 24 ksi, so the connection should yield first since it is designed closer to fy.  is there an advantage to that?  any corrections and pointers would be greatly appreciated.
find a job or post a job opening
you are checking two different limit states - rupture and yielding.  the tension rupture limit state checks the ultimate tensile strength of the material (fu) at the net section while the tension yielding limit state checks the yield strength of the gross section at yield (fy).  the same type checks apply in lrfd as they do with asd.  the difference in safety factors is due to the fact that rupture is a brittle and in general more difficult to predict and erratic limit state as opposed to yielding thus the higher safety factor (and in lrfd the lower phi factor).  
the reason for the two different limit states is that typically the gross strength of a member would well exceed fyag due to strain hardening - however the large elongations induced by the yielding through the entire section would probably render the member ineffective for its intended purpose thus "failing" it (though not actually falling down).  the actual member tearing mode of rupture will occur along the plane of least resistance (at the bolt holes) and requires the member to actually break thus the use of the ultimate tension stress fu.