I do it both ways.
For normal, simple-span beams, I let the steel book dictate the connection
design, and have a note to that effect in the Steel Notes on the drawings.
My defense of this practice is that I don't want the connections to limit
capacity when occupancy changes, etcetera. Of course, I know that
deflections often control, and increased occupancy loads may require
reinforcing to meet the deflection limits. But to me, deflection is not a
safety concern, it's aesthetics. If the next guy goofs up on a deflection
check, he may be embarrassed . . . or maybe, no one even notices. Big deal.
Your example doesn't seem very practical to me. You said, "If you have a
W8x24 GR50 spanning 5 feet and utilizing the uniform load tables that means
you are designing that connection for 39k. The beam has hardly any load on
it and a typical two bolt connection would be adequate."
A two bolt connection, single shear, 3/4" ASTM A325N is good for (2)(9.3k) =
18.6 kips. The uniform load on the beam to create that magnitude of
reaction for a 5-foot span, is 7.44 klf. Simple span moment is 23.3 ft-k.
I'm wondering why you'd use a beam whose moment capacity is 57 ft-k,
assuming Lb=0. What's wrong with a W6x15 (Fy=50), whose moment capacity is
about 25 ft-kips and the beam tables would dictate 20.5 kips reaction
capacity? The two bolt connection would yield 91% of the table-derived
requirement, and that's close enough for me. [Remember, this is
engineering, not science or accounting . . . you guessed at the loads, you
guessed at the strength of materials, and there is a factor of safety.]
In my projects, 99.38% of the beams I design are simple span. When I'm
forced to frame beams into beams, and I'm forced to do a real shear and
moment diagram, I sometimes put the reactions on the drawings. But this
happens only 0.62% of the time, and if I neglect to put the reactions on the
plan, it's OK because I definitely DID check V and R . . . it is safe.
In my opinion, if you have a gazillion connections that would benefit by
your undivided attention, then by all means do it. But if you have a
gazillion beams that are only 5' long, then I'll bet I can come up with a
better framing scheme.
John P. Riley, SE
Blue Grass, Iowa