single angle joist sea

huangyhg

single angle joist seat
i have a situation where there are joists framing into a concrete wall.  very standard situation.  we would like to use an angle for the joist bearing seat that is welded to an embedded plate into the wall.  what kind of design checks do i need to do on the joist seat?  i know i need to make sure the weld is adequate and the headed studs in the wall are adequate.  but what kind of checks do i need to do on the angle itself?  i was looking in chapter f of the specs in the steel manual, but i wasn't having much luck.  is there a check to make sure the horizontal leg of the angle doesn't buckle?  however, according to ch. f, the leg local buckling check only applies when the toe of the leg is in compression, which i don't have.  any help would be much appreciated.  thank you.

i would not worry about buckling of the leg.  i would check the angle thickness as a cantilever.  i would conservatively apply the joist reaction at or near the tip of the angle leg and verify it is ok for bending (you can check shear, but it will likely not be a problem).
when you say verify if the angle leg is okay for bending, do you mean bending about the length of the entire angle or the length of the angle leg?  i tried checking the 6" leg of the angle as a cantilever, and have the moment checked about the point of weld (see sketch) at the embed plate, and of course it failed miserably since the section modulus is so tiny.
]         
]------ )m
]
/-----/
   6"
here is what struturaleit is getting at (i think):
1.  you have a vertical load applied to the horizontal leg of the angle.  this imposes a bending moment on the leg only, trying to bend the horizontal leg downward.  bending is about the leg's horizontal neutral axis...yes, fairly small zx, sx, and ix.
2.  calculate the properties of the leg using the following "beam" properties:  beam width b = horizontal angle length tributary to your load...for joist seats i'd use 6" to 12" at most.  beam depth d = angle leg thickness.
3.  calculate the bending moment on the leg.  i would not use the wall to angle tip distance but rather i'd use half the angle leg width - in your case 6"/2 = 3".  even if the load is applied at the tip, the natural bending of the leg will move the applied reaction point closer to the wall.  moment is joist reaction r times 3".  
4.  since the "beam" is bending about the weak axis, buckling is not a concern.  use the asd or lrfd weak axis bending checks per the aisc spec.
what i have done in simular situations is put gusset plates on each side of the joist. then you have the moment transfering between the plates and the cantilevered end. treat it as a simple span between the plates and a cantilever from the other leg. then check mux/(phi*mnx)+ muy/(phi*mny)<1.0
i believe this is a logical interpretation of the code. what do others think? is what i have done valid?
jae-
that is exactly what i was getting at - thanks!
design of welded structures by blodgett has exactly the condition you are describing in it, if you would like a textbook example
excellent!  that was just the process i was using!  thanks for all the help.  it looks like i may need to add some stiffeners.  for this situation you can assume that it is laterally restrained, right?  but what if you just bolted in the vertical leg of the angle at let's say, 24" o.c. instead of the embed plate. can you assume this is laterally restrained?
the thing that you need to be concerned with in checking the attachment of the angle back to the wall is the spacing of the fastener in relation to spacing of the joists.  
if the load of the joist (spaced at 6' o.c.) requires (3) anchors then you can't just space them at 2' o.c.  you have to consider where the anchor is in relation to where the joist is.  is it spaced evenly between anchors?  is it located right on an anchor?
if you have on anchor located right at the joist then it will pick up all of the load.  
i would check the unstiffened and stiffened seats in the steel manual.  if that doesn't cover your situation i'd check as jae posted.  i'd take the joist reaction in the middle of the bearing ("x" below), using the minimum bearing minus some tolerance, like minimum - ?".  from that location, i'd take the width of angle used for the section properties as the width you'd get by drawing two lines toward the wall at an angle of 30 degrees from the longitudinal axis of the angle, adding to this the width of the seat.  some prefer to take this angle as 45 degrees.  it's up to you really.
code
            |<=  b  =>|
  ___________________________         
            \         /
  30-45 deg  \  ___  /  30-45 deg
       .......\|   |/..........
               | x |
               |   | joist seat
       ________|___|_________
agree with ucfse, but i would take 45 degrees as this more closely matches the plate yield line pattern.
also use same philosophy to check bending at bolts.
csd