member suporting a monorail 9bolted connection0

huangyhg

member suporting a monorail (bolted connection)
     i have this situation. steel fy=50. a pretty long wideflange (w) simply supported, and in the middle, undernith, a monorail (s shape) passing perpendicular to the supporting w beam - bolted to the wideflange! top flange of the mr is against the bottom flange of the w beam. at the intersection i have 2 full depth stiffeners on the w beam and a total of 4 bolts connecting the flanges of the two members (one in each quadrant formed by the web of w and the stiffeners).
    in the bottom flange of the w member i get a general tension stress due to the global bending of the member and also local bending stresses (the load travels from the bolt towards the stiffener on a tributary area and i have the flange behaving as a plate in weak axis bending). both these stresses are having the same direction -> along the member so i can just add them. the results : top fiber of the bottom flange is at 42 ksi (tension), bottom fiber of the bottom flange about 0 ksi.
    i'm not interested in finding a way to reduce the 42 ksi, i'm rather interested in finding the documentation that would say this stress is too much or it is acceptable in such and such situations.

    note: the bending stresses in the w member from the
          global bending are ok.
          the local bending of the flange alone is ok -          (stresses are less than  0.75 fy)
    thank you,
         tibby.
before i pronounced judgement on this situation i would be interested in your assumptions regarding the plate bending.  this is a case where you can pretty much tailor your results based on assumptions.  for example, for your span, did you go the the center line of the stiffener, the edge line of the stiffener, or the edge of the fillet weld?  did you use a beam type analysis or a plate spanning in two directions (to the stiffener and to the web)?
my guess is that you could probably play around with your analysis with legitimate methods and reduce your stresses to something managable.
jedclampett,
i considered the load from one bolt to go in both directions meaning towards the stiffener and towards the web. i have equal distances from the center of bolt to the tip of weld (at stiffener) and to the radius (k1) at web so... i'm assuming the load from 1 bolt is almost equally divided in 2. half towards the web, half towards the stiffener. the problem i stated originally was considering already these conditions.
  thank you.
     tibi.
there is still some conservatism in the analysis.  the flange is somewhat supported in the third direction (in the opposite direction to the stiffener) and in how much of the flange is considered effective along the web (you could assume more).
saying that, i'd add another stiffener.  if you absolutely can't, i'd justify it as follows:
you're in the tension region of the flange.  you're below yield.  even if you exceeded yield, it would only be at the extreme outer fiber.  
without completely understanding your design, i found that o.blodgett "design of welded structures" to be helpfull and where there is some analysis on reinforcing gussets welded to webs and stress distributions around that area. there is some consideration of fatigue to take into account, if you weld the gusset to the tension flange. also i assume you considered the supporting w beam for top flange buckling based on code design limits?.
dooron,
i do have blodgett, it is a great source of info,  but i didn't find what i was looking for in there (for this particular problem).
jedclampett,
i know that adding a stiffener would improve the situation, or solve it, my problem was to find some documentation that would cover this situation ... where the stresses are going over any allowable limits. it's true the average stress on the flange is just the stress due to the global bending of the member (and that was within allowable limits).
      thank you both for your thoughts.
      regards,
          tibi