welded splice

huangyhg

welded splice
i need to design a welded splice at beam to beam connection only taking gravity loads (cantilvered system). one of the beams was about 2" short. i cannot full pen the flanges as i had detailed. the owner does not want to see bolts at the splice and does not want to see the splice plate exposed.(the beams are storefront beams) is it possible to have a single splice plate welded to the webs of the beam on the non exposed sides? . has any one designed a welded splice in such fashion.  

a) replace entire beam? it is the steel contractor's problem.
b) i don't think your proposed solution will work because splices generally (to as 4100) have to be designed for a minimum load of 0.3 x member capacity. check whichever code you are using.
c) can you change to an extended fin plate connection? this might also not work aesthetically.
if it's a simple pin support the loss of flanges for the last 2" shouldn't be a problem. provide a spliced web plate designed for the shear and small maount of bending moment.
i don't think the problem is very difficult with a little bit of thought.  how large is the reaction?
best regards,
ba
the location is meant to be a simple pin support, only downward reaction of about 24k. the connected beams are w24x55.
if it's cantilevered how is it taking only shear and no moment.  also, if it's taking only shear why would you detail a full pen weld of the flanges?  have the steel fabricator cut two plates to fit the 2" gap.  prep the plates and provide full pen welds at top and bottom flanges.  remove the backer bars after the weld is complete (use a ceramic backer bar if necessary).  grind the welds smooth, then provide the shear plate on the un-exposed side as you mentioned.   
your original statement "i need to design a welded splice at beam to beam connection only taking gravity loads (cantilvered system), and the latest "the location is meant to be a simple pin support" are inconsistant and vague. please clarify, a sketch may eliminate a lot of guessing and pointless recommendation.
i assumed the op was talking about a connection between the end of a cantilevered beam and the drop-in span.  i don't know why he would weld the flanges in that case, though.
if that is the case, a plate welded to the inside of the web on each beam would carry shear and torsion.  could also fill in with a two inch plate aligned with the webs of both beams and field welded using the first plate as a backer.
a reaction of 24k is not much for a w24x55.  this would be similar to a single angle connection to a column.
  
best regards,
ba
a sketch always helps in situations like these.
ba:
no offense, you could be right on the spot. what i meant was "everybody is guessing to put the puzzle together, the resulting recommendation/comment may not necessarily to the point.  
the beam to beam connection is at the end of a cantilever. the main beam cantilevers across a column and picks up a smaller span beam. the reaction at end of smaller beam is about 24k.
hope this would help.
i will try to upload a sketch sometime later.
thanks.