structural steel connection design

huangyhg

structural steel connection design
i am working in north america. client/upper management hires a korean detailing company for structural steel connection. their typical shear connection detail is close to shear tab/ beam splice connection. we can use this type of connection as per aisc-asd manual if we incorporate eccentricity moment in connections(refer.pg3-123- vol.2).
i will explain the detailas follows:
beam stops at the edge of girder flange,shear plate is welded to full depth of girder like a stiffener(it is on both side of web). two plates similar to splice connection(both side of webs) make a connection between girder and beam using bolts to fulldepth of beam. (this is 100% similar to splice connection).45 mm is distance between beam bolts and face of girder flange.
koreans are using e=45mm eccentricity (for all beams) and mcon= r*45 kn-mm. my opinion is that eccentricity should be equal to the 1/2 of  flange width of girder (mconn=.5*gider flange width*r) as beam stops outside of girder flange face.
now this moment should be resolved into components and victorically added to shear to get max bolt forces.

please give me your feed back. i would be greatfull to all of you.
check out our whitepaper library.
it appears that the beam bolts are being designed for the 45mm eccentricity.
the eccentricity should be calculated from the girder cg. if you want to be a 100% accurate, then
actual eccentricity = distance of cg of beam bolt group from the girder centroid.
if the 45mm eccentricity is from beam bolt cg to the face of the girder flange, then
eccentricity = ( 45mm + 0.5*girder flange width ) * r
it depends on the the stiffeness of girder and the  connetion plate.
girder support was torsion resistance. if you consider girder length=0, it's the situation of korea's calculation.
thanks both of you. 45mm is distance between beam bolts cg and flange face of girder. e= (45+.5*flange width)is it not too conservative.
i am at home and will attach connection detail and some calculations tomorrow to discuss futher.
girder is modeleld as simple pin connections both sides. moreover w sections has very little or no torsional resistance. torsion will act on girder as mconn (as per aisc) but it would not be resisted by girder instead will be resisted by connection.  once again thanks for your time.
is there a beam framing into the plate on the other side of the girder web?  if so, the two reactions will tend to remove or at least reduce torsion on the girder.
best regards,
ba
some time beam is on both side of girder and some time not. it is a typical shear connection configration either beam is there or not.
i can imagine that torsion is there, but it is taken by connection,  not by a girder. my basic worry is this eccetricity moment(how much) and how to resolve it into connection bolts shears.
rfd23,
if the girder cannot take torsion and the beam is connected on one side only the connection must be designed for an eccentric moment of r(b/2 + 45) as noted earlier by both you and slickdeals.  
bolt shears for eccentric loads on bolt groups are readily available.
best regards,
ba
thanks. baretired an slicdeals.
please see attached connection detail and calculation.
review and discuss.

connection calculation is attached
rfd23,
i do not recognize the beam sizes you are using.  the girder is h390 x 300 with bf = 300.  does that mean a 390 mm deep   
i agree with the comments above.  the korean calculation is treating the shear plate like a splice plate, which is incorrect, because it introduces torsion into the supporting girder.  the three bolts at the supported beam should be designed for a much larger eccentricity.
daveatkins