connection plate eccentricity

huangyhg

connection plate eccentricity
i've got a problem where the increased thickness of a connection plate increases the overall eccentricity of the joint, thereby causing the failure of the other plate.
the attached should clarify what i'm trying to explain. it's the fixed base moment of the thinner plate which is theoretically increasing and failing.
is there a rational explanation why the base moment would not be what it seems, ie. 'axial load x e'?
it seems counter-intuitive that an increase in plate thickness would result in over-stress.
i would make the same assumptions you have, so i would end up thickening or stiffening the bottom plate
can you assume the axial load is not at the center of the upper plate, but is closer to the lower plate?  if the upper plate is oversized, perhaps you can assume this.
daveatkins
perhaps as the load is applied to the upper plate, and it deflects and rotates, then the load is transferred to the edge of the plate rather than remaining at the centre. of course at that stage the edge of the plate has moved and so the applied load is still off-centrw and still applies a significant moment to the lower plate.   
corus
quote:
is there a rational explanation why the base moment would not be what it seems, ie. 'axial load x e'?
are you looking for a way to justify using a smaller moment at this location?  if the plates are stiff enough to take this moment, then it wouldn't need to go to the connection.  like if the plate was really a web of a wide flange beam, then you could get by neglecting the moment.  but if you just have two plates, with relatively little stiffness in that direction, then you're looking at a moment of p*e on the connection.
i don't know if this solves your question, but try think the "load path" is like - assume loaded in center of the upper plate, then the load is transferred from the center to right side face of the upper plate, then to the welds, then from the welds to the lower plate and to the support. break the path down, show loads as you go, then sum the forces & moment. now see what you get.  
based on statics alone your fixed end moment increased 57% i'm not sure why you'd be getting 75% unless you're running a geometric nonlinear analysis and are accumulating p-delta effects.  
the eccentricity is 17.5 which is 75% greater than 10. what am i missing?   
whoops, i did punch that in wrong. statics does it.
thanks all.
i guess i was clutching at straws. if it was a single connection strengthening the lower plate would be an easy decision, but we have about 500 of these.
i think my best option is to duplicate the existing on the other side of the upper plate. this avoids removing the existing and also eliminates all eccentricity.
yeah that was just a simple statics one there.