emperature loads open steel structure

huangyhg

temperature loads open steel structure
i am looking at an industrial steel structure. it is relatively small - 3 levels of steel on a concrete platform with a footprint of 12m x 6m. i have modeled it and want to look at the thermal effects. when i put in a 60 degree temp change i get huge forces. i wasn't expecting those kind of forces and wonder if there isn't another problem somewhere in the model. i modeled the braced open grid flooring with nominal horizontal crossed bracing across the bays. then vertically i have downward pointing v bracing from the top level down and upward pointing v bracing from the bottom level up meeting in the middle.
first of all for somewhere with a very steady temperate climate is 60 degrees too much swing - what do others use? secondly - do you expect large temperature change forces or will they dissipate in movement in the connections and how do you account for that?
check out our whitepaper library.
i've noticed the same thing in the past doing industrial structures. you have to be cautious about apply thermal loads to the portions of the structure close to supports.
thermal forces are "self limiting" forces.  by that i mean that if the structure is allowed to deflect that will limit the magnitude of load that can develop.   
since (in an fem solution) the supports are considered infinitely rigid, the amount of restraint at these locations dramatically increases the forces that results from your temperature loads.
if you replace these "infinitely rigid supports" with some less rigid supports you might be suprised by how small of a deflection it takes to dramatically reduce the forces.
you might even be able to rationalize the force away by saying that you've got enough slop in the bolts and or flexibility in the support pedestals and such to easily accomodate that movement.  
josh

thanks. i would love to rationalise the force away but i would also like to have some sound basis for doing so. any good literature on this you know of?. then we like to have slop in the bolts to accommodate thermal movement but we don't want it for deflection under lateral loads. in fact my support points on a concrete plinth are going to be pretty rigid.
perhaps rearranging the bracing can help. i think k bracing would be better than x bracing because deflection of the beam that the k bracing frames into would relieve some of those forces.