cantilevered roof girders

huangyhg

cantilevered roof girders
i am looking at 30 year old roof.  the framing consists of joists at 6' on center, supported by wide flange girders.  the wide flange girders are cantilevered over the columns varying distances and spliced, i guess to save weight.  i've looked at it with enercalc and can't get any of the girders to work.  the splices appear to be moment splices as well, but i've seen other jobs where they are just shear splices.  is there an easy way to analyze a row of cantilevered and spliced girders.  it seems complicated to figure out what end is supporting what end (and what end is reacting on what end).  especially for 30 years ago when they didn't have computers.  they must have had an easy way to do it.      
check out our whitepaper library.
aisc manual of steel construction has some charts in it for cantilevered construction with moment coefficients.  
thirty years ago was not that long ago, it could have be done on a main-frame computer... but you are probably right.
suggest you try the easy way first - check the beams as if they all have simple supports. this could very easily be how it was done then (with a slide rule or 4-function calculator). you may find, with that assumption, the wf's work.
if i am wrong to propose this, you will find out very quickly, too.
the splices could have been used thinking that they would make things "better"... but in reality, they did not.
that type of engineering problem-solving was common before computers became available - take a complex problem that you do not have the tools to analyze cost-effectively and "transform" it into an "simplified" model that you can address. then add things, like the splices, to tie everything together. a shame that this approach it did not always give the right answer.
if they are shear-connected only, i would take the reactions from the beams in between the column as loads on the end of the cantilever.
------------o---------o------------
^       ^                    ^         ^
is that what you have? the circles represent shear connections between columns.  now you have a simply-supported middle beam and two continuous beams with an overhang and point load at the end of the overhang.
most single story frames with cantilevered girders used shear splices for the suspended span. occasionally, you might run across one with moment splices.
frames with shear splices can be analyzed statically but those with moment frames should be analyzed as continuous beams.
years ago, some engineers assumed that the splice or point of inflection was a braced point. some engineers positively braced this point while others did not. this may be part of the reason that things are not checking out.