Evaluating existing roof structure
Evaluating existing roof structure
(OP)
I am evaluating an existing structural steel roof framing system for new point loads. The building is from the 50s and the framing consists of a continuous three-span girder over columns, with secondary beams spanning between girders. The roof deck is precast concrete on a wide-ribbed 3” deep steel deck – never seen precast concrete come with a steel deck before, but maybe this was a common system back in the day? The existing drawings show all member sizes but does not show the original roof design loads. I am going to have new point loads acting at the underside of the girder.
I have estimated the dead loads based on the existing drawings and determined the snow loads based on the latest edition of the code (likely lower than the original design snow loads). I performed a code check of the existing roof based on these loads and I am finding that the existing roof does not check out. It all comes down to the unbraced length in negative flexure that I have used in my calculations. I am taking the full span as the unbraced length and calculating the omega 2 factor (Canadian code, I believe the US equivalent is the Cb factor). This is the correct approach – but back in the 50s I am sure they used inflection points as brace points hence why this roof is not checking out. I am also using lower yield strength prevalent for the 50s.
How would I go about adequately bracing the girder bottom flange?
Would it be in the form of adding a knee-brace from the secondary roof beams to the girder bottom flange and design it for 2% of the compression flange force? Is that all it takes to bring my girder from 180% overstress to PASS?
Locating my knee-bracing to avoid existing services will be a challenge; I may only be able to place these knee-braces at regions of positive flexure – say at mid-span, yet this would reduce my bottom flange unbraced length by half for negative flexure? Seems strange.
I have estimated the dead loads based on the existing drawings and determined the snow loads based on the latest edition of the code (likely lower than the original design snow loads). I performed a code check of the existing roof based on these loads and I am finding that the existing roof does not check out. It all comes down to the unbraced length in negative flexure that I have used in my calculations. I am taking the full span as the unbraced length and calculating the omega 2 factor (Canadian code, I believe the US equivalent is the Cb factor). This is the correct approach – but back in the 50s I am sure they used inflection points as brace points hence why this roof is not checking out. I am also using lower yield strength prevalent for the 50s.
How would I go about adequately bracing the girder bottom flange?
Would it be in the form of adding a knee-brace from the secondary roof beams to the girder bottom flange and design it for 2% of the compression flange force? Is that all it takes to bring my girder from 180% overstress to PASS?
Locating my knee-bracing to avoid existing services will be a challenge; I may only be able to place these knee-braces at regions of positive flexure – say at mid-span, yet this would reduce my bottom flange unbraced length by half for negative flexure? Seems strange.






RE: Evaluating existing roof structure
RE: Evaluating existing roof structure
RE: Evaluating existing roof structure
Is there any chance that the secondary beams are already bracing the bottom flange by way of torsional beam restraint? If the tops of the girders and infill beams are at the same elevation and the infill beams are at least 60% of the depth of the girders, I would say so for most conventional shear connections. Of course, if your secondary beams run over top of the girders then it's a different ball game.
Pretty much. Nowadays, I prefer the bracing force/stiffness in the American steel manual for this kind of thing. They did a killer job of it other than readability, which is atrocious.
If you consider the bottom half of the beam as a column in compression, then cutting the unbraced span in half would quadruple the axial capacity for elastic buckling.
Yeah, I know what you mean. I think that it feels strange because we're taught to think of lateral torsional buckling (LTB) as the buckling of the compression chord. For most purposes, that's an apt analogy. Not always though. If you more accurately envision LTB as rotation about a point somewhere in space aligned with the vertical axis of the beam, it be comes apparent that the deformation of the entire beam span between brace points is involved in buckling, not just the parts of the beam where the flange being considered is in compression.
I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.