Lateral Support … (at midspan)

As part of a recent post (here) we looked at the required size glulam beam to support a roof load (snow plus dead) of 21,000 pounds across a span of 32 feet. (Yeah, a lot of snow.) In that post the load was delivered as a concentrated load, as though from a column atop the beam, at mid-span. (See sketch.) In that framing arrangement the column was assumed to provide no lateral support for the beam (and rightly so); the beam was only supported laterally (and against rotation) at the ends.

Now let’s look at the same load, again applied mid-span, but as from girders, framed to the sides, … with the girder framing also providing lateral support (and, generally, rightly so). (See other sketch.)`Load wise’, the only thing that changes is the `buckling length’ of the beam. Now that the beam is supported laterally at mid-span, the unsupported length of the beam is 16 feet, instead of 32. Instead of potentially buckling in a single arc, end-to-end, the beam, if it buckles at all, will buckle in an `S’ shape, held in place at ends and mid-span, and buckling in-between, the curved shape reversing where held in place by the girders (S-shape not shown, sorry). Re-running our calcs with the unsupported length of 16 feet, we find that a smaller size beam, 6-3/4 x 28.5 works! … `one lam’ smaller than unsupported. I thought that by cutting the unsupported length is half (which gets `squared’ in the calcs) … I’d get an even smaller section to work. But no.

It turns out that the Volume factor, CV, takes over.