Camber in Open-Web Steel Bar Joists for Roof Framing?

[DCEngr1]

Engineers . . .

I have a load-bearing concrete-block industrial building built in 1935 that I have examined. The roof is framed with open-web steel bar joists; 50 foot span, 5-foot on-center spacings. A fire occurred in a machine inside the building. The direct-destructive fire damage is limited to the machine and the roof framing - joists, roof deck and roofing materials.

Three bar joists were directly over the fire origin. I measured the elevations of 6 joists at the distance from bearings to the center of the fire origin from the ends of each joist; comparing the elevations to the bottom chord at the fire origin to the elevations at the bearing points. At the two joists closest to the fire origin, the trusses were flat. Measurements away from the fire origin - extending from the closest joist to joists farther away, the joists extending from one side of the origin appeared to have a rise of the bottom chord of 0.125, 0.3125, and 0.4 inches. Visually, the joists and their components do not appear to have sustained any deformation; though supported ribbed metal roof deck had deformed between 4 of the joists.

Is it typical that steel bar joists for roof framing were, or are now, cambered as part of the design/mfgr.? If you are aware of good information on this question, and/or can direct me to current pertinent information regarding analysis of steel bar joists for reuse/replacement following a fire event, please assist me with links, etc. Again, this structure was built in 1935.

Thank you for your help.

Dana

 

[Struct123ure]

Picture with some comments would be very helpful.
Short answer is yes, you could have cambered joists. I usually camber no in 1/4" intervals becuase there is a limit to what is measurable on site, but the numbers you gave "0.125, 0.3125, and 0.4 inches" seem a little weird, I would expect them to be consistent.

 

[jayrod12]

I think his answers kind of do make sense. As he measures the centre-joist camber, as he gets further away from the heat affected area, more camber remains. I bet it would level out at something like 0.4" or 0.5" pretty quickly.

Our joists are always specified to be cambered for a minimum amount.

 

[JLNJ]

Joists are manufactured with default camber based on span (unless otherwise specified). A 50' span will get you about 1" of factory camber.

Current joists, that is. I'm pretty sure there was no uniform spec in the early 1930s.

EDIT: I see that the SJI spec goes all the way back to the late 1920s. No reference to camber that I can see and the joists tended to be shallow and relatively short-spanning.

 

[JAE]

I found a 1930 Macomber joist document that didn't directly specify camber but did have a reference to the AISC specification saying the design would conform to it.

The early AISC specs didn't require camber, but my recollection is that camber was just something everyone knew to do in the fabrication process. There were AISC tables showing how to determine camber based on span/depth and loading levels.

I agree with jayrod here that the measured numbers do make sense - fire heated bottom chord more than top chord and the resulting expansion dropped the roof joists a bit.