|
|
||
|
|
|
|
For example, local buckling at the end of a thin-walled floor joist or at the ends of a number of adjacent floor joists tends to convert the effective boundary condition there from clamped to a kind of simple support. Under simply-supported end conditions each floor joist wil sag under the dead weight of the stuff on the floor more than it did under the initial clamped condition. This additional sagging of the floor will generate more tension where the joists are attached by bolts to adjacent columns. This additional tension (“catenary action”) may cause some of the bolts to fail in tension, with possible consequent separation of the floor joists from the columns. After bolt failure, the columns, now laterally unsupported along longer vertical lengths than they were designed for, may buckle, causing the detached floor to crash down upon the floor below it, resulting in similar failure of the analogous bolts on that lower floor, and so on. (“Pancaking” of the second, third and further floors may be caused by shearing of the bolts rather than by bolt tension due to beam “catenary action”.)
For a discussion of beam “catenary action” see the paper:
Jennifer Righman McConnell, Thomas Cotter and Tiera Rollins, “Finite element analysis assessing partial catenary action in steel beams”, Journal of Constructional Steel Research, Vol. 109, pp 1-12, June 2015
In the literature on steel frame analysis “catenary action” is generally caracterized as helping prevent progressive collapse because catenary action decreases the maximum vertical flexural deflection of the beam. A harmful effect of “catenary action” is the increased chance that end bolts may fail under tension.
The image presented here is from:
https://fire-research.group.shef.ac.uk/current.html
2012- Guan Quan (Research student, funded by EPSRC & China Scholarship Council)
Shear buckling in the vicinity of beam-column connections in fire
The Cardington composite frame fire tests indicated that shear buckling of beams, as well as beam bottom flange buckling, in the vicinity of the beam-column joints, is very prevalent under fire conditions. These phenomena can have significant effects on both the force redistribution between the bolt rows at the face of the column and beam deflections at high temperatures.
This research is currently aimed at investigating the local buckling behaviour in the vicinity of beam-column joints at elevated temperature. The behaviour being studied includes shear buckling of the beam, local buckling of the bottom flange of the beam and shear buckling of the column. Theoretical models will be created for the three buckling zones at elevated temperature. On this basis, three corresponding component-based models will be created and implemented in the software Vulcan, in order to study their influences on the behaviour of the whole structure, and particularly on its progressive collapse in fire.
For more on progressive collapse see the collection of images in the “Architecture” gallery about the 2001 collapse of the World Trade Center Buildings 1, 2 and 7 in New York City.
Page 220 / 410