Ding and Wang 2007
This paper studies joint failure due to fire with beam to column connections. There are not a great deal of studies on joint failure due to failure as it is complex. Traditionally, joints had a thickness as high as the connected members, thus the fire exposed area is low compared to the mass, thus slowing down temperature rise in the joints. This method was found, on some occurrences to cause a greater effect on the tension components of the joints, causing joint failure.
Experimental Study, Results, and Discussion
This paper includes an experimental study on fire behavior of steel beam to concrete filled tubular column frames, taking into account various types of joints. The types of joint are fin plate, end plate, reverse channel, and T-stub. Ten tests were completed, and fire exposure was maintained in eight set ups until the structure failed in the joints under tension. The fire tests were conducted in the Fire Testing Laboratory at the University of Manchester, using a furnace which is lined with ceramic fibre materials, which transfer head to the specimen. A honeycomb ceramic wall was used to ensure uniform heating. In a frame such as the frames tested, the beam would support a concrete slab, which was not used in these tests. In order to simulate the heating effect of the slab, the top flange of the beam was wrapped in a 15mm thick ceramic fibre blanket to slow the temperature rise of the top flange. The columns were restrained from lateral movement. Two transverse loads were applied to the top of the beam using two hydraulic jacks. Strain gauges were not used as the temperatures were high, however displacements and temperature distributions were measured. In the tests with the bolted T-stub connection, the beam deflection increases rapidly after 23-25 minutes of loading, before failing between 30 and 38 minutes. The connections themselves were twisted, and the bolts exhibited a large rotation. There was no sign of column failure. The two tests with the reverse channel joints exhibited similar time of significant deflection and failure. The second test with the reverse channel, which had a thicker tube thickness, failed at a much later time, of 51 minutes, where fracture was observed in the steel tube at the joint location. Both tests with the extended endplate joints experienced rapidly increasing deflection at 23 minutes, and failed within 25 minutes, and the endplates were significantly distorted. The concrete core was not damaged, and the tubes did not fail. The reactions of the steel beam can be split into three phases: slight bending and deflection due to thermal bowing, increased vertical deflection and twisting due to reduction in strength of steel, and finally, catenary action and reduced rate of deflection. After reviewing the failure of the connections, the reverse channel has the best features including construction cost, ability to carry catenary action, and high ductility.
Reference
Ding, J., and Wang, Y.-C. (2007). “Experimental study of structural fire behaviour of steel beam to concrete filled tubular column assemblies with different types of joints.” Engineering Structures 29 (12), December, pp. 3495-3502 doi:10.1016/j.engstruct.2007.08.018