| Abstract: | In fires in large compartments like enclosed car parks, airport terminals and industrial halls, the uniform distribution of gas temperature of post-flashover stages are unlikely to occur; in these cases, the thermal actions of a localised fire must be taken into account. In order to design steel structures for a localised fire, very detailed data concerning the development of temperatures in steel is required. EN 1991-1-2 presents a simplified model for calculating the temperatures in ceiling slabs and in the beams that may support such slabs; however, no simplified calculation model for the heat transfer in vertical elements, such as columns, is yet available. There is a need for more experimental data on real scale structures exposed to localised fires. A research project on the evaluation of temperatures in steel columns exposed to localised fires was carried out at the University of Coimbra. Full-scale natural fire tests were used to test columns, instead of conducting the usual furnace tests. This paper presents and discusses the results of the experimental tests on unprotected hollow steel columns exposed to localised fires, each of them simulating a distinct fire scenario according to different fire loads, positions and ventilation conditions. During the fire tests, real measurements showed flame heights and burning times different to those preliminarily estimated: flame heights had been conservatively predicted; while, the duration of the burning had been significantly underestimated. |
