Investigation of fire-induced collapse scenarios for a steel high-rise building

The paper deals with the problem of understanding and evaluating the structural response of steel buildings to fire and outlines a general framework for the structural fire safety design of high-rise building. Among all building typology, the fire design of high-rise buildings is particularly challenging with respect to both non-structural and structural design aspects: the enhanced design difficulties in providing i) a safe and prompt vertical evacuation of the building and ii) an effective vertical compartmentalization for avoiding vertical fire spread, refer both to non-structural aspects (architectural design choices and active measures) and won’t be investigated in detail in this paper; the paper focus instead on structural design aspects, with specific reference to the iii) enhanced susceptibility of high-rise buildings to disproportionate collapse, due to the significant vertical elevation of the structural system and to the complex and often untraditional design. With reference to fire-induced collapses, a particular dangerous situation is represented by the spread of failures to elements not directly involved in the fire, i.e. element that due to their location or because of greater insulation have still a relatively low temperature at the time of failure: in this respect, the example of a high rise building is presented in the paper, where, depending on the fire scenarios considered and as different beam-column stiffness ratio, fire damages can remain localized to the heated zone or involve other elements in the failure. Disproportionate damages induced by fire can therefore be avoided with a robust design of the structural system. This requires however a study of the response to fire of the structure as a whole, which is a quite difficult task, especially for complex structures such as high-rise buildings. Aim of the paper is thus to exemplify a method for performing this kind of investigations and to outline problematic aspects in the modeling and in the interpretation of the results.