Analysis of progressive collapse in single-story buildings affected by local fire

Abstract: When a building is exposed to fire, it is required to remain structurally stable for a period of time. The regulations do however allow some types of localised failures within this time frame. The damage area of these failures must be contained and remain proportional to the initial triggering action and not continue into a widespread collapse, commonly referred to as a progressive collapse. In order to prevent progressive collapses, it is necessary to first identify which types of failures that could result in a progressive collapse. In a recent study (Iqbal N., Ph.D. thesis, Luleå University of Technology, 2016), single-storey steel frame buildings affected by localised fires were analysed. In the study it was identified that an initial failure in the truss’ top chord could potentially result in a progressive collapse. The reason for this is because when the top chord fails, the truss and its roof sheeting deflect and transitions into only handling catenary forces. The catenary forces present in the roof sheeting are then transferred to the adjacent trusses which therefore risks collapsing. The analysis could however not determine the possibility of progressive collapses and how factors such as truss span length affect the possibility of progressive collapses. The purpose of this thesis therefore became to analyse how span length affect the roof sheeting’s catenary forces and try to determine if a failure in the top chord could result in a progressive collapse. To answer this, finite element analyses where conducted on two different truss models with varying span lengths, i.e. 18- and 36-meter. Each model consisted of three trusses along with columns, bracings, and roof sheeting. Additionally, a hand calculation model was adopted to determine the strength of the catenary forces. From the finite element analysis, it could be seen that the adjacent trusses of the 36-meter truss model became grossly deformed. Hence indicating that a longer span length would increase the possibility of a progressive collapse. However, the hand calculation model used to calculate the strength of the catenary forces indicated that catenary forces present in the roof sheeting of the longer truss model, was relatively weak compared to the shorter truss model. The reason for this could not be determined, but some adjustments to the hand calculation model might be necessary to make it compatible with the analysed truss model. Consequently, it was impossible to determine the possibility of a progressive collapse. Additionally, during this work it was identified that other factors, such as truss model, bay length and roof sheeting thickness, could affect the possibility of progressive collapses. Hence, further work is necessary to determine the possibility of a progressive collapse.