Effects of non-fire-rated floor/roof-to-wall connections on the fire behaviour of LSF walls

This research study focuses on light gauge steel framed (LSF) walls with non-fire rated floor/roof-to-wall connections made of steel bearers and cleats. The thermal behaviour of these LSF walls exposed to fire on one side was assessed first using a fire test of a small-scale cavity insulated LSF wall. A thermal finite element (FE) model was then developed and validated against the obtained fire test results. The thermal FE model was then modified to simulate the behaviour of full-scale LSF walls by including the effect of plasterboard joints in the walls. Using the temperatures of stud hot and cold flanges predicted by the modified thermal FE model and a sequentially coupled structural FE model developed for LSF walls without floor/roof-to-wall connections in a previous study, a modified sequentially coupled structural FE model was developed for LSF walls with floor/roof-to-wall connections made of steel bearers and cleats. FE analysis results showed that the connections with cleats located near the wall studs could lead to localised higher temperatures on stud flanges as well as localised failures in the wall studs, but without any premature collapse of the walls. For the connections with cleats located between the wall studs, temperatures and failure times of the wall studs were minimally affected. This paper presents the details of this study, its results and design recommendations for such LSF walls. • Investigated the fire behaviour of LSF walls with non-fire rated floor/roof-to-wall connections using bearers and cleats. • Fire test of a cavity-insulated LSF wall and its validated thermal FE model were used in this investigation. • Used modified thermal model and sequentially coupled structural models to determine the FRLs of full-scale LSF walls. • Localised higher stud temperatures and associated failures are observed depending on the distance between cleats and studs. • Even when cleats are near the studs, the reduction in failure times is less than 15% and is negligible in many cases.