Among the scientific domains which are used for Fire Dynamics prediction, FLuid Mechanics plays a special role, because it is the interface among very different scientific domains, propagation the alea towards people and goods. Moreover, this branch of fluid mechanics is not simple at all: the flow encompass large density variations, and as a consequence, models are used well beyond their validation domain. This may lead to large uncertainties. The present document aims at showing these challenges, and proposing a review of recent progress and research perspectives for Fluid Mechanics in the context of building fires.
