Experimental and analytical investigations on the response of structural building frames further to a column loss

Recent events such as natural catastrophes or terrorism attacks have highlighted the necessity to ensure the structural integrity of buildings under exceptional events. Design requirements are proposed in some codes but are generally not satisfactory. In particular, it is not demonstrated that, even if these requirements are respected, a structure subjected to an exceptional event will really behave properly.
A European RFCS project called “Robust structures by joint ductility” has been set up in 2004, for three years, with the aim to provide requirements and practical guidelines so as to ensure the structural integrity of steel and composite structures under exceptional events through an appropriate robustness.
The investigations performed at the University of Liège, as part of this European project, are mainly dedicated to the exceptional event “loss of a column in a steel or steel-concrete composite building frame”; the main objective is to develop a simplified analytical procedure to predict the frame response further to a column loss. The development of this simplified procedure is detailed in two complementary PhD theses: the thesis of Demonceau J.-F. and the thesis of Luu N.N.H. Present paper describes experimental and analytical studies carried out in [Demonceau, 2008]. In particular, a simplified analytical procedure for the prediction of the global frame response when significant membrane forces develop further to a column loss will be described; it allows: (i) to predict the development of the catenary action in a frame with joints subjected to combined bending moment and tension loads and (ii) to compute the requested rotation capacity at the joint level according to the loads applied on the frame.