Robustness; impact; design guidelines; recommendations; alternative load path; dynamic effect; analytical approach; loss of a column; numerical approach; membrane effects; membrane forces; caternary action
Abstract :
[en] The present document details the design guidelines which emanates from the RobustImpact project. The reasoning used is depicted below in Figure 1-1. The first step is to define the impact scenario and its key parameters. Then the impact can be characterized as a hard impact or as a soft impact. The first option assumes that the kinetic energy of the impactor is absorbed by the impactor itself while the second option assumes that the kinetic energy of the impactor is mainly absorbed by the column, thus leading to plastic deformations and the activation of geometrical non-linear effects in the column. If the soft impact is considered, the buckling resistance of the column must be examined in order to evaluate if the impacted column shows sufficient residual resistance after the impact. If this is not the case, the structure will resist if it is able to sustain the load initially carried by the impacted column by redistributing it in the damaged structure, i.e. if it is able to activate alternative load path. The resistance brought by other structural elements will then be evaluated. If the impacted column is a peripheral column, the yield line theory is used to evaluate the resistance of both slab and beams. In case of a central column, Bailey’s method evaluating the slab resistance taking into account the membrane effect is used to determine the contribution of the slab. The contribution of the beams is then quantified by evaluating the corresponding P-u curve. All the methods have been validated using FEM softwares. Since the project concerns structures whose columns are impacted by vehicles, the accidental load combination must be considered. In addition, the yield strength of the steel should be considered as 𝑓𝑦,𝑑𝑦𝑛= 𝜇.𝑓𝑦> 𝑓𝑦 . Indeed was proved than the steel mechanical properties are better under dynamic loading. Nevertheless, there is for the moment no available quantification for the majoring coefficient that should apply to 𝑓𝑦 : this could be done realizing tests.
