Computational modeling of cold-formed steel: characterizing geometric imperfections and residual stresses

Thin-walled, cold-formed steel members exhibit a complicated post-buckling regime that is difficult to predict. Today, advanced computational modeling supplements experimental investigation. Accuracy of computational models relies significantly on the characterization of selected inputs. No consensus exists on distributions or magnitudes to be used for modeling geometric imperfections and for modeling residual stresses of cold-formed steel members. In order to provide additional information existing data is collected and analyzed and new experiments performed. Simple rules of thumb and probabilistic concepts are advanced for characterization of both quantities. The importance of the modeling assumptions are shown in the examples. The ideas are summarized in a preliminary set of guidelines for computational modeling of imperfections and residual stresses.