A Comparison Study of Modeling Methods of Simple Steel Connections Subject to Column Removal Scenarios

When investigating the global behavior of steel and composite structures, numerical modeling is preferred to physical testing as the latter is much more costly and time-consuming. To date, various models have been used to simulate the behavior of steel and composite connections subjected to catenary action. This is a key attribute to structural robustness under column removal scenarios. The main difference of these methods lies in the type of elements chosen to model the connections, including three-dimensional solid elements, component-based model and plastic hinge model. Although each model has its own merit and strength, not all of them are suitable for building up the whole structural model for analysis of global behavior, which is required for assessing the robustness of structures against progressive collapse. To evaluate the pros and cons of different methods for whole structural model, a systematic comparison study of connection elements used by these methods, i.e. solid, spring (or fiber) and hinge elements, is presented. All three types of models are validated by physical test data and they show good agreement with test data. Based on accuracy and time, some helpful suggestions on selection of modeling method are provided.