| description abstract | Motivated by the magnitude of discrepancies that are typically encountered in blind analysis contests between numerical model predictions and test data, a methodology is presented to incorporate modeling uncertainties in the assessment of capacity-designed steel moment-resisting frames (MRFs) under earthquake loading. Sources of modeling uncertainties are identified in order to define a set of variables that control the seismic response of steel MRFs. For each variable, statistical distributions that rely on experimental databases are deduced. Special attention is paid to the truncation limits to enable the generation of individual parameter samples that have an actual physical meaning. Besides strength modification factors for various steel grades, distributions are offered for the parameters that are used to model wide-flange (composite) beams, steel hollow structural section columns and damping. Both intracomponent and intercomponent interdependencies are explicitly discussed in an attempt to propose correlation coefficients that comply with the current design philosophy and construction sequence. Although the focus is strictly on collapse of capacity-designed steel MRFs, the proposed methodology can be utilized to examine less severe limit states and/or existing structures where capacity-design principles do not necessarily apply. A 4-story steel MRF, tested full-scale at the E-Defense facility to collapse, is employed as a case study to demonstrate the applicability of the proposed methodology. It is shown that although the steel MRF examined is insensitive to modeling uncertainties regarding both collapse and story mechanism prediction, local response parameters can vary considerably versus the ones observed during the test. Through a parametric investigation on the strong-column-weak-beam (SCWB) ratio, the parameters that have an impact on the associated predictions are identified. Implications for capacity design, collapse capacity and residual drift are also discussed, highlighting the benefits that higher SCWB ratios can have on seismic response. | |
