| Seismic design criteria for steel frames underwent extensive revisions and modifications after widespread damage to moment resisting connections was discovered after the 1994 Northridge earthquake, as a result of a coordinated research project funded primarily by the Federal Emergency Management Agency. Although the SAC research project created copious new knowledge about steel moment frames, much research is still needed to fill in specific gaps about seismic behavior, modeling, and design of steel frame systems. This dissertation attempts to address some of these issues and focuses on a promising type of structural system known as a Reduced Beam Section (RBS) moment-resisting frame.; Analytical models are developed for beam column elements, panel zones and RBS connections to enable realistic modeling of a RBS structural system. All the models are based on bounding surface plasticity theory and are calibrated using available test results. The beam-column model is particularly versatile and gives good results even under extreme load regimes such as in the case of steel braces subjected to reversed cyclic loading.; To investigate the behavior of RBS moment frames, 3 buildings that are 4, 8, and 16 stories tall are designed following the latest seismic design provisions. A series of nonlinear static (pushover) and dynamic analysis are performed according to provisions in FEMA-273 and FEMA-350. Both material and geometry nonlinearities are considered in the analyses, which are conducted for seismic hazard levels corresponding to 2% and 50% probabilities of exceedance in 50 years. The effects of a variety of parameters such as P-delta effect, level of damping, and higher mode effects on column axial force, panel zone distortion, and interstory drift are investigated.; Nonlinear dynamic analysis results show that RBS frames designed according to current provisions can deliver satisfactory seismic performance. Pushover results suggest that the frames will not perform well in a strong earthquake. This contradiction shows that the pushover criteria in FEMA-350 need to be refined. The effect of P-delta on frame behavior can be significant, and was observed to cause excessive demands. An evaluation of the analysis uncertainty factor in FEMA-350 shows that it is conservative for all cases considered. |
