Report - PEER - University of California, Berkeley

frame, the influence of the floor connections was included in the analyses. However,the lateral load resistance from the gravity columns and out-of-plane bending of theprecast wall system was neglected.As demonstrated elsewhere (Pampanin et al. 2000), the beams and columns in theRUAUMOKO models were represented with the beam-column elements while tworotational springs per nodal location modeled the hybrid connections at the beam-tocolumnand column-to-footing interfaces. The use of two springs to model eachhybrid connection was to represent the moment contributions of the mild steelreinforcement and prestressing steel separately. The moment–rotation responseenvelopes of the springs were derived using the modified PRESSS analysis procedurereported by Celik and Sritharan (2004). The modified Takeda hysteresis and bi-linearelastic models were used to define the cyclic behavior of the springs representing themild steel reinforcement and post-tensioning tendons, respectively. The combinationof using two cyclic models for the precast connections enabled the hysteretic energydissipation and re-centering capability of the hybrid frames to be characterizedsatisfactorily. To account for the influence of flexural cracking, the moment of inertiafor the beam-column elements was taken as a fraction of that corresponded to theuncracked concrete gross section (I g ). Based on the test observations reported for thePRESSS building (Priestley et al. 1999), 0.6I g , I g , and 0.5I g were used for the columnsin the first story, all other columns, and beams, respectively.4. ASSESSMENT PROCEDURESeismic performance of the two hybrid frame buildings was assessed using amultiple-level performance objective that enabled examination of damage statesunder four levels of earthquake ground motion. Consistent with the Appendices G andI of the SEAOC Blue Book (Seismology Committee 1999), the four earthquake levelswere identified as EQ-1, EQ-II, EQ-III and EQ-IV (see Figure 4), which respectivelycorresponded to 33%, 50%, 100% and 150% of a design level earthquake.Using the strong segments of recorded input motions from small to largeearthquakes, earthquake ground motions compatible with the EQ-1, EQ-II, EQ-III andEQ-IV spectra shown in Figure 4 were previously developed (Sritharan et al. 1999).These motions, as a continuous sequence as shown in Figure 5, were essentially usedin the performance assessment of the hybrid frame buildings. A sufficient number oftime steps with zero accelerations was included in the input sequence shown in Figure5 to examine the free vibration response of the buildings at the end of each earthquakesegment. Because the hybrid buildings were dimensioned at 60 percent scale, the timestep and accelerations were multiplied by 0.6 and 1.67, respectively, during thedynamic analyses.Based on the performance-based seismic design concept presented in the SEAOCBlue Book, the minimum requirements chosen for the precast hybrid frame buildingswere fully operational, operational, life safety and near collapse performances whensubjected to EQ-I, EQ-II, EQ-III and EQ-VI, respectively. The maximum transient450