Quantification of variability in simulated seismic performance of RC wall buildings

Abstract

This paper quantifies the variability in the seismic performance of slender reinforced concrete (RC) wall buildings simulated using different fiber based nonlinear models. The models use the same material constitutive relationships and other analysis inputs (e.g., damping) so that the quantified variability in the building performance is caused by the numerical modeling approach rather than user-selected parameters. These models were first validated against the measured shake-table test behavior of a 7-story RC wall building. Then, nonlinear dynamic analyses of the 7-story wall test specimen and three RC wall archetype buildings of 4-, 8-, and 12-stories were conducted to quantify the variability in the resulting seismic performance assessment parameters and damage fragility curves. The variability in the numerical model responses was evaluated using ground motion suites corresponding to three hazard levels with 50%, 10%, and 2% probability of exceedance in 50 years, while the damage fragility curves were obtained following the FEMA P695 methodology. The variability was quantified based on global response parameters (interstory drift ratio and roof drift ratio) and local response parameters (plastic hinge rotation and curvature at the critical length), since these parameters are commonly used in seismic performance assessment of RC buildings. Based on the results, it is strongly recommended that the performance assessment of slender RC wall buildings be done using global response parameters rather than local response parameters, since the variability of the predicted response among the nonlinear models was significantly smaller for global response parameters.