Search

The US consensus standard for seismic evaluation and retrofit of existing buildings, ASCE/SEI 41, establishes provisions for seismic analysis procedures that vary in complexity and fidelity. Although ASCE/SEI 41 provides detailed nonlinear dynamic procedures, most engineers rely on simpler methods to evaluate building seismic performance and retrofit, particularly the ASCE/SEI 41 linear procedures and, more recently, the FEMA P-2018 methodology for evaluating collapse potential. Under ideal conditions, these procedures identify similar structural deficiencies. However, evaluation outcomes in practice may differ due to the complexity of real building response, approximations used in modeling and analysis, and level of intentional conservativism that reflects the limitations of the procedures. To quantify these differences, this study considers six reinforced concrete buildings that sustained damage in real earthquakes or in shake table tests and compares the performance assessed by the ASCE/SEI 41 linear and nonlinear dynamic procedures, as well as the FEMA P-2018 seismic evaluation methodology. The results show that for these highly damaged buildings, the overall performance level estimated from the ASCE/SEI 41 linear procedures is consistent with observed damage. In general, the procedures also correctly identify the story with the most damage and the component failure mode. However, the ASCE/SEI 41 linear procedure generally underpredicts drift response and greatly overpredicts peak floor accelerations. Though these are not directly used to evaluate structural performance, they are related to component deformation and force demands, respectively. Moreover, the linear procedures predict damage in components that would be precluded by yielding or failure of other components in the load path. Results from the FEMA P-2018 methodology for the six buildings provide more distinction between buildings than the ASCE/SEI 41 Collapse Prevention performance level. The results also suggest the FEMA P-2018 limit-state mechanism analysis can provide supplemental information to support and improve the ASCE/SEI 41 linear procedures. [ABSTRACT FROM AUTHOR]

ASCE/SEI 41 is the consensus US standard for the seismic evaluation and retrofit of existing buildings. Although the performance-based engineering standard is based on decades of research and has been significantly vetted by ASCE and other committees, it is unclear how well the evaluations capture the seismic response of real building systems. This article examines six, primarily nonductile, reinforced concrete buildings, including four damaged in earthquakes and two experimental structures tested and damaged on shake tables, to compare the measured response and observed damage to simulated outcomes produced following ASCE/SEI 41 nonlinear dynamic procedure. The results show that the simulations are generally able to capture the story mechanism and peak transient story drift demands at the critical story (predicted values are typically within ±20% of the measured values). However, drifts at non-critical stories and floor accelerations at all stories show greater error relative to the measured responses. At the component level, the simulations, in most cases, correctly identify the location(s) of the critical component(s) and the failure mode (e.g. flexure vs shear). However, the extent of the damage is overestimated in some cases. These results form the basis for recommendations for column, beam, and wall modeling procedures that can be used to improve ASCE/SEI 41. [ABSTRACT FROM AUTHOR]