Citation

Krawinkler H. Struct. Des. Tall. Spec. Build. 2006; 15(5): 515-531.

Affiliation

Department of Civil and Environmental Engineering, Terman Engineering Center, Stanford, CA 94305-4020

Copyright

(Copyright © 2006, John Wiley and Sons)

DOI

unavailable

PMID

unavailable

Abstract

Nonlinear analysis is becoming a popular tool for performance evaluation of structural systems at the life safety and collapse prevention levels. The challenge is to develop analytical tools and component models that make it possible to conduct such analysis with a sufficient degree of confidence. An exact replication of reality is desirable but not critical because of many sources of uncertainty, such as ground motion intensity and frequency content, and modeling uncertainties, which cannot be eliminated by the most accurate analysis. As summarized in this paper, there are several options to nonlinear analysis, and the simplest one that gets close to reality is usually the best one. Observation of behavior during the analysis (whether pushover or time history) is perhaps more important than a rigorous quantitative evaluation based on maximum recorded force or deformation values. In the author's opinion, the main objective of nonlinear analysis is to understand behavior rather than fulfill code-specified limit values. This paper presents arguments why this is a critical issue, even though it is understood that reality in engineering practice usually dictates priority to fulfilling code specified limits.