SAC Phase 1 Analytical Studies of Building Performance
Project Title:
Evaluation of Seismic Performance of an 11-story Steel Moment Frame Building
During the 1994 Northridge Earthquake
Sub-contractors:
Farzad Naeim, John A. Martin and Associates, Inc.
Roger DiJulio, Jr., John A. Martin and Associates, Inc.
Kalman Benuska, John A. Martin and Associates, Inc.
Andrei M. Reinhorn, State University of New York at Buffalo
Chen Li, State University of New York at Buffalo
Project Summary:
This report presents results of performance analysis studies on an 11-story steel moment
frame building located in west Los Angeles which was damaged during the January 17, 1994
Northridge earthquake. Extensive field inspections of damage were performed prior to the start
of this investigation. These inspections/tests revealed some degree of damage in more than
28% of the moment frame joints of the building
Extensive two and three-dimensional, linear and nonlinear, static and dynamic analyses of the
building are conducted as a part of this study to investigate and correlate observed damage with
various elastic and inelastic analytical damage predictors. Statistical correlation of
observed damage patterns with various analytical damage indicators are presented.
Elastic damage indicators studied include beam and column demand/capacity ratios (DCR)
calculated based on Load and Resistance Factor Design (LRFD) formulations. Inelastic damage
indicators considered include beam, column, story and building damage indices which are based
on the calculated amount of permanent deformation (rotation, curvature) and levels of dissipated
energy. The results include the following important observations:
The damage observed was much more extensive than that predicted by analyses.
The spatial distribution of incipient cracks suggests that some incipient cracks probably
existed before the Northridge earthquake occurred.
The average beam DCR at damaged connections is 20% higher than average DCR for all
connections (damaged and undamaged). This ratio for columns is 33%. However, there is no
direct correlation between increased DCR and actual damage.
Elastic beam stresses correlate better than other analytical indicators with the severity
of the observed damage; the inelastic column damage index is the second best.
Inelastic beam damage indices are very useful in predicting damage, although they miss
many of the cases of observed damage. The inelastic column damage indices are less helpful
compared to beam damage indices. This is mainly due to the fact that the two-dimensional
inelastic analysis does not consider the three-dimensional state of stress existing in many
of the damage columns.
