SAC Phase 1 Analytical Studies of Building Performance
Performance of a 13-Story Steel Moment-Resisting Frame Damaged in the 1994 Northridge Earthquake
Chia-Ming Uang, University of California, San Diego
Qi-Song Yu, University of California, San Diego
Ali Sadre, Esgil Corporation
David Bonowitz, Nabih Youssef and Associates
Nabih Youssef, Nabih Youssef and Associates
This report summarizes the results from case study of a thirteen-story steel moment frame
(SMF) building, which sustained extensive damage to its welded connections during the 1994
Northridge earthquake. The project is located approximately five kilometers from the epicenter,
and the ground motion records in the basement, as well as the sixth floor and the twelfth floor
were made available. The building had been surveyed and data were collected on the fractured
joints. Original design drawings were also readily accessible. The structure has no pronouned
irregularities or discontinuities in its load paths. The subject building is a prime candidate
for this research for all of the above reasons
The primary objective was to verify the accuracy of present analytical tools in successfully
predicting the pattern and severity of the documented connection failures. Absence of any notable
torsional or stiffness irregularities in this building simplified the dynamic analyses, which along
with access to recorded ground motions allowed the model to be calibrated with high precision.
Statistical reports on the failed connections indicate that the building suffered more severe
damage in the reference north-south (N-S) direction than the reference east-west (E-W) direction.
Results from elastic time-history analyses correlated well with the measured response in the
E-W direction, but not so will in the N-S direction. Correlation began to deteriorate as the
damage became more extensive and the elastic model was unable to simulate the fractures to the
welded moment connections. The demand/capacity ratios (DCRs) of the beams computed from the
response spectrum analyses were as high as 1.18 in the N-S direction, which was nearly double the
amount in the E-W direction. Beams with higher DCRs were concentrated between the second and
seventh floors, which coincided well with the observed damage on the west side of the building.
In the E-W direction, the maximum story drift ratio was calculated at 1% with a roof drift ratio
of 0.55%. These values suggested that the ground motion did not force the building into the
post-yield state in the E-W direction and connections primarily fractured will before the plastic
capacity of the beams were developed.
Modeling the panel zone was among the most significant adjustments for the inelastic analyses.
The design base shear was calculated for the building per 1994 Uniform Building Code (UBC).
Structural overstrength was then estimated to be 2.6 at ultimate level based on an inelastic static
push-over analysis. Inelastic time-history analyses improved the correlation in the N-S direction.
It also indicated that a significant number of panel zones had yielded. Thus, the authors
speculated that panel zone must have been a major source of energy dissipation during the
From this limited study, on a regular structure, it appears that current analytical and
modeling tools for both elastic as well as inelastic analyses are quite reliable. They can be
instrumental in predicting with a reasonable degree of accuracy the intensity and pattern of
damage expected during severe seismic events.