 |
SAC Publications
Report No. SAC 95-04
|
Link Here
for the Reference Library and Downloadable Documents |
Analytical and Field Investigations of Buildings Affected by the
Northridge Earthquake of January 17, 1994, by A. Alali, J. Anderson,
J. Beck., K. Benuska, D. Bonowitz, R. DiJulio, J. Dunlea, T. Eimani, M.
Englehardt, F. Filippou, G. Hart, L. Ho, S. Huang, A. Husain, A. Jain,
J. Kariotis, H. Kim, K. Kim, H. Krawinkler, C. Lee, C. Li, R. Lobo, B.
May, F. Naeim, T. Paret, D. Polidori, A. Reinhorn, T. Sabol, A. Sadre,
K. Sasaki, J. Stewart, C. Thiel, C. Uang, J. Uzarski, M. Vanik, M. VanWinkler,
N. Youssef, and Q. Yu, December, 1995.
Executive Summary
This volume includes the technical reports documenting detailed structural
analyses of nine steel moment-resisting frame buildings and ambient vibration
studies of three of them. All buildings were affected by the January 17,
1994 Northridge Earthquake. These studies comprise Task 3.1 and 3.2 of
Phase I of the SAC Steel Project. The goals of this task were to identify
specific causes of the frame damage, to assess the adequacy of available
analytical methods to predict the severity and distribution of the damage,
to identify conditions under which more severe damage might occur, and
to identify small amplitude periods and mode shapes of lower modes of vibration
of structures.
A solicitation for proposals to perform these detailed analyses was
circulated during December, 1994. Over thirty proposals were received for
this task, which included more than twenty buildings of interest. Nine
Subcontractors were selected, and nine buildings were included in this
study. Note that in the majority of cases, the Subcontractors consisted
of teams of practicing engineers and university researchers. Included were
buildings that addressed a wide variety of parameters and issues, including
the following:
-
The buildings ranged in height from 2 to 17 stories.
-
The complexity of the different buildings ranged from simple rectangular
plans without any irregularities to highly irregular, complex configurations
which included participation by other structural elements (masonry walls,
e.g.) at lower levels.
-
The majority of the buildings were located in the San Fernando Valley.
Buildings located in West Los Angeles and Santa Monica were also analyzed.
-
Damage to the different buildings ranged from severe to none at all.
-
Varying amounts of frame redundancy were incorporated in some buildings.
-
Two of the subcontractors analyzed pairs of buildings with very different
levels of damage despite locations on the same site; one at the California
State University at Northridge campus, and the other at a Woodland Hills
medical center.
-
Two independent analyses of one building were performed.
-
One building in the survey although structurally complete, did not have
interior nonstructural elements and finishes in place, effectively reducing
the building mass and damping.
-
Three of the analytically studied buildings have strong-motion instruments.
One had only a basement instrument. One only had a roof instrument. The
third had a series of instruments distributed over the height of the building.
The general capabilities of four different analytical procedures were assessed
as to their ability to predict damage states and susceptibility to increased
aftershock damage. These four procedures included the following:
-
Equivalent static elastic methods
-
Elastic dynamic analysis methods, both response spectrum and time history
-
Static nonlinear analyses ("push-over")
-
Nonlinear dynamic analysis methods
To ensure compatibility of the analytical results obtained by different
investigators and to provide a direct means of comparing between different
buildings within this sub-Task, a standard set of modeling and analysis
standards and a common set of ground motion characteristics (response and
time histories) were established for use by all of the analytical investigation
subcontractors. These so-called "baseline" analysis procedures were intended
to represent simple analytical approaches that would be commonly used in
standard building design (e.g., centerline frame dimensions, simple foundation
models, no slab or non-frame column participation in lateral resistance.).
Baseline assumptions were provided for both the elastic and inelastic analyses.
The response spectrum specified was an equal hazard spectrum for the Northern
San Fernando Valley, with a recurrence interval of 475 years. Baseline
strong-motion time histories included the following: 1940 El Centro, 1978
Tabas Iran, 1994 Northridge Sylmar County Hospital, and the 1994 Canoga
Park record. These records were selected for purposes of comparison to
previous standard records (El Centro), representative records from the
1994 Northridge Earthquake (Sylmar County Hospital and Canoga Park), and
a larger near-field event that includes a long duration pulse.
In addition to the baseline analyses, the investigators were encouraged
to modify their analytical models to include more accurate representations
of the actual structural systems of the building. These improvements included:
varying the assumed level of damping and modeling the participation of
panel zones, composite floor slabs, non-frame columns, and foundation flexibility.
The performance of the enhanced analytical models could then be compared
with that of the baseline results. In addition, a simple fracture element
was developed as part of the Phase 1 project for use in nonlinear analyses.
A suite of simulated time histories was developed for each of the building
sites in the analytical investigation portion of this Sub-Task that did
not have strong-motion instruments. These were prepared as a portion of
the work in Task 4 of the SAC Steel Project (SAC
Technical Report 95-03). The suite of time histories included nine
records for each site. These time histories were developed to provide a
mechanism for estimating the demands that these buildings actually experienced.
Various analytical results were collected and evaluated as indicators
of damage. For the elastic analyses, these results included the roof displacement
ratio, interstory drift ratios, and Demand/Capacity ratios (DCR's) for
the various members of the frames. For the inelastic analyses, the roof
displacement ratio, interstory drift ratios, and inelastic joint and member
demands were evaluated. In addition, various investigators attempted to
develop other methods for assessing the results.
A number of general observations from these analyses were of use in
the development of the SAC Interim
Guidelines (SAC 95-02, FEMA 267). The following general trends
were identified in these analyses:
-
All of the analytical procedures were able, in at least a limited fashion,
to provide an indication of the locations within the building where connection
damage was most probable. That is, analytical indicators could be identified
in all cases that would provide a better indication of damage locations
than random sampling.
-
None of the procedures or indicators evaluated were very reliable in predicting
specific locations for damage. Correlation between connection demand indicators
and damage was better if incipient root cracks (Type WI damage) were removed
from the sample.
-
Connection demand parameter indicators such as relative values of DCR or
inelastic rotation demand appear to be somewhat more reliable in predicting
damage than more global indices such as interstory drift. The postulation
of absolute values for DCRÕs in relation to damage was not possible
with the data available.
-
Inelastic analyses tend to provide somewhat better reliability than elastic
analyses in identifying damage patterns. Examples to the contrary were
also found in this sample.
-
In taller buildings, higher mode effects appear to have been the cause
of a concentration of damage in the upper stories. Standard push-over analyses
cannot easily identify these effects.
-
Enhanced analytical procedures beyond the simple baseline assumptions led
to improved correlation of damage analytical predictions location and type.
Global factors such as interstory drift were not materially affected, but
local response predictors were modified, and the results improved. The
degree of improvement varied greatly from case to case. Modeling of three-dimensional
effects showed a pronounced improvement for buildings that were susceptible
to torsional motions.
-
-
The ground motions generated by the Northridge Earthquake did not generate
large interstory drifts, DCR's or inelastic joint rotation demands on the
moment frame buildings in this study (maximum drifts on the order of 1.5%
or less, DCR's with a maximum of 2.5, and joint rotations up to 0.02).
Other ground motions considered in the study generated much larger joint
demand values and expected interstory drifts. Increased levels of structural
damage could be expected under these scenarios.
A complete analysis of the data and summary report has been prepared to
condense the large amount of information presented in this volume, and
to make a number of comparisons between the various analytical studies.
This summary report, which follows the Executive Summary in Part I of this
two volume set, provides an excellent overview and comparison of the results
of this task. The reader is strongly encouraged to read this summary report.
In addition to the analytical investigations, this technical report
includes the results from post-earthquake ambient vibration surveys of
three of the buildings. These surveys indicated significantly shorter modal
periods than the analytical investigations, but identified a consistent
ratio between the measured and analytical values.
As a quality assurance measure, all SAC Steel Project Investigations
were overseen by a Technical Advisory Panel (TAP). The panel for the Task
3 investigations were specialists in the fields of elastic and inelastic
structural analysis, modeling techniques and structural steel system behavior.
