SMIP 2003 Seminar

We have developed a modification to the ground motion model of Abrahamson et al. (1997) that takes into account the effects of basin generated surface waves. The main feature of our model is that for response spectral accelerations at periods of 4 and 5 seconds, the Abrahamson and Silva (1997) model for soil sites should be scaled by a factor of 1.65 in order to represent the ground motions on soil sites located within sedimentary basins. Our finding that no scaling of the Abrahamson and Silva (1997) model is required for periods shorter than 4 seconds reflects the fact that most of the deep soil recordings on which that model is based come from sedimentary basins. The fact that our first order model does not have a dependence on distance to the basin edge or the depth of the basin makes it easily applicable in ground motion calculations, especially those for probabilistic seismic hazard analyses, which typically involve the calculation of ground motions from many different scenario earthquakes occurring on a variety of faults surrounding the site. We have identified many second order basin effects that are not quantified in our first order model of basin effects.

A Design Ground Motion Library (DGML) is being developed that will contain selected recorded acceleration time histories considered to be suitable for use by engineering practitioners for the time history dynamic analysis of various facility types in California and other parts of the Western United States. The DGML will include: (1) the electronic library of selected time histories and their associated ground motion parameters and supporting information on the earthquake source, travel path, and site characteristics; and (2) detailed guidelines for forming and scaling sets of time histories for applications. The characteristics of the seismic environment, including earthquake magnitude, faulting mechanism, source-to-site distance, nearfault directivity conditions, and site conditions, and the damaging characteristics of time histories are being incorporated into criteria for selecting and binning records for the library.

The California Integrated Seismic Network (CISN) is a consortium of institutions engaged in statewide earthquake monitoring. The five core members are the California Geological Survey, the California Institute of Technology, the U.S. Geological Survey Offices at Pasadena and Menlo Park, and UC Berkeley. The California Office of Emergenc y Services (OES) participates as an ex-officio participant. The TriNet project initiated in southern California with FEMA support was a prototype for the statewide CISN project. The CISN has a statewide Engineering Data Center, a southern California seismic Data Center at Pasadena, and a northern California seismic Data Center in the Bay area. The CISN Engineering Data Center is operated by the CGS Strong Motion Instrumentation Program (CSMIP) in cooperation with the USGS National Strong Motion Program (NSMP). A primary goal of the Engineering Data Center as well as the other two Data Centers is to provide robust and rapid information products after an earthquake, with products ranging from the ShakeMap to strong- motion data and calculated parameters. A high-speed T-1 computer communication network, or Intranet, which connects all CISN partner agencies has been recently established, depicted in Figure 1. The T-1 communication ring consists of dedicated segments connecting each center and OES, and is completely independent of the Internet. The ring became operational early in 2003. The Internet is also used in communication between the centers, but the T-1 ring provides the secure, reliable backup to the Internet that is needed. A second level of backup, using OES satellite communication channels, is also planned. Effective communication requires standardized protocols. Data packets and formats for the exchange of parametric data and waveforms are being finalized by the CISN Program Management Committee and its Standards Committee. With the completion of the T-1 ring, the CISN partners have begun routine exchange of strong-motion parametric data between centers while the protocols for exchange of waveforms are being standardized. When routine waveform exchange begins, planned for later in 2003, it will be possible to assemble the strong-motion data for all strong-motion stations in the State at CISN, in forms convenient for use by the earthquake engineering or seismology communities, an unprecedented advance.

The ATC-54 Report, Guidelines for Utilizing Strong-Motion Data and ShakeMaps in Postearthquake Response, prepared by the Applied Technology Council for the California Geological Survey, provides guidance on (1) the use of near-real-time computer-generated ground-motion maps in emergency response, and (2) the use and interpretation of strong-motion data to evaluate the earthquake response of buildings, bridges, and dams in the immediate postearthquake aftermath. Guidance is also provided on the collection of data describing the characteristics and performance of structures in which, or near which, strong-motion data have been recorded.

This paper describes real-time damage and loss estimation using the HAZUS earthquake loss estimation technology and ShakeMap data, and provides an example comparison of predicted and observed losses for the 1994 Northridge earthquake.

HAZUS [NIBS, 1999, Kircher et al., 1997a/1997b, Whitman et al., 1997] is the standardized earthquake loss estimation methodology developed by the National Institute of Building Sciences (NIBS) for the United States Federal Emergency Management Agency (FEMA). HAZUS was originally developed to assist emergency response planners to "provide local, state and regional officials with the tools necessary to plan and stimulate efforts to reduce risk from earthquakes and to prepare for emergency response and recovery from an earthquake."

HAZUS can also be used to make regional estimates of damage and loss following and earthquake using ground motion, ShakeMap, data provided by the United States Geological Survey (USGS) as part of Tri-Net in Southern California [Wald et al., 1999] or by other regional strong-motion instrumentation networks.

A magnitude 4.3 earthquake was recorded on an array of accelerometers at Pacoima Dam on January 13, 2001. The records are used for two purposes: (1) to analyze the effects that canyon topography has on the ground motion along the abutments, and (2) as input for a system identification study, leading to a calibrated finite element model of Pacoima Dam. The quantified amplification and time delay characteristics of the 2001 abutment motions serve as a basis for generating records to replace ones that went off-scale during the 1994 Northridge earthquake. These generated records were then used in the finite element model to verify that nonuniform ground motion caused by the topography has a significant impact on the dam response. Forced vibration tests were also conducted in July/August 2002.

As part of the Strong-Motion Instrumentation Program in California (CSMIP), several bridge structures have been instrumented to record accelerations for the duration of each earthquake event that strikes. The subject of this paper is the investigation of the measured and calculated responses of the heavily instrumented 10-span connector bridge (53-2795F) at the 5/14 Interchange, subjected to a recorded earthquake. A detailed computer model of the bridge structure was developed, allowing comparisons to measured response quantities for many of the 42 sensor locations. In order to compare measured and computed results it was necessary to develop global axes that were common to both the bridge model and structure instrumentation.

This paper describes results of the CSMIP-funded project to develop correlations of observed building performance with measured ground motion. Much of the information presented in the paper is taken from King et al. (2002), which described the progress of the project to date at last year’s SMIP02 Seminar. Motion-damage relationships in the form of lognormal fragility curves and damage probability matrices have been developed for wood frame, steel moment frame, and concrete frame buildings – building types for which there are enough samples in the database to warrant statistical analysis. The ground motion parameters that were found to exhibit relatively higher correlations with building performance were used in the analysis. Building performance is characterized in terms of damage states and performance levels. The resulting relationships are compared to those published in ATC-13 (ATC, 1985) and HAZUS99 (FEMA, 1999). The comparison shows that the relationships developed in the project are quite different from the published models; however, the loss estimates resulting from the application of the models are similar.

The objective of this investigation is to evaluate the FEMA-356 Nonlinear Static Procedure (NSP) and a recently developed Modal Pushover Analysis (MPA) procedure using recorded motions of buildings that were damaged during the 1994 Northridge earthquake. It is found the FEMA-356 NSP typically underestimates the drifts in upper stories and overestimates them in lower stories. The MPA procedure provides much-improved estimates of the response compared to the FEMA-356 NSP. In particular, the MPA procedure, unlike the FEMA-356 NSP, is able to capture the effects of higher modes.

The historic Oakland City Hall experienced extensive damage in the 1989 Loma Prieta earthquake, and was subsequently the focus of an intensive process of testing, historic evaluation, dynamic structural analysis, and retrofit design using seismic isolation. A comprehensive post-earthquake study done by a team of architects and engineers, and reviewed by FEMA and SHPO, concluded that seismic isolation was the most cost-effective and behavior- effective method to protect the landmark building and its occupants from seismic hazards. The retrofit was completed in 1995. Numerous technical challenges and questions were confronted in the course of this pioneering project. The resolution of these issues, and the features of the seismic design and analysis are discussed.

The Oakland City Hall was strengthened after the 1989 Loma Prieta earthquake and instrumented with 21 sensors by the California Strong Motion Instrumentation Program in 1995. This paper describes the sensor locations in the City Hall and the instrumentation objectives. Low amplitude strong-motion records that were obtained from the instrumentation at the City Hall during the magnitude 4.9 Gilroy earthquake of May 13, 2002 are also presented and discussed.