SMIP 2016 Seminar

The Utilization of Ground Motion Simulation (UGMS) committee of the Southern California Earthquake Center (SCEC) is currently developing risk-targeted Maximum Considered Earthquake (MCER) maps for possible inclusion as an amendment to the ASCE 7-16 edition of the Los Angeles City Building Code (LACBC). These maps are scheduled for release in 2017. The maps will be based on 3-D numerical ground-motion simulations and ground motions computed using the empirical ground-motion prediction equations (GMPEs) from the Pacific Earthquake Engineering Research (PEER) Center NGA West2 project. A web-based lookup tool, similar to the USGS lookup tool, will be posted so users can obtain the MCER response spectrum for a specified latitude and longitude and for a specified site class or 30-m average shear-wave velocity, Vs30. The acceleration ordinates of the MCER response spectrum will be provided at multiple natural periods in the 0 to 10-sec band; values of SDS and SD1, per the requirements in Section 21.4 of ASCE 7-16, will also be listed. ​.

A new procedure for rapid post-earthquake safety evaluation of bridges has been developed, using existing strong motion records, fragility curves and ground motion data immediately available following an earthquake that will provide the engineer or person directly in charge of the bridge to make a more informed decision to close or keep a bridge open to traffic. The recently constructed Carquinez I80 West Bridge (Alfred Zampa Memorial Bridge) was selected to demonstrate the procedure. This paper describes the detailed time history finite element analysis conducted using strong motion data for the 26 scenario earthquake events and the development of the fragility curves using shake table test results on reinforced concrete columns tested through five damage states to final failure. Fragility functions are developed for various seismic parameters for each damage state and calibrated for maximum drift ratios for inclusion into the rapid safety evaluation of the Carquinez Bridge.

Ninety-four buildings, with a total of 1045 distinct seismic event and building direction records, were selected from the CSMIP database to identify modal quantities (i.e., natural periods and equivalent viscous damping ratios). The selected buildings include steel and reinforced concrete moment resisting frames (i.e., SMRF, and RCMRF), reinforced concrete walls (RCW), concentrically braced frames (CBF), eccentrically braced frames (EBF), masonry walls (MAW), precast concrete walls (PCW), reinforced concrete tilt-up bearing walls (RCTUW), unreinforced masonry (URM), and WOOD. Simplified and practical equations for modal quantities along with variation of such parameters to structural system types, building height, amplitude of excitation, and system identification technique for a subset of buildings were reliable data was available are presented.

Substructure method is commonly used in engineering practice to take Soil-Structure Interaction (SSI) effects into account in seismic design. In this method, soil is modeled using discrete spring elements—ideally Impedance Functions (IF)—that are attached to the superstructure; and the Foundation Input Motions (FIMs) are applied at the remote ends of these springs. While the application of the substructure method is simple and its computational cost is low, the determination of FIMs and the IFs are generally challenging. This paper presents results of a two-year project during which a new method was developed to identify IFs and to back-calculate FIMs from earthquake data recorded at CSMIP-instrumented buildings. The method features a flexible-based Timoshenko beam idealization of the superstructure and its soil-foundation system, and is based on updating the parameters of this model such that its responses match real-life data. Details of the said method are briefly reviewed first, followed by a presentation of the results it produced on currently available CSMIP data.

One-dimensional (1D) ground response analyses are often used with an expectation that they provide an unbiased estimation of site effects, and therefore improve upon site response estimates from ergodic models (i.e. site terms in ground motion models, GMMs). We use California vertical array data to (1) investigate the degree to which 1D analysis provides results compatible with observation, thus checking the typical assumption, and (2) quantify epistemic uncertainty in site response estimates from ground response analysis. Objective (1) was discussed in a previous CSMIP conference paper and a brief update is provided here. We present our methodology and preliminary results for quantifying epistemic uncertainty in site response as estimated from 1D analysis. We decompose prediction residuals into between- and within-site components, and take the between-site standard deviation as a quantification of epistemic uncertainty. Preliminary results suggest values ranging from 0.35-0.5 in natural log units.

This paper presents progress made towards the identification of spatial variability in bridge Foundation Input Motions (FIMs) in a project funded by the California Geological Survey. The term spatial variability denotes here the differences in amplitude and phase of seismic motions recorded over extended areas, and it is well accepted that lifeline structures, such as long bridges, are prone to its effects, because they extend over relatively long distances. The specific objective of the project is to identify FIMs from response signals recorded by instrumented bridges during the South Napa 2014 earthquake, decompose them to bedrock motions and site effects, and finally quantify the spatial variability for each bridge structure. In this progress paper, we present initial observations, data selection, and the theoretical basis of the methodology that will be employed to process the recorded data. The overall methodology comprises two novel approaches (i) for output only identification of bridges under multiple support excitations, and (ii) for blind identification of bedrock motions and site effects from two (or more) ground surface motions (FIMs). The said two methods are briefly described and numerically verified in the present paper. The first method will be employed to extract FIMs from spatially sparse measurements of bridge responses, while the second one will be used to further identify the site effects and bedrock motions from the recovered FIMs.

The 2014 South Napa mainshock caused significant damage in the Northern California Bay Area. Time series from a foreshock, mainshock, and three aftershocks were collected from various agencies. These were processed following the Pacific Earthquake Engineering Research Center (PEER) standard data-processing methods, and a ground-motion database was developed. Metadata such as fault style, source-to-site distance, average shear wave velocity in the top 30 m (Vs30), and basin depth were collected. Shear wave velocity profiles were also measured by the Spectral Analysis of Surface Wave Dispersion (SASW) technique at selected strong-motion stations. These datasets were combined in the ground motion database and compared to the Ground Motion Models (GMMs) from the NGA-West2 studies to evaluate the regional attenuation of these events. Time series at two geotechnical downhole array sites were also collected from 29 earthquakes to calculate apparent wave velocities from wave travel times and empirical transfer functions to understand wave amplification. Characteristics of pulse-like records from the South Napa and NGA-West2 databases were also analyzed to compare near-fault regions between these databases. The influence of pulse-like records was also investigated using inelastic response spectra to understand the damage potential on structures. These observed ground- motion characteristics are summarized in this study.

This paper presents the one-dimensional (1D) site response analysis (SRA) of three geotechnical downhole arrays in California subjected to both strong and weak earthquake shakings. The arrays were initially assessed in terms of effectiveness of 1D SRA using taxonomy exercise. Then SRA were performed utilizing finite element program LS- DYNA to study the site effects of the selected arrays. Lastly, the predictions were compared with the recorded counterparts and the uncertainties of the 1D SRA models were evaluated. Among the analyzed ground motions, we focus on the analysis results of the mainshock and aftershock of 2014 Mw 6.0 South Napa Earthquake.

Identifying a measure of ground motion intensity that is well correlated with strongly nonlinear response is desirable not only for reducing the required number of response history analyses but also for establishing criteria for selecting ground motion time histories for conducting such analyses. The most commonly used ground motion intensity is the response spectral ordinate of a 5% damped system with a period equal to the fundamental period of the structure being analyzed. In this study we explore and evaluate alternate measure of ground motion intensity with emphasis with those that are well correlated with strongly nonlinear response of multi-degree-of-freedom system and for estimating the probability of collapse of a structure. Preliminary results indicate that using the average of spectral ordinate over a relative wide range of periods including both periods shorter and longer than the fundamental period of vibration leads to significant reductions in the record-to-record variability of ground motion intensities triggering collapse. Other alternatives measure of intensity but based on time domain features of acceleration times histories are also being explored.