SMIP 2019 Seminar

A unique opportunity for gaining insights is facilitated by availability of the CSMIP Eureka Channel Bridge seismic records. Of special interest is the recorded response of a bridge pier at the deck, pile cap and within the underlying pile foundation. In this study, recorded response from the strongest to date 2010 Ferndale earthquake (PGA of about 0.25g), along with other available low-amplitude events are employed to evaluate the pile foundation, and overall bridge seismic response. Finite Element modeling is employed along with the optimization framework SNOPT, to derive salient characteristics of the overall bridge system response.

This study aims at validating the modeling approaches of ordinary bridge structures suggested by Caltrans (SDC 2013; referred to as SDC models) and sophisticated models suggested by researchers, herein termed as Stick models. The validation is conducted using the CSMIP sensor data of four ordinary bridges in California with seat-type, and monolithic abutments. The backbone curves of the structural components of the Stick model including shear keys, abutment piles, and backfill soil are updated using Particle Swarm Optimization. This study yields the guidelines for calibration of parameters of bridge structural components and suggests improvements for modeling approaches of such bridges.

We derive non-ergodic site response for California sites using an expanded version of the NGA-West2 database. We then investigate the degree to which different site response analysis methods capture observations. An ergodic site term provides a baseline against which other models are compared. Here we emphasize site-specific ground response analysis for sites with in situ VS measurements. We describe the assignment of damping to individual soil layers using geotechnical models and site-specific spectral amplitude decay parameter . We provide data-model comparisons for cases in which ground response analyses provide variable levels of effectiveness.

This paper presents the results of a study on the identification of earthquake input excitations for CSMIP-Instrumented buildings. The true earthquake input motions exciting buildings may not be available for various reasons. For example, when there is Soil-Structure Interaction (SSI) effects, the recorded signal at the foundation level, which is commonly used as input excitation, is a part of the building’s response. Also, the waves scattered from a vibrating building can alter the wave field around the building, so the so-called recorded Free-Field Motions (FFMs), another input motion candidate, could be polluted with these reflecting waves. Moreover, if there is significant Kinematic SSI, what actually a building experiences as input excitation is different from FFM and foundation response. These unmeasured motions are called Foundation Input Motions (FIMs) and have to be identified from recorded building’s responses. In this paper, we propose various methods to carry out this task along with their verification, validation and real-life applications.

The Ridgecrest Earthquake Sequence began on July 4, 2019 with a Mw 6.4 earthquake at 10:33 am PDT at a depth of 8.7 km. The epicenter was located about 18 km east-northeast of the City of Ridgecrest within the Naval Weapons Station China Lake (NWSCL) property. This event was preceded by several small foreshocks a few days prior to the event. Surface rupture from this event was expressed as a zone of surface faulting over 17 km long, consisting of several strands with en-echelon stepovers, striking northeast-southwest with left-lateral displacement of the ground surface. Rupture appears to have propagated from the epicenter toward the southwest.

The Ridgecrest Earthquake sequence included a foreshock event on July 4 2019 (M6.4) and a M7.1 mainshock event on July 5 2019. These events occurred in the Eastern California Shear Zone, near Indian Wells Valley, south of China Lake and west of Searles Valley. GEER partnered with several organizations to collect perishable data and document the important impacts of these events, including the US Geological Survey, the California Geological Survey, the US Navy, the Southern California Earthquake Center, and local utilities. Critical geotechnical features of this event are extensive left-lateral (M6.4 event) and right-lateral (M7.1 event) surface ruptures over fault segments of variable complexity and width as well as across extensional and compressive step-over zones. We also document lifeline performance at fault crossings (gas, water, electrical), mainshock slip and afterslip, liquefaction and lateral spreading features, and liquefaction effects on structures. These effects are documented using field (ground) mapping and aerial imagery that will support subsequent development of high-resolution digital elevation models. Over 1200 ground motions were recorded from the foreshock and mainshock alone, with many additional aftershock records. The data demonstrate significant impacts of site response and rupture directivity on ground motion attributes.

The overall scope of this study is to evaluate the acceleration amplification effects of nonstructural components using recorded earthquake responses of buildings and nonstructural components. Specifically, two separate, yet complementary efforts are undertaken, namely: 1) characterizing nonstructural component amplification effects using a large set of building earthquake responses that are available in the CESMD strong motion database, and 2) identifying the dynamic characteristics of instrumented nonstructural components integrated within a full- scale building shake table test program. Findings from this study are intended to supplement current seismic design provisions of nonstructural systems with evidence obtained from recorded data.

The overarching goal of this research is to derive innovative methodologies for analyzing diverse sensing streams to yield actionable information that directly support post-earthquake response, emergency management, and disaster recovery. As a step towards this goal, the objective of this study is to use data collected from the California Strong Motion Instrumentation Program (CSMIP) to demonstrate that a particular class of system identification (SI) techniques – namely, subspace identification (SI) or recursive subspace identification (RSI) – is especially suitable for rapid, post-disaster, structural health assessment. The advantage of SI/RSI is that it is an input-output and data-driven method, where only structural response records (e.g., acceleration records) are needed for extracting dynamic properties of the structure. In this paper, both SI and RSI were be applied to assess CSMIP-instrumented buildings with acceleration records from past ground motion events. The result verified that building dynamic characteristics (i.e., natural frequencies and mode shapes) could be clearly identified using all the recorded data simultaneously. In addition, the RSI algorithm was also employed for analyzing data recorded from the Northridge earthquake event. Time-varying modal properties of the building were also examined.

The rotational fixity of column base connections in Steel Moment Resisting Frames (SMRFs) strongly influences their seismic response. However, approaches for estimating base fixity have been validated only against laboratory test data. In the present study these approaches are examined based on strong motion recordings from two instrumented SMRF buildings in California. Three-dimensional simulation models are constructed for these buildings, including the gravity framing and nonstructural stiffness. For each building, the base fixities are parametrically varied. These include pinned and fixed bases, as well as intermediate fixities determined from previously developed models that are appropriate to simulate the specific types of base connections used in the buildings. The simulated response of these buildings is compared to strong motion recordings to inform optimal approaches for simulating column bases.