Search

Continuous advancements in scientific and engineering understanding of earthquake phenomena, combined with the associated development of representative physics-based models, is providing a foundation for high-performance, fault-to-structure earthquake simulations. However, regional-scale applications of high-performance models have been challenged by the computational requirements at the resolutions required for engineering risk assessments. The EarthQuake SIMulation (EQSIM) framework, a software application development under the US Department of Energy (DOE) Exascale Computing Project, is focused on overcoming the existing computational barriers and enabling routine regional-scale simulations at resolutions relevant to a breadth of engineered systems. This multidisciplinary software development—drawing upon expertise in geophysics, engineering, applied math and computer science—is preparing the advanced computational workflow necessary to fully exploit the DOE’s exaflop computer platforms coming online in the 2023 to 2024 timeframe. Achievement of the computational performance required for high-resolution regional models containing upward of hundreds of billions to trillions of model grid points requires numerical efficiency in every phase of a regional simulation. This includes run time start-up and regional model generation, effective distribution of the computational workload across thousands of computer nodes, efficient coupling of regional geophysics and local engineering models, and application-tailored highly efficient transfer, storage, and interrogation of very large volumes of simulation data. This article summarizes the most recent advancements and refinements incorporated in the workflow design for the EQSIM integrated fault-to-structure framework, which are based on extensive numerical testing across multiple graphics processing unit (GPU)-accelerated platforms, and demonstrates the computational performance achieved on the world’s first exaflop computer platform through representative regional-scale earthquake simulations for the San Francisco Bay Area in California, USA.

This paper presents a model for distributing zones of liquefaction and nonliquefaction for use in regional liquefaction risk analysis. There are two broad methodologies that have been used to evaluate liquefaction risk on the regional scale: (a) application of site-specific procedures using soil properties inferred from geology, or (b) application of geospatial proxies for liquefaction. The first approach will tend to predict similar liquefaction probabilities across broad areas with similar geology, water table depths, and shaking intensities. The second approach yields the probability of liquefaction, which can be interpreted as the portion of the area affected by liquefaction ((Formula presented.)). Neither approach, however, gives an informed prediction of the spatial distribution of liquefaction and the resulting displacements, which are particularly important for assessments of seismic risk for spatially distributed infrastructure systems. We propose a methodology for incorporating spatial correlation into a geospatial proxy for liquefaction to create maps of liquefaction and nonliquefaction for a given earthquake scenario. First, we describe a latent Gaussian process that is assumed to govern the spatial distribution of liquefaction. Next, a database of empirical observations of liquefaction is used to obtain the coefficients that describe that latent Gaussian process. The proposed model yields random realizations of maps of liquefaction and nonliquefaction conditioned on a map of (Formula presented.). Such maps can be used to constrain the area over which displacements are estimated using soil properties inferred from geology and are therefore a critical component in reducing bias in assessments of liquefaction risk at the regional scale.

One-dimensional site response analyses (1D SRAs) with shear-wave velocity (VS) randomization are commonly performed to estimate median site-specific amplification factors (AFs) under the implicit assumption that this approach yields a realistic response. In this work, an investigation is conducted to determine the appropriate amount of VS randomization (σlnVs) needed to capture a median response that accounts for 2D VS spatial variability effects. Results from 2D SRAs and 1D SRAs with VS randomization show that the median 2D seismic responses are generally higher than 1D responses at the site’s fundamental frequency, and that higher VS variability has a mild impact on the median 2D seismic response amplitude at the fundamental frequency, whereas it significantly reduces the median 1D response. Findings indicate that the 84th percentile AFs based on 1D SRAs conducted with VS randomization using σlnVs = 0.25, approximate well with the more realistic median 2D SRA-based AFs around the fundamental frequency, while the 70th to 60th percentiles might be more appropriate at higher frequencies. The benefit of using percentiles of the 1D SRA-based AFs higher than the median is shown for different site conditions and supported by comparisons against empirical data from four downhole sites.

Computational simulations have become central to the seismic analysis and design of major infrastructure over the past several decades. Most major structures are now “proof tested” virtually through representative simulations of earthquake-induced response. More recently, with the advancement of high-performance computing (HPC) platforms and the associated massively parallel computational ecosystems, simulation is beginning to play a role in increased understanding and prediction of ground motions for earthquake hazard assessments. However, the computational requirements for regional-scale geophysics-based ground motion simulations are extreme, which has restricted the frequency resolution of direct simulations and limited the ability to perform the large number of simulations required to numerically explore the problem parametric space. In this article, recent developments toward an integrated, multidisciplinary earth science-engineering computational framework for the regional-scale simulation of both ground motions and resulting structural response are described with a particular emphasis on advancing simulations to frequencies relevant to engineered systems. This multidisciplinary computational development is being carried out as part of the US Department of Energy (DOE) Exascale Computing Project with the goal of achieving a computational framework poised to exploit emerging DOE exaflop computer platforms scheduled for the 2022–2023 timeframe.

The existing observational database of the regional-scale distribution of strong ground motions and measured building response for major earthquakes continues to be quite sparse. As a result, details of the regional variability and spatial distribution of ground motions, and the corresponding distribution of risk to buildings and other infrastructure, are not comprehensively understood. Utilizing high-performance computing platforms, emerging high-resolution, physics-based ground motion simulations can now resolve frequencies of engineering interest and provide detailed synthetic ground motions at high spatial density. This provides an opportunity for new insight into the distribution of infrastructure seismic demands and risk. In the work presented herein, the EQSIM fault-to-structure computational framework described in a companion paper, McCallen et al., is employed to investigate the regional-scale response of buildings to large earthquakes. A representative M = 7.0 strike-slip event is used to explore the distribution and amplitude of building demand, and comparisons are made between building response computed with fault-to-structure simulations and building response computed with existing measured near-fault earthquake records. New information on the distribution and variability of building response from high-performance parallel simulations is described and analyzed, and favorable first comparisons between building response predicted with both fault-to-structure simulations and real ground motions records are presented.

This study presents a critical assessment of the reliability of current base plate connections in steel special moment resisting frames (SMRFs). Using a probabilistic outlook, this research evaluates the reliability of exposed column base (ECB) connections in SMRFs designed based on the current seismic design provisions; it suggests (and implements) a statistical approach to compute resistance factors for three modes of failure (concrete bearing, base plate yielding at tensile interface, and anchor bolt fracture) of ECB connections to achieve a target reliability index, β, of 4.5. Since ECB connections are limited to short buildings, therefore, this study is conducted on two-story and four-story SMRFs which are analyzed using a suite of 120 ground motions originating from strike-slip and reverse faults. ECB connections for the two-story building are designed to simulate pinned connection, while the bases of the four-story building represent moment connections. Detailed methodology for calculating the β of ECB connections is presented considering the three limit states in a moment–axial load interaction curve. Results indicate that the implementation of current seismic provisions results in β ~ 3.3 for non-moment resisting ECB connections for all tried combinations of resistance factors. For moment resisting ECB connections, however, only the designs based on a resistance factor for concrete bearing failure mode less than the current 0.65 result in an acceptable reliability factor of β > 4.5.

The Next-Generation Liquefaction database is a resource for the geotechnical hazard community. It is publicly available online under the following digital object identifier (DOI): 10.21222/C2J040. The database organizes objective liquefaction data into tables and fields (columns of information), with the relationships among the tables and fields described by a schema. The data are organized into tables pertaining to (1) sites, including geotechnical and geophysical site investigation data, surface geology information, and laboratory test data; (2) earthquake events, including source and ground motion information; and (3) observations of sites following events. The schema was vetted through community outreach efforts involving multiple workshops and meetings. Users can view the data, download existing data, and upload new data through a geographic information system (GIS)-based graphical user interface. Information uploaded to the database is reviewed by a database working group to verify consistency between uploaded data and source documents. The database is replicated in DesignSafe where users can interact with the data using Python scripts in Jupyter notebooks, view point cloud data using Potree, and interact with geospatial data using QGIS.

Methods for the seismic design of base connections in steel moment frames are well-developed and routinely utilized by practicing engineers. However, design loads for these connections are not verified by rigorous analysis. This knowledge gap is addressed through nonlinear time history simulations using design-level seismic excitation that interrogate demands in column base connections in 2-, 4-, 8-, and 12-story steel moment frames, featuring base connections that reflect current U.S. practice. The results indicate that: (1) for exposed base plate connections, lower bound (rather than peak) estimates of axial compression are suitable for design because higher axial forces increase connection strength by delaying base plate uplift; (2) even when designed as pinned (as in low-rise frames), base connections carry significant moment, which can be estimated only through accurate representation of base flexibility; and (3) the failure of embedded base connections is controlled by moment, which may be estimated either through overstrength or capacity-based calculations.

This paper presents and validates an analytical formulation, denoted as Performance Prediction Equations (PPEs), that relates the seismic response engineering demand parameter (EDP) of buildings to earthquake parameters such as magnitude, epicentral distance, and type of faulting. PPEs are conceptually novel and can be readily included in any hazard calculation program to directly estimate EDP hazard curves. The PPEs presented herein are based on the linear-ization of response spectrum analysis (RSA) formulation for estimation of the seismic response of multi-degree-of-freedom (MDOF) models for planar structural systems. Equations for mean and variance are provided for floor displacement, interstory drift ratio, and normalized base shear. The input parameters needed to apply the proposed PPEs are the modal properties of the structural system and the selection of an existing ground motion model (GMM). The proposed PPEs are validated against simulated results using a set of planar building models and the Campbell-Bozorgnia 2014 GMM. The comparison confirms that the proposed PPEs provide an accurate estimate of the statistics of the said EDPs.

Accurate measurements of the time-dependent deformations of a building during earthquake excitation are essential for interpretation of the dynamic response of the as-built system and for quantifying the seismic demands. Traditional approaches for monitoring building systems are based on strong motion accelerometers mounted at selected elevations. However, accelerometer-based systems do not directly measure the deformations of the structure, and can have significant limitations that make it challenging to correctly measure deformations, particularly permanent deformations from inelastic response. In the study described herein, computational simulations and experiments were combined to evaluate the potential of a new optically based sensor to directly measure time-dependent deformations of a building, including inelastic deformations. The sensor methodology includes corrections for localized structural member rotations and can provide estimates of the absolute accelerations at each floor. A laser-based system utilizing a recently developed discrete diode position sensor (DDPS) is evaluated, and the ability of such a system to measure earthquake induced transient deformations characterized by building interstory drift is demonstrated.

Seismic levee performance is most readily computed for short segments having consistent geometry, soil conditions, and seismic demands. Spatial variations of seismic demands and of segment capacities significantly influence system risk, which is critical for flood protection because any segment failure within the system can cause inundation. We present a methodology to compute the probability of seismic levee system failure conditional on individual segment fragility and spatial correlations of demands and of capacities. Seismic demands are estimated from ground motion prediction equations; their correlation is available in the literature. Capacities and their correlation are derived from levee damage observations from a levee system in Japan shaken by two earthquakes. We find seismic capacities to exhibit positive correlations over shorter distances than for demands. System fragility is computed using Monte Carlo simulations where segment demand and capacity realizations are generated to account for spatial correlations. We find that the probability of system failure is lower than would be obtained under an assumption of no correlation and that damage probability increases as the number of components in the system increases.

Four large-scale centrifuge tests were performed at the NEES@UCDavis equipment site to study the cyclic behavior of levee structures resting atop soft organic peat. The model configurations using a non-liquefiable levee focused on the seismic deformation potential of peat during primary consolidation and secondary compression. The tests performed with a sandy levee studied the liquefaction potential of saturated loose sand fill overlying soft peat as well as the levee-peat-interaction under different loading conditions. The models were subjected to scaled ground motions representative of the Sacramento/San Joaquin Delta. System instrumentation consisted of linear potentiometers, pore pressure sensors and accelerometers. Slow data recorded at 1 Hz document the settlements during spin up, application of ground motions, and spin down. Fast data sampled at 4,167 Hz measured the dynamic response of the system, the excess pore pressure increase and immediate settlements. The project is archived at the NEES data repository under nees.org/warehouse/project/1161 .

We characterize the seismic fragility of levees along the Shinano River system in Japan using field performance data from two M 6.6 shallow crustal earthquakes. Levee damage is quantified based on crack depth, crack width, and crest subsidence for 3,318 levee segments each 50 m long. Variables considered for possible correlation to damage include peak ground velocity (PGV), geomorphology, groundwater elevation, and levee geometry. Seismic levee fragility is expressed as the probability of exceeding a damage level conditioned on PGV alone and PGV in combination with other predictive variables. The probability of damage (at any level) monotonically increases from effectively zero for PGV < 14cm/s to approximately 0.5 for PGV ≈ 80 cm/s. Of the additional parameters considered, groundwater elevation relative to the levee base most significantly affects fragility functions, increasing and decreasing failure probabilities (relative to the PGV-only function) for shallow and deep groundwater conditions, respectively.

A model levee resting atop soft compressible peaty organic soil in the Sacramento/San Joaquin Delta was shaken by forced vibration to study the seismic deformation potential of the underlying peat and measure dynamic levee-peat interaction. Forced vibration testing occurred over a frequency range of 0 Hz to 5 Hz and produced force amplitudes applied to the embankment crest that induced elastic to nonlinear levee-foundation responses. Available data include acceleration records from sensors mounted on the model levee and on the ground surface near the model levee, and acceleration and pore pressure measurements from sensors embedded in the underlying peat. A remote data acquisition system measured settlements and pore pressures over a span of more than a year, encompassing time before and after the dynamic testing. Small pore pressures were generated in the peat during testing although embankment settlements from cyclic loading were small..

A two-dimensional nonlinear dynamic finite element (FE) model was developed and calibrated against dynamic centrifuge tests to study the behavior of soil-pile-structure systems in liquefied and laterally spreading ground during earthquakes. The centrifuge models included a simple structure supported on pile group. The soil profiles consisted of a gently sloping clay crust over liquefiable sand over dense sand. The FE model used an effective stress pressure dependent plasticity model for liquefiable soil and a total stress pressure independent plasticity model for clay, beam column elements for piles and structure, and interface springs that couple with the soil mesh for soil-structure interaction. The FE model was evaluated against recorded data for eight cases with same set of baseline parameters. Comparisons between analyses and experiments showed that the FE model was able to approximate the soil and structural responses and reproduce the lateral loads and bending moments on the piles reasonably well. © 2013, Earthquake Engineering Research Institute.