Earthquake Engineering Applications Research Articles

Determining the number of soil layers using the surface wave method based on decision tree and random forest

The surface wave method is an efficient non-invasive technique to infer shear wave velocity profiles and has wide applications in civil engineering, earthquake engineering, and geophysics. Inversion analysis is a key step in this method. However, inferring the soil layer number is a challenging problem in inversion analysis and needs improvement. Machine-learning algorithms, including decision tree and random forest, are used in this study to infer the number of soil layers based on the dispersion curve of the surface wave. This process includes two steps for obtaining the model. In Step 1, a large set of synthetic dispersion curves is produced, and in Step 2, a model using machine-learning algorithms is trained based on the synthetic data. The analysis results showed that the obtained model could gain more than 75% accuracy in the test data. Comparing the results from the two machine learning models indicated that the random forest method produced a more favorable prediction. Furthermore, investigating the importance of features implied that the longest and shortest wavelengths were more important than other features.

Deep learning-based modeling of the cyclic behavior of replaceable fuse buckling-restrained braces (BRBs)

This paper presents a novel approach for predicting the behavior of buckling restrained braces (BRBs) using long short-term memory (LSTM) and residual neural network (ResNet) models. Nonlinear finite element analysis (NLFEA) usually requires 3D modeling of complex parts and assigning various material and simulation properties before conducting the analysis, which can be time-consuming and computationally costly. LSTM networks are particularly well-suited for modeling sequential data, such as time series, and ResNet is a discriminative deep learning-based model that can handle deeper architectures without the issue of vanishing gradient. Developing a general LSTM model requires a comprehensive database for training, cross-validation, and blind testing. Even if such a database were available, concatenating data from different BRB specimens would not be feasible when modeling a data series. Such a method would intertwine the features of the individual specimens, rendering the data useless for machine-learning purposes. On the other hand, a well-trained LSTM can predict cyclic behavior for specimens that resemble the training data. As such, a group of well-trained models can produce superior predictions when used on individually familiar data. Hence, there is a need for a prescreening classification stage. The framework presented herein utilizes an AI classifier (ResNet) that automatically recognizes the BRB specimen type of the input data and directs it to the appropriate LSTM model for prediction. The presented framework demonstrates excellent predictions with an accuracy greater than 99%. It is customizable, adaptable, and scalable to include different BRB specimens. The input data consisted of the displacement of the BRB specimens based on the AISC 341–10 qualification recommendations, while the output was the corresponding force (hysteretic response to cyclic loading). The proposed framework can be valuable in earthquake engineering applications as it demonstrates superior predictions.

Stochastic emulation with enhanced partial‐ and no‐replication strategies for seismic response distribution estimation

AbstractModern performance earthquake engineering practices frequently require a large number of time‐consuming non‐linear time‐history simulations to appropriately address excitation and structural uncertainties when estimating engineering demand parameter (EDP) distributions. Surrogate modeling techniques have emerged as an attractive tool for alleviating such high computational burden in similar engineering problems. A key challenge for the application of surrogate models in earthquake engineering context relates to the aleatoric variability associated with the seismic hazard. This variability is typically expressed as high‐dimensional or non‐parametric uncertainty, and so cannot be easily incorporated within standard surrogate modeling frameworks. Rather, a surrogate modeling approach that can directly approximate the full distribution of the response output is warranted for this application. This approach needs to additionally address the fact that the response variability may change as input parameter changes, yielding a heteroscedastic behavior. Stochastic emulation techniques have emerged as a viable solution to accurately capture aleatoric uncertainties in similar contexts, and recent work by the second author has established a framework to accommodate this for earthquake engineering applications, using Gaussian Process (GP) regression to predict the EDP response distribution. The established formulation requires for a portion of the training samples the replication of simulations for different descriptions of the aleatoric uncertainty. In particular, the replicated samples are used to build a secondary GP model to predict the heteroscedastic characteristics, and these predictions are then used to formulate the primary GP that produces the full EDP distribution. This practice, however, has two downsides: it always requires minimum replications when training the secondary GP, and the information from the non‐replicated samples is utilized only for the primary GP. This research adopts an alternative stochastic GP formulation that can address both limitations. To this end, the secondary GP is trained by measuring the square of sample deviations from the mean instead of the crude sample variances. To establish the primitive mean estimates, another auxiliary GP is introduced. This way, information from all replicated and non‐replicated samples is fully leveraged for estimating both the EDP distribution and the underlying heteroscedastic behavior, while formulation accommodates an implementation using no replications. The case study examples using three different stochastic ground motion models demonstrate that the proposed approach can address both aforementioned challenges.

Emerging technologies and advanced analyses for non-invasive near-surface site characterization

The in-situ small-strain shear modulus of soil and rock materials is a parameter of paramount importance in geotechnical modeling. It can be derived from non-invasive geophysical surveys, which provide the possibility of testing the subsurface in its natural and undisturbed condition by inferring the velocity of propagation of shear waves. In addition, for soil dynamics and earthquake engineering applications, the small-strain damping ratio plays a relevant role, yet its estimation is still challenging, lacking consolidated approaches for its in-situ evaluation. Recent advancements in instrumentation, such as distributed acoustic sensing (DAS), combined with advanced analysis methodologies for the interpretation of seismic wave propagation (e.g., machine learning and full waveform inversion), open new frontiers in site characterization. This paper presents and compares some advanced applications of measuring 1D and 2D variations in shear wave velocity and attenuation in-situ with reference to a specific case history.

A multi-layered model of Newtonian viscous liquid, fiber-reinforced and poro-elastic media over a self-weighted half-space to investigate the SH-wave interactions

The present study is carried out to investigate the regulation phenomena of SH-waves into a multi-layered structure of Newtonian viscous liquid, fiber-reinforced and poro-elastic strips laid over a gravitating half-space. The classical dynamical coupled theory has been employed to examine the problem. The variable separable technique and Whittaker special function have been used to determine the exact solution of the governing equations. A frequency equation for the SH-wave has been developed for this boundary conditions based problem. Results indicate that the presence of reinforcement, porosity, initial stress, viscosity, and gravitational field in the structure possesses a prominent impact on velocity of the wave. This model contains huge potential to deal with many commercial and industrial applications in acoustical engineering, geo-technical engineering, ultrasonic, earthquake engineering, and geophysics.

Hazard‐consistent simulated earthquake ground motions for PBEE applications on stiff soil and rock sites

AbstractGround motion record selection is a standard step in state‐of‐the‐art performance‐based earthquake engineering (PBEE) applications. It links the structural response to seismic hazard of the site of interest. In this process, suites of hazard‐consistent ground motion recordings of a wide range of intensity levels are selected (and often scaled) from a database of ground motions to be used as input to nonlinear dynamic response analysis. The ideal practice would be to select ground motions from a database of recorded accelerograms in such a way that they are consistent with the seismic hazard at as many intensity measure (IM) levels as possible and for all important causative scenarios. However, the available databases are not heavily populated, especially for the large‐magnitude short‐distance scenarios. Therefore, synthetic ground motions are the natural candidates to enlarge the database of recordings for scenarios that are important to the site hazard but lack, completely or partially, recorded accelerograms. The usage of such recordings, however, can be recommended only after they are proven to be “realistic” through a battery of seismological and statistical tests. In this study, we propose and implement a multi‐fold testing framework for such an evaluation. We first generate a ground motion database of simulated ground motions (SDB) whose values of the parameters of the causative earthquakes are, to the extent possible, the same of those that caused the ground motions in a mirror database of recorded ground motions (RDB). We then utilize the conditional spectrum (CS) approach computed for a hypothetical rock site in Perugia, Italy, to select hazard consistent records separately from the RDB and SDB databases. A comprehensive set of comparisons of distributions of several IMs, computed from the selected sets, and of distributions of engineering demand parameters (EDPs), that such record sets induce on simple SDoF structures suggests the adequacy of using the proposed simulated ground motions for structural response analyses of structures on stiff and rock sites.

Explainable probabilistic deep learning framework for seismic assessment of structures using distribution‐free prediction intervals

AbstractA new probabilistic framework is proposed for providing a distribution‐free prediction interval (PI) of seismic responses required for various earthquake engineering applications. The framework overcomes the limitation of point prediction models and avoids the complexity of traditional probabilistic methods. The framework utilizes a few assumptions of probability distributions and requires no prior assumed statistical distribution for the PI. Ensemble probabilistic deep learning models (DLMs) are used to provide quality‐driven PIs of seismic responses for low‐ to mid‐rise buildings with limited irregularity. Considering these systems and ground motions with the aid of Monte Carlo simulation and nonlinear time‐history analysis (NLTHA), huge datasets are generated for training. To have an insight into the probabilistic DLM, explainable artificial intelligence techniques are used. The superiority of the proposed framework in quantifying uncertainties is validated by comparison with the conventional Bayesian method. In addition, its applicability is investigated by providing bounds of seismic fragility curves, life cycle cost, and resilience index obtained by NLTHA for a benchmark case study model. The results showed that the proposed framework is robust and outperforms the conventional Bayesian method in uncertainty quantification for the considered dataset.

Coseismic site response and slope instability using periodic boundary conditions in the material point method

This paper proposed the explicit generalized-α time scheme and periodic boundary conditions in the material point method (MPM) for the simulation of coseismic site response. The proposed boundary condition uses an intuitive particle-relocation algorithm ensuring material points always remain within the computational mesh. The explicit generalized-α time scheme was implemented in MPM to enable the damping of spurious high frequency oscillations. Firstly, the MPM was verified against finite element method (FEM). Secondly, ability of the MPM in capturing the analytical transfer function was investigated. Thirdly, a symmetric embankment was adopted to investigate the effects of ground motion arias intensity (Ia), geometry dimensions, and constitutive models. The results show that the larger the model size, the higher the crest runout and settlement for the same ground motion. When using a Mohr-Coulomb model, the crest runout increases with increasing Ia. However, if the strain-softening law is activated, the results are less influenced by the ground motion. Finally, the MPM results were compared with the Newmark sliding block solution. The simplified analysis herein highlights the capabilities of MPM to capture the full deformation process for earthquake engineering applications, the importance of geometry characterization, and the selection of appropriate constitutive models when simulating coseismic site response and subsequent large deformations.

Resonant Column Testing Procedure for Microbial-Induced Carbonate-Precipitated Sands

Abstract One of the fundamental inputs to site response analysis is the characterization of the dynamic properties of the soil, namely the shear modulus and material damping ratio. Because of soil’s nonlinear behavior, these properties change with induced shear strains, and modulus reduction and damping (MRD) curves have been proposed to capture that cyclic shear strain dependency. Microbial-induced carbonate precipitation (MICP) is a natural cementation process that induces the precipitation of calcium carbonate, bonding soil particles together. Laboratory experiments have demonstrated that MICP can improve the mechanical behavior of soils and mitigate their liquefaction potential. However, MRD curves have not been developed for MICP-treated soils, which hinders further evaluations of their suitability as a ground improvement technique for geotechnical earthquake engineering applications. To the best of our knowledge, this paper provides the first empirical study measuring the shear strain–dependent dynamic properties of MICP-treated sands for different cementation levels (i.e., lightly to heavily cemented). Outcomes from this work include empirical models of the maximum shear modulus and minimum shear strain damping ratio of MICP-treated clean sands and the corresponding mean MRD curves. The experimental program includes MICP-treated specimens of Ottawa 20-30 sand tested with a resonant column (RC) device incorporating a modified RC porous disk that (1) minimizes the disturbance between treatment and installation of MICP-treated samples prior to being tested in the RC device, (2) prevents slippage between the specimen and the modified RC porous disk during RC shearing, and (3) enables repeatability of the MICP-treated sample preparation. We find that the level of cementation influences the MRD curves of MICP-treated sands. Linear elastic and volumetric shear strain thresholds for MICP-treated sands are smaller than those for untreated sands, whereas the initial shear modulus for treated soils is larger than its counterpart for untreated sands.

A multiaxial inertial macroelement for deep foundations

A hyperplastic macroelement is proposed as an efficient means for simulating nonlinear and multi-directional soil-piles interaction in the nonlinear analysis of structures. The macroelement relates the generalised forces exchanged between the superstructure and the foundation to the corresponding displacements and rotations of the foundation raft. The inertial effects developing under dynamic loading can be reproduced by coupling the macroelement with the participating masses of the soil-foundation system, obtainable with a modal identification of the pile group. Failure conditions of the foundation are described by a hyper-ovoidal ultimate limit state surface in the force space, which is calibrated through standardised procedures. The plastic behaviour, controlled by a series of yield surfaces with kinematic hardening, produces the desired directional coupling and allows to model the cyclic response and the irreversible deformation of the foundation. The macroelement is implemented in the analysis framework OpenSees for a prompt use in earthquake engineering applications. Its predictions under monotonic loads are validated against experimental data and advanced, fully coupled numerical modelling, while its dynamic response is tested on a reference case study.

A DesignSafe earthquake ground motion database for California and surrounding regions

This article presents a ground motion database for California and its close surroundings (i.e. areas near the border in Nevada, Oregon, and Arizona) from earthquakes between 1999 and 2021. This data set includes events with magnitudes larger than 3.2 and focal depths less than 40 km, and it is available on DesignSafe. Ground motion records and events included in this data set are collected from 65 different seismic networks and processed with an automated software tool called gmprocess, which was developed by the United States Geological Survey (USGS). Path measures such as rupture distance and epicentral distance are computed, 5%-damped spectral accelerations, duration metrics, and other ground motion intensity measures (IMs) are provided for records that pass the quality assurance check performed by the gmprocess toolkit. The quality of processed ground motions is also screened by using outlier detection algorithms and a multiple wave-train arrivals identification algorithm. In addition, site metadata are provided, including wave velocity information (from proxy-based time-averaged shear-wave velocity for the top 30 m, Vs30, and from P- and S-wave measured velocity profiles when available), predominant frequency measured from microtremor-based horizontal-to-vertical ratios (mHVSR), and site-specific (high-frequency spectral decay) [Formula: see text] values computed from multiple ground motions recorded at sites when available. The final database contains 287,804 three-component ground motions recorded at 3709 stations from 2641 earthquakes with magnitudes and distances ranging from 3.2 to 7.2 and 0.15 to 335 km, respectively. This ground motion database contributes to advancing both engineering seismology studies and earthquake engineering applications in shallow crustal tectonic settings.

The “Glass Beads” Coupling Solution for Borehole and Posthole Accelerometers: Shaking Table Tests and Field Retrievability

AbstractBorehole accelerometers are designed to record strong ground movements at depth. They have become an important complement to surface accelerometers for seismic and earthquake engineering applications. Borehole accelerometers present several implementation challenges, including their coupling with the geological environment. One possible coupling solution is the use of small glass beads, which are placed inside the borehole casing with the purpose of filling in any empty space between the sensor and the casing walls. We carried out a test on a shaking table, over a wide range of peak ground accelerations (PGA from 0.17 to 1.64g), allowing the comparison of the signals between a surface accelerometer and a borehole accelerometer coupled through the glass beads. These tests show that there is almost no difference between the surface and borehole accelerometer signals between 0.5 and 25 Hz, and only very small differences outside this band (0.2–0.5 Hz and 25–80 Hz). Furthermore, experience from multiple vertical accelerometric arrays show that an installation using glass beads is “reversible”, that is, 30 yr after the initial installation it can still be possible to easily extract the accelerometers for repair or replacement, without any problems or damage to the sensors.

Application of Soil Moisture Active Passive (SMAP) Satellite Data in Seismic Response Assessment

The proven relationship between soil moisture and seismic ground response highlights the need for a tool to track the Earth’s surface soil moisture before and after seismic events. This paper introduces the application of Soil Moisture Active Passive (SMAP) satellite data for global soil moisture measurement during earthquakes and consequent events. An approach is presented to study areas that experienced high level of increase in soil moisture during eleven earthquakes. Two ancillary datasets, Global Precipitation Measurement (GPM) and Global Land Data Assimilation (GLDAS), were used to isolate areas that had an earthquake-induced increase in soil moisture from those that were due to hydrological processes. SMAP-based soil moisture changes were synthesized with seismic records developed by the United States Geological Survey (USGS), mapped ground failures in reconnaissance reports, and surface changes marked by Synthetic Aperture Radar (SAR)-based damage proxy maps. In the majority of the target earthquakes, including Croatia 2020, Greece 2020, Indonesia 2018, Taiwan 2016, Ecuador 2016, and Nepal 2015, a relationship between the SMAP soil moisture estimates and seismic events was evident. For these events, the earthquake-induced soil moisture response occurred in liquefaction-prone seismic zones. The New Zealand 2016 event was the only study region for which there was a clear inconsistency between ΔSMSMAP and the seismic records. The promising relationship between soil moisture changes and ground deformations indicates that SMAP would be a useful data resource for geotechnical earthquake engineering applications and reconnaissance efforts.

Geostatistical Framework for Estimation of VS30 in Data-Scarce Regions

ABSTRACTThe time-averaged shear-wave velocity in the upper 30 m (VS30) is widely used as a proxy for site characterization in building codes. Regional estimations of VS30 often use either slope-based, terrain-based, or geological approaches as a proxy. This technique has proven useful at a number of locations globally, and slope-based estimates formed the basis of the original global VS30 model implemented by the U.S. Geological Survey. Geostatistical models involve the study of potentially spatially correlated parameters. Modeling challenges arise when parameters are scarce or uncertain, and traditional geostatistical workflows cannot be implemented in all settings. In this study, the benefits of the spatial extents of VS30 proxies are used to supplement local data to implement a methodology for improving estimates using a multi-Gaussian Bayesian updating framework. This methodology is presented in the context of a data-scarce region, specifically, the Kathmandu Valley in Nepal. Using geostatistical approaches typically used by the petroleum industry, this article develops a novel practice-oriented framework for VS30 estimation that can be adapted for use on a region-by-region basis. This framework provides an informed estimate and assessment of the uncertainties in which quantification of VS30 is required in geotechnical earthquake engineering applications.

3D vision‐based out‐of‐plane displacement quantification for steel plate structures using structure‐from‐motion, deep learning, and point‐cloud processing

AbstractIn this paper, a novel accurate and economical 3D computer vision‐based framework is proposed to quantify out‐of‐plane displacements of steel plate structures. First, a sequence of image frames of the steel plate structures of interest is collected. Second, using image association, structure‐from‐motion, and multi‐view stereo algorithms, a 3D point cloud of the steel plate structures and their surroundings is created. Third, an efficient 3D object detection method based on convolutional neural networks is developed and implemented to identify the steel plate structures in the 3D point cloud. Last, the out‐of‐plane displacements of the steel plate structures are quantified using point cloud postprocessing algorithms. The proposed framework has been implemented on a steel plate damper and a full‐scale steel corrugated plate wall panel, which are commonly used in structural and earthquake engineering applications. The results indicate the developed framework can successfully localize the steel plate components in the 3D scene and accurately quantify the out‐of‐plane structural displacements with an average accuracy of ∼1 mm. The implementation shows the proposed framework can accurately and efficiently quantify the out‐of‐plane displacements of steel plate structures in realistic engineering applications.

On occurrence disaggregation of probabilistic seismic hazard

AbstractDisaggregation of probabilistic seismic hazard allows to quantify how much one or more earthquake scenarios contribute to the occurrence [exceedance] of a ground motion intensity measure threshold of interest (x) at the construction site. The scenario is usually defined in terms of magnitude (M), source‐to‐site distance (R), and possibly includes the standardized residual (ε) of the ground motion model considered in the hazard analysis. Analytically, in case occurrence is of interest, disaggregation provides the joint probability density function of or conditional on the event, that is, or . Occurrence disaggregation is important for a number of earthquake engineering applications, and it is typically addressed in the literature in an approximated manner, considering as the conditioning event , with being an arbitrary finite width of the interval. This short communication undertakes a deeper examination of occurrence disaggregation clarifying that: (i) no approximation is needed in the case of disaggregation in terms of magnitude and distance (i.e., when is sought); (ii) is theoretically degenerate, and as such, its approximation via finite can lead to misleading results; (iii) if is chosen coherently with the discretization of the domain used in the hazard integral, it leads to approximated , enabling the conclusion that occurrence disaggregation does not add information with respect to disaggregation.

Empirical Correlations between the Spectral Input Energy and Spectral Acceleration

ABSTRACT Correlating different intensity measures (IMs) enables the joint consideration of multiple IMs in various performance-based earthquake engineering applications, such as ground motion selection. The correlations between the spectral input energy and spectral accelerations have not been previously examined. This study develops a correlation model quantifying the uncertainties due to GMPE selection and a finite sample size. The results show a strong correlation between the spectral amplitudes at the same period, and the correlation decreases as the period difference increases. With the proposed model, ground motion characteristics reflected by the input energy can be explicitly considered in performance-based earthquake engineering applications.

Characteristics of earthquake source and ground motions in Northern Vietnam investigated through the 2020 Moc Chau M5.0 earthquake sequence

• The strike-slip rupture mechanism has been verified by the moment tensor inversion. • The isoseismal map of earthquake intensity has been constructed based on field survey. • This earthquake is associated with the activation of the rank II Da River fault. • The site condition of each station is determined using HVSR for further applications. • Earthquake engineering application issues of this area have been clearly addressed. On July 27, 2020, a shallow earthquake with a moment magnitude (M w ) of 5.0 occurred near Moc Chau in Northwestern Vietnam. Several shallow aftershocks, clustered in a small area, followed the mainshock. The mainshock caused damage to infrastructure in the source area, and considerable shaking was felt in many high buildings in surrounding cities. This event was the first recorded significant earthquake by a newly operational Vietnam broadband seismic network. The ground motions of this earthquake sequence were adequately recorded by this network. The strike-slip rupture mechanisms of the events were evaluated through moment tensor inversion analysis. After this earthquake, a field survey was conducted. A combined analysis of damage patterns and source mechanisms revealed that this earthquake sequence might be associated with the active right-lateral Da River fault. The site classification of each station that recorded data on the Moc Chau series was evaluated by computing the horizontal-to-vertical spectral ratio. This earthquake sequence was analyzed to evaluate the characteristics of the earthquake source and the seismic wave propagation in Northwestern Vietnam and to address the potential earthquake engineering applications of this information.

Editorial: Artificial Intelligence-Powered Methodologies and Applications in Earthquake and Structural Engineering

Editorial: Artificial Intelligence-Powered Methodologies and Applications in Earthquake and Structural Engineering

Bandgaps for flexural waves in infinite beams and plates with a periodic array of resonators

ABSTRACT The subject of seismic metamaterials, inspired from novel ideas in optics and acoustics, has attracted great attention in the last decade for potential applications in earthquake engineering. Simple structure systems, like beams and plates, with periodically attached mechanical resonators provide a simple physical model to interpret the existence of certain frequency bandgap in dispersion relations and to simulate the mechanism of flexural energy attenuation. In this work, we consider simple structure systems of beams and plates with periodically attached resonators. The resonator is composed of a spring, a damper and a mass attached along the beam direction. We utilize the Timoshenko beam model and the Mindlin plate theory to incorporate the shear effect. The plane wave expansion method together with the Bloch theorem is used to expand the governing field into an eigenvalue problem of an infinite complex system, allowing us to characterize the band structures of the dispersion relations. Local resonance and Bragg bandgaps are identified and examined. The effect of thickness ratios, the damping ratio and the shear modulus are exemplified to demonstrate how these factors will affect the formation of bandgaps. This formulation demonstrates a feasibility that a periodic array of mechanical resonators together with suitable material and geometric parameters of beams and plates can be designed to tune with the dispersion behavior in the control of flexure waves. This study may open up new potential in the control of wave propagation in complex continuum systems through the interaction of adequately designed resonators.