Original Ground Motion Research Articles

An Empirical Spatiotemporal Input Model for Regional Seismic Ground Motions

The development of the economy and technology has attracted greater attention to regional seismic resistance, a key technology for which is the simulation of regional seismic ground motions. This study proposes a simple and highly operable method for the generation of regional seismic ground motions, which includes five steps. First, a seismic ground motion recorded at a seismic station is selected as the original ground motion. The original ground motion is then decomposed into eight sub-signals by wavelet packet transform (WPT). Next, the eight sub-signals are adjusted by the frequency attenuation model. The generated ground motion is then amplitude modulated by the peak ground acceleration (PGA) attenuation model. The frequency attenuation model and the PGA attenuation model are obtained by fitting seismic data from the 1999 Chi-Chi earthquake. Finally, the propagation of the ground motion is considered. The proposed method can provide a time history of seismic ground motions for any point in a region. To investigate the rationality and applicability of the proposed method, the characteristics of the generated seismic ground motions are compared with those of true seismic ground motions. The average error of the PGA of the generated ground motions and the true ground motions is 0.072[Formula: see text][Formula: see text]; thus, the proposed method is suitable for use in small-scale regions. Furthermore, 4-story and 15-story structures are used for structural response analysis. The story drift ratio error between the generated ground and true ground motions is found to be within an acceptable error range. The proposed method provides a new way to generate ground motions for regional seismic investigation.

Automated parameterization of velocity pulses in near‐fault ground motions

AbstractProper parameterization of near‐fault ground motions is of critical importance in earthquake engineering, and this process is traditionally performed by directly fitting an analytical pulse model to the original motion. Yet such a process is usually limited by the trial‐and‐error procedure, which is strongly dependent on the initial guesses and may converge to local rather than global minimums. In this study, we propose a progressive (step‐by‐step) iterative approach that can achieve a fully automated parameterization of the velocity pulse contained in near‐fault motions. Assuming that a velocity pulse can be characterized by a pulse model with four key parameters, the approach is conducted by iteratively matching the pulse model to the smoothed ground motion, and the parameterized pulse is analytically derived by best fitting to the smoothed motion not only in the time domain but also in the spectral domain. Specifically, the velocity time history of interest is initially smoothed by a moving average filter so that the low‐frequency content can be filtered out of the original motion, from which the pulse amplitude as well as its epoch is accordingly determined. The coherent velocity pulse is then progressively extracted by performing the nonlinear least‐square‐fitting of the pulse model to the filtered low‐frequency content, during which the remaining parameters, that is, the pulse period, the number of cycles and the phase of the pulse, are estimated successively. Finally, the above procedure is applied repeatedly to the original ground motion by changing an empirical factor controlling the extent of smoothing of the motion so that convergence to local minimums that frequently occurs in the trial‐and‐error procedure could be largely avoided, and best match of the spectrum of the extracted pulse to that of the original motion can be acquired. Fitting quality of the velocity pulses is examined by comparing with existing methods. Prospective applications of the proposed procedure include the stochastic simulations of near‐fault ground motions and parametric investigations of the influence of velocity pulses on various engineered structures.

An ameliorative duration of pulse-like ground motion records for nonlinear structural analysis

The near-field pulse-like ground motion contains obvious long-period and large pulse characteristic, which makes the nonlinear response of the structure exceed the general prediction of seismic response results. In order to reserve the obvious pulse component while selecting the strong ground motion duration part of the near-field pulse-like ground motion record without causing a large loss of spectral characteristics, the paper proposes an ameliorative truncated strong ground motion duration—useful duration. It is proposed based on the elastic acceleration response spectrum. Its principle is to make the acceleration response spectrum corresponding to useful duration and that corresponding to the original ground motion record as close and consistent as possible. The research results show that from the perspective of elastic response spectrum matching, the response spectrum matching error corresponding to useful duration is smaller than that of significant duration. In addition, the five types of periods corresponding to useful duration are closer to the period of original ground motion. Ulteriorly, from the displacement ratio spectrum of the elasto-plastic single-degree-of-freedom system, it is found that the average value of the elasto-plastic displacement ratio spectrum of useful duration and the original ground motion record is 0.98, which is very close to 1.00. It shows that the truncated ground motion can be used to replace the original ground motion to fully predict the nonlinear response of typical structures. Additionally, the effectiveness is further verified in the nonlinear response of three reinforced concrete frame structures, and the nonlinear response results of the structures are calculated using the time-history of useful duration, with less discreteness.

Convolutional variational autoencoder for ground motion classification and generation toward efficient seismic fragility assessment

AbstractThis study develops an end‐to‐end deep learning framework to learn and analyze ground motions (GMs) through their latent features, and achieve reliable GM classification, selection, and generation of simulated motions. The framework is composed of an analysis workflow that transforms and reconstructs GMs through short‐time Fourier transform (STFT), encodes and decodes their latent features through convolutional variational autoencoder (CVAE), and classifies and generates GMs by grouping and interpolating latent variables. A benchmark study is established to confirm the minor difference between original GMs and the corresponding reconstructed accelerograms. The encoded latent space reveals that certain latent variables are directly linked to the dominant physical features of GMs. Resultantly, clustering latent variables using the k‐means algorithm successfully classifies GMs into different groups that vary in earthquake magnitude, soil type, field distance, and fault mechanism. By linearly interpolating two parent latent variables, simulated GMs are generated with consistent class information and matching response spectra. Furthermore, seismic fragility models are developed for a steel frame building and a concrete bridge using different sets of GMs. Using five classified, top‐ranked motions, regardless of recorded or simulated accelerograms, can achieve reasonable and efficient fragility estimates compared to the case that adopts 230 GMs. The proposed deep learning framework addresses two compelling questions regarding seismic fragility assessment: How many GMs are sufficient and what types of motions should be selected.

Loss estimation of curved bridges considering the incidence angle and spatial variability of earthquake ground motion

The consideration for the angle of seismic incidence (ASI) is critical for the seismic design of curved bridges. Previous studies indicate that the effect of ASI weakens as the damage develops throughout the structure. Since curved bridges often extend over long distances and cross varied geological conditions, the spatial variability of earthquake ground motions (SVEGMs) may significantly affect their responses. However, the concurrent effects of ASI and SVEGM on the system performance of curved bridges have yet to be explored, and there are currently few reliable methodologies for conducting relevant research. To this end, this study takes the monetary repair costs as the metric for representing the entire performance of the bridge, and the impact of ASI on the total loss of long multi-span curved bridges considering the spatially variable nature of earthquake ground motions is investigated in a probabilistic manner. A multi-directional loss assessment approach is derived based on the extension of an existing loss methodology, and an already constructed, 638-m-long, twelve-span prestressed concrete curved bridge is selected as the test-bed. Twenty non-stationary and spectrum-compatible spatially variable earthquake ground motion (SVEGM) pairs are generated based on the Eurocode 8 (EC8) design spectrum and soil conditions of the bridge site for each ASI. Three seismic input motion scenarios are assumed according to the characteristics of SVEGMs, and variabilities of bridge loss with respect to the ASI under the scenario earthquakes are investigated for each excitation case. On these bases, the original ground motion records are amplified by 2 and 4 times, respectively, and 4320 (3 earthquake intensity levels × 3 excitation scenarios × 24 ASIs × 20 ground motion pairs) nonlinear response-history analyses (NRHAs) are implemented to identify the influence induced by the ground motion intensity on the loss variability. Results indicate that the impacts of ASI and SVEGMs weaken as the ground shaking intensity increases and the damage accumulates in the structure.

Multi-pulse decomposition for nonlinear seismic analysis of structural systems

Large sympathetic, resonance-like, structural behaviour to earthquake excitations with analogous frequency content often plays a critical role in determining its maximum seismic response. Earthquake excitation typically contains a broad spectrum of non-stationary frequency content, wave packets, which are difficult to observe from the recorded time series. Therefore, identifying the root cause of large responses (which act sympathically with the input but do not achieve full-resonance) of a structure is problematic. Hence, this paper proposes a new multi-pulse decomposition method of ground motions, through which components of a ground motion within a specific period range are determined. In this method, a ground motion is approximated with a Gauss-Fourier wave packet series. The decomposed components, wave packets, contain information about its time-position, frequency, amplitude, pulse width and phase angle. Unlike the Ricker (Morlet) wavelet the Gauss-Fourier wave packet is not limited to symmetrical pulses. One ensemble of 40 near-fault ground motions and one ensemble of 44 far-fault ground motions are used to demonstrate the application and efficiency of the proposed method. The method is shown to be precise in reconstructing the original ground motion using its decomposed components. It is also concluded that the method is accurate in replication of elastic and inelastic response spectra of ground motions within a specific period range. It is demonstrated that for some structure/ground motion combinations, only a few Gauss-Fourier components are required to faithfully describe response behaviour. This highlights that, for these systems, most of the recorded earthquake time series acts like noise on a much simpler wave-packet signal.

Improved endurance time analysis for seismic responses of concrete dam under near-fault pulse-like ground motions

This study aimed to modify the Endurance Time Analysis (ETA) method using a novel prediction model to evaluate the seismic demands of concrete dams subjected to near-fault pulse-like excitations. To this end, a two-dimensional concrete gravity dam–reservoir–foundation system was employed in the numerical model. Twenty as-recorded pulse-like (original) ground motions were decomposed into representative low- and high-frequency components using the continuous wavelet transform method. The contributions of these components toward the dynamic responses under different intensity measures were investigated, based on an Incremental Dynamic Analysis (IDA). The effect of Peak Ground Acceleration (PGA) on the responses was considered, and the response prediction model was further modified, including some dimensionless parameters. Coupled with the response results under residual motion obtained via the ETA method, the estimated response under pulse-like motion can be derived using the proposed ETA prediction method. Based on a comparison of the average results of the IDA and improved ETA methods under pulse-like motion, the average error was found to be approximately 10%. Furthermore, 10 as-recorded earthquakes were used to verify the proposed procedure. The results revealed that the ETA method can be applied to pulse-like motions, which reduces the calculation time and contains pulse properties.

A hybrid characterization framework for structure-significant pulse-like features in ground motions

In earthquake engineering and engineering seismology, it is fundamental to identify pulse-type ground motions and characterize their pulse-like features, as seismologists relate them to the fault rupture and engineers relate them to the structural response. Pulse-type ground motions have been observed to exhibit peculiar attributes both in the response spectrum and in the time series. The existing pulse identification methods rarely leverage the distinct feature in response spectrum sufficiently, and consequently only the mimicry in time series sometimes loses the impulsive effect of the original ground motion on the structures. Based on the self-similarity revealed by dimensional analysis, the current study developed a hybrid pulse characterization framework that starts from subregion-to-subregion similarity matching in response spectral space and finally achieves resemblance in time domain, for pulse-like features in either ground velocity or acceleration. The framework directly identifies the pulses that dominate the structural response. The framework is versatile in terms of allowing various pulse waveforms, switching among different types of response spectra, and coping with the ground motions that contain multiple structure-significant pulse-like features.

Seismic response of high-speed railway simple-supported girder track-bridge system considering spatial effect at near-fault region

High-speed railway are mostly long-line structures stretching for several kilometers, and the seismic response analysis of individual stations cannot accurately determine the overall safety of the structure. In this study, 53 stations with rupture distances less than 50 km were selected from the Taiwan Chichi earthquake, and they were divided into three regions according to the rupture location. The seismic response and spatial distribution characteristics of High-Speed Railway Track-Bridge System (HSRTBS) under the original ground motion as well as the amplitude scaling ground motion of each station were investigated. The results showed that: (1) Under the original ground motion, the average seismic response of HSRTBS in the middle region (MR) is the largest, while the one in the forward region (FR) is the smallest. The displacements and internal forces of the structural components follow a logarithmic relation with Rrup/a (ratio of rupture distance to peak acceleration) and they tend to be stable when Rrup/a is larger than 100. (2) There is a correlation between the seismic response of HSRTBS and the fault angle. In FR and MR, the seismic response of the HSRTBS is the smallest at a fault angle 90°, while, at the backward region (BR) the minimum response of the HSRTBS is noted for fault angle 45°.

Effects of the Predominant Pulse on the Inelastic Displacement Ratios of Pulse-Like Ground Motions Based on Wavelet Analysis

Having a predominant pulse is the main feature for pulse-like ground motions differing from others. To investigate the influence of the predominant pulse on the inelastic displacement ratios of pulse-like ground motions, the wavelet analysis method is used to extract the predominant pulse. The results indicate that the inelastic displacement ratios of the pulse-removed parts obtained by subtracting the extracted pulse from the original pulse-like ground motions are close to the results of non-pulse-like ground motions. The ratio of the energy of the extracted pulse to the energy of the original ground motion is used to represent the pulse intensity. The results indicate that the pulse period determines the locations in which the inelastic displacement ratios would have noticeable increments, and the pulse intensity determines the degree of the increments. Besides, the effects of five commonly used parameters (PGV, PGD, PGV/PGA, Arias intensity Ia, and soil condition) on the inelastic displacement ratios of pulse-like ground motions and their relations to the pulse period and the pulse intensity are studied. Finally, a new model, in which the influence of pulse intensity is considered, to predict the inelastic displacement ratios of pulse-like ground motions is proposed.

A wavelet-based approach for truncating pulse-like records

The paper discusses the fast seismic performance assessment of structures using truncated pulse-like ground motion records. It is shown that it is possible to truncate a pulse-like signal using a novel wavelet-based definition that identifies the duration of the predominant velocity pulse. The truncated time-history is still able to efficiently reproduce the increased seismic demand that near-field records typically produce. Substituting the original ground motion with the truncated signal, can significantly accelerate structural analysis and design. The truncated signal is the part of the original accelerogram that coincides with the duration of the predominant pulse, which is identified using a wavelet-based procedure, previously proposed by the authors. Elastic and inelastic response spectra and nonlinear time-history analyses for SDOF (single-degree-of-freedom) systems are first studied. Subsequently a nine-storey steel frame is examined in order to demonstrate the performance of the proposed approach on a multiple-degree-of-freedom system. The proposed approach is found very efficient for pulse-like ground motions, while it is also sufficient for many records that are not characterized as such.

Quantitative identification of near-fault pulse-like ground motions based on variational mode decomposition technique

Near-fault ground motions have distinctive features different from typical far-field ground motions. The approach detects near-fault pulse-like ground motion based on the variational mode decomposition (VMD) technique. First, the original ground motion can be decomposed into a series of intrinsic mode functions (IMFs) with the VMD technique. The low-frequency signal is obtained by synthesizing the IMFs with the center frequency lower than an adaptive frequency threshold proposed in this paper. Second, the pulse signal and the pulse period are obtained from the best-fitting low-frequency signal by the peak-point method. Finally, 596 records are utilized to calibrate the classification standard. It is concluded that these ground motions with relative energy indicator greater than 0.27 can be classified as pulse-like. The approach is further used to identify the pulse-like ground motion with a significant velocity pulse at the beginning, which may be caused by the forward-directivity effects.

Automated detection of velocity pulses in ground motions based on adaptive similarity search in response spectrum

Ground motions with strong pulse-like features are of special concern for structural engineers. Detection and characterization of the pulse-like features are fundamental in seismic hazard analysis, ground motion selection/simulation, and structural analysis. From the perspective of earthquake engineering, a pulse-like feature in a ground motion is of significance only if it dominates certain structural response to the ground motion. In order to identify such kind of pulse-like features, the current study establishes the so-called response spectra associated with various velocity waveforms through dimensional analysis of structural response to a generic velocity pulse. The congruence and the shifts between the ΠPV and the pseudo-velocity response spectra associated with identical velocity waveform allow automatic detection of velocity pulses through adaptive similarity search in bi-logarithmic spectrum curves. Following this approach, velocity pulses are detected and characterized in hundreds of historical ground motion records. The majority of the detection results are comparable to the velocity pulses identified through wavelet analysis in literature, while significant differences are noticed for a few cases. Further investigation shows our approach is able to capture the velocity pulse that retains the impulsive effect of the original ground motion on structures and thus provides a worthy supplement to the existing pulse identification methods.

Seismic analysis of steel concentrically braced frames with consideration of fling-step effect

Near-field earthquakes with a fling-step effect cause a permanent displacement at the end of the displacement record, which can affect the response of structures. Fling step is induced by permanent static displacement of the ground that is noticeable in strike–slip faults. This effect is usually omitted from the main record using the standard processing methods and, thus, the fling-step effect cannot be considered to obtain seismic response of different structures. To investigate the effect of near-field earthquakes with the fling-step effect, 14 original (non-filtered) ground motions with different fling displacement are selected. The selected records are applied to special concentrically braced frame (SCBF) structures with 3, 6, 9 and 12 stories to cover most structural periods. To investigate seismic response of the structures, two main parameters including lateral displacement and story–drift ratio are selected. A nonlinear time history analysis is used to obtain the displacement and story–drift ratio responses. The results show that the fling-step effect has no significant impact on any of the considered structures, and its effect can be neglected in case of special concentrically braced frame structures.

Seismic response of mid-rise steel MRFs: the role of geometrical irregularity, frequency components of near-fault records, and soil-structure interaction

Near-fault (NF) ground motions may induce higher demands on structures compared to far-fault records. This is often due to the velocity pulses, called forward-directivity pulses, which are commonly observed in the fault-normal ground motion components. Hence, some researchers accept such pulses as the representative of pulse-like motions. This paper is devoted to studying the nature of NF events by evaluating their impact on the structural response. To this end, a group of ten-story steel moment-resisting frames with regular and setback configuration were subjected to NF pulse-like records. The records were decomposed to distinct frequency components, including the extracted pulses (low-frequency components) and the residual (high-frequency) parts. The effect of TP/T1 (the ratio of pulse period to the first-mode structural period) was also evaluated by classifying the earthquake records into four groups, indicating TP/T1≈0.5, 1, 2, and > 2. Furthermore, soil-structure interaction (SSI) was considered through the Beam on the Nonlinear Winkler Foundation theory. Incremental dynamic analyses were conducted under the action of the original ground motions, extracted pulses and residual parts. The results deeply question the possibility of using directivity pulses as the representative of original excitation in dynamic time history analyses, especially for setback structures. The residual component plays an important role in seismic response of such structures in most cases of TP/T1. The considerable effect of this frequency component is also observed in regular structures when TP/T1≈1. Meanwhile, SSI can increase structural drift demands, especially in setback structures, when subjected to NF pulse-like records.

A Modified Algorithm to identify the strongest velocity pulse in three orthogonal components of ground motions

An algorithm for the identification of the strongest velocity pulse is proposed based on wavelet analysis, which considers three orthogonal components of the ground motion. According to the NGA-West2 database, 6288 strong motion records were analyzed, and further, the classification results were compared with those obtained from the existing identification method of Shahi and Baker (2014) [1]. Among the results, our algorithm classified more records as pulse-like ground motion and has expanded the pulse identifications. In other words, the proposed algorithm can reflect the three orthogonal characteristics of the original ground motion records. Besides, the relationships of pulse amplitude, period with magnitude and fault distance were analyzed. Based on the analysis of near-fault ground motions containing velocity pulses, the following conclusions can be made: (1) The peak of the strongest velocity pulse gradually reduces with the increase of the fault distance, while gradually rises with the increase of the magnitude. Comparing with the fault normal or fault parallel orientation component, the strongest velocity pulse component has larger energy and smaller peak velocity. (2) For the relationship between the velocity pulse period and magnitude, results show that when Mw < 7, the predicted value of our model is larger than that of the Shahi and Baker (2014) [1] model; however when Mw ≥ 7, the prediction results from these two models are fairly close. The difference between the two models decreases as magnitude increases. That is because the strongest velocity pulses occupied more in the horizontal direction under a large magnitude.

Comparison of Near-Fault Velocity Pulse-Like Ground Motions from the 2018 Mw 6.4 Hualien, Taiwan, Earthquake with the Next Generation Attenuation (NGA)-West2 Ground-Motion Models and Directivity Models

ABSTRACTPulse-like ground motions cause severe damage in structures at certain periods. Hence, pulse effects need to be considered during probabilistic seismic hazard analysis and seismic design in the near-fault region. Traditional ground-motion models used to quantify the hazard posed by pulse-like ground motions may underestimate them, but they are relatively suitable for describing the residual ground motions after extracting pulses. Nevertheless, the applicability of Next Generation Attenuation-West2 Project (NGA-West2) models to pulse and residual ground motions has not been evaluated. Moreover, the applicability of recently developed directivity models, including the Shahi and Baker (2011; hereafter, SB2011), Chang et al. (2018; hereafter, Chang2018), and Rupakhety et al. (2011; hereafter, Rupakhety2011) models, has not been investigated for this event. Here, based on the abundance of pulse-like ground motions recorded during the Mw 6.4 Hualien earthquake, the applicability of NGA-West2 models and directivity models was quantitatively evaluated. In summary, (1) The applicability of NGA-West2 models to the observed original and residual ground motions varies significantly at different periods. The suggests that NGA-West2 models overestimate the original and residual ground motions for short periods (T&amp;lt;1.0 s), but are suitable for describing the residual ground motions yet underestimate the original ground motions for long periods (T≥1.0 s). (2) Pulse periods and amplification bands due to pulses are unusually larger than previous events. Similar to the Chang2018 model, the plateau of this event starts and ends at the periods of 0.70 and 1.1 times the pulse period. However, the Chang2018 and SB2011 models underestimate the constant ordinate of this plateau. Spectral ordinates of the spectral shape curve due to pulses for the short period (∼Tn&amp;lt;1.3 s) are smaller than the predictions from the Rupakhety2011 model. The trend was reversed for long periods (∼Tn&amp;gt;3.0 s). Compared with the Rupakhety2011 model, the peak location of the spectral shape curve is shifted to the long period. These results will be helpful for updating these models in the near future.

Summarized IDA curves by the wavelet transform and bees optimization algorithm

Incremental dynamic analysis (IDA), as an accurate method to evaluate the parameters of structural performance levels, requires many non-linear time history analyses, using a set of ground motion records which are scaled to different intensity levels. Therefore, this method is very computationally demanding. In this study, a new method is presented to estimate the summarized (16%, 50%, and 84% fractiles) IDA curves of a first-mode dominated structure using discrete wavelet transform and bees optimization algorithm. This method reduces the number of required ground motion records for the prediction of the summarized IDA curves. At first, a subset of first list ground motion records is decomposed by means of discrete wavelet transform which have a low dispersion estimating the summarized IDA curves of equivalent SDOF system of the main structure. Then, the bees algorithm optimizes a series of factors for each level of detail coefficients in discrete wavelet transform. The applied factors change the frequency content of original ground motion records which the generated ground motions records can be utilized to reliably estimate the summarized IDA curves of the main structure. At the end, to evaluate the efficiency of the proposed method, the seismic behavior of a typical 3-story special steel moment frame, subjected to a set of twenty ground motion records is compared with this method.

An improved defining parameter for long-period ground motions with application of a super-tall building

This paper focuses on defining long-period ground motions and analyzing the effect of them on building structures. Increasing attention has been paid to the effect of long-period ground motions on building structures. However, long-period ground motions are not quantitatively defined due to the lack of numerous ground motion investigations. In this paper, based on the statistical analysis of 39,744 ground motions, an improved frequency-domain defining parameter, namely βl, is firstly proposed to identify long-period ground motions without the necessity of seismological information. Then, based on the definition, long-period ground motions are selected for the time history analysis of a super-tall building structure. The structural responses are obtained under three scenarios: the original ground motion record, the ground motions scaled to minor earthquakes and major earthquakes. To assess the validity of the proposed parameter βl in identifying long-period ground motions, correlation between the structural response and the defining parameter is obtained.

An empirical approach of accounting for the amplification effects induced by near-fault directivity

How to incorporate near-fault directivity effect into the ground motion prediction equations (GMPEs) is an active research topic in recent years. Many studies demonstrated that the directivity could amplify the spectral accelerations (S a) in a range close to the pulse period (T p). This study intends to propose a new empirical narrowband model for including such effects. The ratio of the S a from the original ground motions to the S a from the residual ground motions is employed to quantify this amplification effects, under the assumption that the S a from the residual ground motions are well predicted by the non-directivity-included GMPEs. Differing from previous models that suggested the amplified S a arise in a smooth bell-shaped pattern, the proposed model assumes that the peak amplification occurs as a plateau in the intermediate range. This plateau, having a constant ordinate of about 3.31, starts and ends at the period of 0.75 and 1.09 times the T p, respectively; while in the other ranges the amplification decreases away from the two sides of this plateau. To further simplify the proposed model, two additional controlling periods are derived to determine the period range with significant amplification effects. Then the proposed model is expressed in the form of a piecewise function comprising five line segments. Finally, it is concluded that the non-directivity-included GMPEs, when multiplying the proposed model, can reasonably approximate the S a of the original ground motions in the statistical sense within the T p-related range.