Seismic response and correlation analysis of a pile-supported wharf to near-fault pulse-like ground motions

Earthquake investigations have shown that near-fault pulse-like(NF-P)ground motions have unique characteristics compared to near-fault non-pulse-like(NF-NP)and far-field(FF)ground motions.It is necessary to study the seismic response of pile-supported wharf(PSW)structures under NF-P ground motions.In this study,a three-dimensional finite element numerical model is created to simulate a PSW.By imparting three types of ground motion,the engineering demand parameters(EDPs)of PSW under NF-P ground motions were analyzed and compared,in which EDPs are the maximum displacement and bending moment of the piles.Twenty intensity measures(IMs)were selected to characterize the properties of ground motions.The correlation between IMs and EDPs was explored.The results show that the piles present larger displacement and bending moment under NF-P ground motions compared to NF-NP and FF ground motions.None of the IMs have a high correlation with EDPs under NF-P ground motions,and these IMs are more applicable to FF ground motions.The correlation coefficients between EDPs and IMs under three types of ground motion were obtained,which will provide a valuable reference for the seismic design of PSWs.

Response spectrum of seismic design code for zones lack of near-fault strong earthquake records

It was shown from the study on the recently near-fault earthquake ground motions that the near-fault effects were seldom considered in the existing Chinese seismic code. Referring to the UBC97 design concept for near-fault factors, based on the collected world-widely free-site records of near-fault earthquakes ground motions classified by earthquake magnitude and site condition, the attenuation relationship expressions of the acceleration spectrum demand at the key points within the long period and moderate period were established in term of the earthquake magnitude and the site condition. Furthermore, the near-fault factors＇ expressions about the earthquake magnitude and the fault distance were deduced for the area lack of near-fault strong earthquake records. Based on the current Chinese Building Seismic Design Code, the near-fault effect factors and the modified design spectral curves, which were valuable for the seismic design, were proposed to analyze the seismic response of structures.

Spatial distribution of near-fault ground motion

Near-fault strong ground motion of strike-slip and dip-slip of vertical and inclined rectangular fault in half-space and layered half-space is analyzed by dislocation source model. The Fourier spectra ratio of ground motion is adopted to study the characteristics of near-fault ground motion. For both slip models, near-fault strong ground motion with high amplitude is located in a narrow belt area along the projection of the fault on the ground and mainly controlled by the sub-faults nearby. Directivity of strike-slip fault is more dominant in long period for components perpendicular to the fault, and more dominant in long period for components parallel to the fault for dip-slip fault. The deeper the location of the source is, the more slowly the amplitude of ground motion attenuates. There is obvious hanging wall effect in ground motion of inclined fault, and the spatial distribution of ground motion is asymmetric which coincides with observational data. Finally, a fitting function of spatial distribution for near-fault ground motion is proposed and compared with near source factors of the 1997 Uniform Building Code of USA.

Effect of near-fault earthquake on bridges:lessons learned from Chi-Chi earthquake

The objective of this paper is to describe the lessons learned and actions that have been taken related to the seismic design of bridge structures after the Chi-Chi,Taiwan earthquake.Much variable near-fault ground motion data was collected from the rupture of Chelungpu fault during the Chi-Chi earthquake,allowing the seismic response of bridge structures subjected to these near-fault ground motions to be carefully examined.To study the near-fault ground motion effect on bridge seismic design codes,a two-level seismic design of bridge structures was developed and implemented.This design code reflects the near-fault factors in the seismic design forces.Finally,a risk assessment methodology,based on bridge vulnerability,is also developed to assist in decisions for reducing seismic risk due to failure of bridges.

Response spectra for nuclear structures on rock sites considering the near-fault directivity effect

Near-fault ground motions, potentially with large amplitude and typical velocity pulses, may significantly impact the performance of a wide range of structures. The current study is aimed at evaluating the safety implications of the near-fault effect on nuclear power plant facilities designed according to the Chinese code. To this end, a set of near- fault ground motions at rock sites with typical forward-directivity effect is examined with special emphasis on several key parameters and response spectra. Spectral comparison of the selected records with the Chinese and other code design spectra was conducted. The bi-normalized response spectra in terms of different comer periods are utilized to derive nuclear design spectra. It is concluded that nuclear design spectra on rock sites derived from typical rupture directivity records are significantly influenced both by the earthquake magnitude and the rupture distance. The nuclear design spectra specified in the code needs to be adjusted to reflect the near-fault directivity effect of large earthquakes.

Investigation of effect of near-fault motions on substandard bridge structures

The objective of this study was to investigate the effects of near-fault ground motions on substandard bridge columns and piers. To accomplish these goals, several large scale reinforced concrete models were constructed and tested on a shake table using near- and far-field ground motion records. Because the input earthquakes for the test models had different characteristics, three different measures were used to evaluate the effect of the input earthquake. These measures are peak shake table acceleration, spectral acceleration at the fundamental period of the test specimens, and the specimen drift ratios. For each measure, force-displacement relationships, strains, curvatures, drift ratios, and visual damage were evaluated. Results showed that regardless of the measure of input or response, the near-fault record generally led to larger strains, curvatures, and drift ratios. Furthermore, residual displacements were small compared to those for columns meeting current seismic code requirements.

Design spectra including effect of rupture directivity in near-fault region

In order to propose a seismic design spectrum that includes the effect of rupture directivity in the near-fault region, this study investigates the application of equivalent pulses to the parameter attenuation relationships developed for near-fault, forward-directivity motions. Near-fault ground motions are represented by equivalent pulses with different waveforms defined by a small number of parameters （peak acceleration, A, and velocity V; and pulse period, Tv）. Dimensionless ratios between these parameters （e.g., ATv/V, VTv/D） and response spectral shapes and amplitudes are examined for different pulses to gain insight on their dependence on basic pulse waveforms. Ratios of ATv/V, VTv/D, and the ratio of pulse period to the period for peak spectral velocity （Tv-p） are utilized to quantify the difference between rock and soil sites for near-fault forward-directivity ground motions. The ATv/Vratio of recorded near-fault motions is substantially larger for rock sites than that for soil sites, while Tvp/Tv ratios are smaller at rock sites than at soil sites. Furthermore, using simple pulses and available predictive relationships for the pulse parameters, a preliminary model for the design acceleration response spectra for the near-fault region that includes the dependence on magnitude, rupture distance, and local site conditions are developed.

Representation of near-fault pulse-type ground motions

Near-fault ground motions with long-period pulses have been identified as critical in the design of structures. To aid in the representation of this special type of motion, eight simple pulses that characterize the effects of either the flingstep or forward-directivity are considered. Relationships between pulse amplitudes and velocity pulse period for different pulses are discussed. Representative ratios and peak acceleration amplification can exhibit distinctive features depending on variations in pulse duration, amplitude and the selected acceleration pulse shape. Additionally, response spectral characteristics for the equivalent pulses are identified and compared in terms of fixed PGA and PGV, respectively. Response spectra are strongly affected by the duration of pulses and the shape of the basic pulses. Finally, dynamic time history response features of a damped SDOF system subjected to pulse excitations are examined. These special aspects of pulse waveforms and their response spectra should be taken into account in the estimation of ground motions for a project site close to a fault.

An investigation of the influence of near-fault ground motion parameters on the pile's response in liquefiable soil

The performance of a soil-pile system can be significantly influenced by ground motion parameters. However, few research efforts have been performed to provide a complete description of the influence of key ground motion parameters on the pile’s behavior in liquefiable soil. In this study, a three-dimensional finite element（FE） model, incorporating a multisurface plasticity solid-fluid fully coupled formulation soil constitutive model, is developed and calibrated based on centrifuge test data. Seventy-two near-fault non-pulse-like（NF-NP） and seventy-two near-fault pulse-like（NF-P） ground motion records are studied with the calibrated FE model to distinguish the effects of several common ground motion parameters soon afterwards. Base on the parametric study results, a simple index, RPGV/PGA（i.e., the ratio of peak ground velocity（PGV） to peak ground acceleration（PGA））, shows its capability on characterizing the pile behavior under both NF-NP and NF-P ground motions. Furthermore, two equations are developed to characterize the relationships between the RPGV/PGA as well as the maximum pile’s moments and displacements. In general, this study can be helpful to gain new insights on the influence of typical index parameters for near-field ground motions on the response of the pile foundation in liquefiable soil.

Near-fault directivity pulse-like ground motion effect on high-speed railway bridge

The vehicle-track-bridge(VTB)element was used to investigate how a high-speed railway bridge reacted when it was subjected to near-fault directivity pulse-like ground motions.Based on the PEER NAG Strong Ground Motion Database,the spatial analysis model of a vehicle-bridge system was developed,the VTB element was derived to simulate the interaction of train and bridge,and the elasto-plastic seismic responses of the bridge were calculated.The calculation results show that girder and pier top displacement,and bending moment of the pier base increase subjected to near-fault directivity pulse-like ground motion compared to far-field earthquakes,and the greater deformation responses in near-fault shaking are associated with fewer reversed cycles of loading.The hysteretic characteristics of the pier subjected to a near-fault directivity pulse-like earthquake should be explicitly expressed as the bending moment-rotation relationship of the pier base,which is characterized by the centrally strengthened hysteretic cycles at some point of the loading time-history curve.The results show that there is an amplification of the vertical deflection in the girder's mid-span owing to the high vertical ground motion.In light of these findings,the effect of the vertical ground motion should be used to adjust the unconservative amplification constant 2/3 of the vertical-to-horizontal peak ground motion ratio in the seismic design of bridge.

Improving the performance of conventional base isolation systems by an external variable negative stiffness device under near-fault and long-period ground motions

Recent studies have shown that base-isolated objects with long fundamental natural periods are highly influenced by long-period earthquakes. These long-period waves result in large displacements for isolators, possibly leading to exceedance of the allowable displacement limits. Conventional isolation systems, in general, fail to resist such large displacements. This has prompted the need to modify conventional base isolation systems. The current work focuses on the development of an external device, comprising a unit of negative and positive springs, for improving the performance of conventional base isolation systems. This unit accelerates the change in the stiffness of the isolation system where the stiffness of the positive spring varies linearly in terms of the displacement response of the isolated objects. The target objects of the present study are small structures such as computer servers, sensitive instruments and machinery. Numerical studies show that the increase in the damping of the system and the slope of the linear function is effective in reducing the displacement response. An optimal range of damping values and slope, satisfying the stability condition and the allowable limits of both displacement and acceleration responses when the system is subjected to near-fault and long-period ground motions simultaneously, is proposed.

Identification of acceleration pulses in near-fault ground motion using the EMD method

In this paper, response spectral characteristics of one-, two-, and three-lobe sinusoidal acceleration pulses are investigated, and some of their basic properties are derived. Furthermore, the empirical mode decomposition （EMD） method is utilized as an adaptive filter to decompose the near-fault pulse-like ground motions, which were recorded during the September 20, 1999, Chi-Chi earthquake. These ground motions contain distinct velocity pulses, and were decomposed into high-frequency （HF） and low-frequency （LF） components, from which the corresponding HF acceleration pulse （if existing） and LF acceleration pulse could be easily identified and detected. Finally, the identified acceleration pulses are modeled by simplified sinusoidal approximations, whose dynamic behaviors are compared to those of the original acceleration pulses as well as to those of the original HF and LF acceleration components in the context of elastic response spectra. It was demonstrated that it is just the acceleration pulses contained in the near-fault pulse-like ground motion that fundamentally dominate the special impulsive dynamic behaviors of such motion in an engineering sense. The motion thus has a greater potential to cause severe damage than the far-field ground motions, i.e. they impose high base shear demands on engineering structures as well as placing very high deformation demands on long-period structures.

Engineering characteristics of near-fault vertical ground motions and their effect on the seismic response of bridges

A wide variety of near-fault strong ground motion records were collected from various tectonic environments worldwide and were used to study the peak value ratio and response spectrum ratio of the vertical to horizontal component of ground motion, focusing on the effect of earthquake magnitude, site conditions, pulse duration, and statistical component. The results show that both the peak value ratio and response spectrum ratio are larger than the 2/3 value prescribed in existing seismic codes, and the relationship between the vertical and horizontal ground motions is comparatively intricate. In addition, the effect of the near-fault ground motions on bridge performance is analyzed, considering both the material nonlinear characteristics and the P-A effect.

Near-fault ground motions with prominent acceleration pulses:pulse characteristics and ductility demand

Major earthquakes of last 15 years （e.g., Northridge 1994, Kobe 1995 and Chi-Chi 1999） have shown that many near-fault ground motions possess prominent acceleration pulses. Some of the prominent ground acceleration pulses are related to large ground velocity pulses, others are caused by mechanisms that are totally different from those causing the velocity pulses or fling steps. Various efforts to model acceleration pulses have been reported in the literature. In this paper, research results from a recent study of acceleration pulse prominent ground motions and an analysis of structural damage induced by acceleration pulses are summarized. The main results of the study include： （1） temporal characteristics of acceleration pulses; （2） ductility demand spectrum of simple acceleration pulses with respect to equivalent classes of dynamic systems and pulse characteristic parameters; and （3） estimation of fundamental period change under the excitation of strong acceleration pulses. By using the acceleration pulse induced linear acceleration spectrum and the ductility demand spectrum, a simple procedure has been developed to estimate the ductility demand and the fundamental period change of a reinforced concrete （RC） structure under the impact of a strong acceleration pulse.

Fractal characterization and frequency properties of near-fault ground motions

This study explores the irregularity and complexity of strong earthquake ground motions from the perspective of fractal geometry, and constructs a relation with the frequency content of the ground motions. The box-counting fractal dimensions and five representative period parameters of near-fault ground motions from the Chi-Chi and Northridge earthquakes are calculated and compared. Numerical results indicate that the acceleration and velocity time histories of ground motions present the statistical fractal property, and the dominant pulses of near-fault ground motions have a significant influence on their box dimensions and periods. Further, the average box dimension of near-fault impulsive ground motions is smaller, and their irregular degree of wave forms is lower. Moreover, the box dimensions of ground motions reflect their frequency properties to a large extent, and can be regarded as an alternative indicator to represent their frequency content. Finally, the box dimension D of the acceleration histories shows a considerably negative correlation with the mean period T. Meanwhile, the box dimension of the velocity histories Dye is negatively correlated with the characteristic period T and improved characteristic period Tgi.

Strength reduction factors for seismic analyses of buildings exposed to near-fault ground motions

To estimate the near-fault inelastic response spectra, the accuracy of six existing strength reduction factors （R） proposed by different investigators were evaluated by using a suite of near-fault earthquake records with directivity-induced pulses. In the evaluation, the force-deformation relationship is modelled by elastic-perfectly plastic, bilinear and stiffness degrading models, and two site conditions, rock and soil, are considered. The R-value ratio （ratio of the R value obtained from the existing R-expressions （or the R-p-T relationships） to that from inelastic analyses） is used as a measurement parameter. Results show that the R-expressions proposed by Ordaz ＆ Perez-Rocha are the most suitable for near-fault ground motions, followed by the Newmark ＆ Hall and the Berrill et al. relationships. Based on an analysis using the near-fault ground motion dataset, new expressions of R that consider the effects of site conditions are presented and verified.

Dynamic response of concrete face rockfill dam affected by polarity reversal of near-fault earthquake

In China,an increasing number of high concrete face rockfill dams(CFRDs)are located in high intensity earthquake zones,some of which are close to the seismic fault line.Recordings suggest that near-fault ground motions are characterized by large one-sided velocity pulses.The conventional dynamic analysis of dams,however,neglects the features of strong ground movements.In this study,under different ground motion levels some numerical dynamic studies considering the one-sided broadband pulses of near-fault earthquakes are presented for CFRDs based on a generalized plasticity model for rockfill materials.The results indicate that the displacements of dam crest corresponding to positive and reverse input of near-fault ground motion make a significant difference,while the displacements of the dam crest under artificial seismic waves are similar.Furthermore,using the horizontal and vertical components as simultaneous excitations near the faults,the displacements of the dam crest before and after reversing the motion produce a larger difference than that using a single component.More importantly,the difference of horizontal displacements of the dam crest caused by polarity reversal of near-fault ground motions increases with the increase of earthquake intensity.Due to the randomness and uncertainties of earthquakes,using a stochastic near-field motion input as excitation without considering the polarity(i.e.,positive vs reversed waveform),does not necessarily obtain a conservative result.

Study on inelastic displacement ratio spectra for near-fault pulse-type ground motions

In displacement-based seismic design, inelastic displacement ratio spectra （IDRS） are particularly useful for estimating the maximum lateral inelastic displacement demand of a nonlinear SDOF system from the maximum elastic displacement demand of its counterpart linear elastic SDOF system. In this study, the characteristics of IDRS for near-fault pulse-type ground motions are investigated based on a great number of earthquake ground motions. The influence of site conditions, ratio of peak ground velocity （PGV） to peak ground acceleration （PGA）, the PGV, and the maximum incremental velocity （MIV） on IDRS are also evaluated. The results indicate that the effect of near-fault ground motions on IDRS are significant only at periods between 0.2 s - 1.5 s, where the amplification can approach 20%. The PGV/PGA ratio has the most significant influence on IDRS among the parameters considered. It is also found that site conditions only slightly affect the IDRS.

Spatial distribution of near-fault ground motion

Near-fault strong ground motion of strike-slip and dip-slip of vertical and inclined rectangular fault in half-space and layered half-space is analyzed by dislocation source model. The Fourier spectra ratio of ground motion is adopted to study the characteristics of near-fault ground motion. For both slip models, near-fault strong ground motion with high amplitude is located in a narrow belt area along the projection of the fault on the ground and mainly controlled by the sub-faults nearby. Directivity of strike-slip fault is more dominant in long period for components perpendicular to the fault, and more dominant in long period for components parallel to the fault for dip-slip fault. The deeper the location of the source is, the more slowly the amplitude of ground motion attenuates. There is obvious hanging wall effect in ground motion of inclined fault, and the spatial distribution of ground motion is asymmetric which coincides with observational data. Finally, a fitting function of spatial distribution for near-fault ground motion is proposed and compared with near source factors of the 1997 Uniform Building Code of USA.

Dynamic-Response Analysis of the Branch System of a Utility Tunnel Subjected to Near-Fault and Far-Field Ground Motions in Different Input Mechanisms

There are few studies on the dynamic-response mechanism of near-fault and far-field ground motions for large underground structures,especially for the branch joint of a utility tunnel(UT)and its internal pipeline.Based on the theory of a 3D viscous-spring artificial boundary,this paper deduced the equivalent nodal force when a P wave and an SV wave were vertically incident at the same time and transformed the ground motion into an equivalent nodal force using a self-developed MATLAB program,which was applied to an ABAQUS finite element model.Based on near-fault and far-field groundmotions obtained fromtheNGA-WEST2 database,the dynamic responses of a utility tunnel and its internal pipeline in different inputmechanisms of near-fault and far-field groundmotions were compared according to bidirectional input and tridirectional input,respectively.Generally,the damage to the utility tunnel caused by the near-fault ground motion was stronger than that caused by the far-field ground motion,and the vertical ground motion of near-fault ground motion aggravated the damage to the utility tunnel.In addition,the joint dislocation of the upper and lower three-way joints of the pipeline in the branch systemunder the seismic action led to local stress concentrations.In general,the branch system of the utility tunnel had good seismic performance to resist the designed earthquake action and protect the internal pipeline fromdamage during the rare earthquake.