Search

Seismic fragility analysis is one probabilistic method that represents the conditional probabilities of exceeding different predetermined damage states for various given earthquake hazard levels, which is the essential part of probabilistic seismic risk assessment. It has gained much popularity recently because much more comprehensive and richer results than traditional probabilistic methods can be provided. The classical seismic fragility analysis of slopes is always performed through analytical approaches. In the present study, a novel seismic fragility analysis framework combined with shaking table model test results is proposed for probabilistic seismic performance and risk assessment of slopes. The shaking table model tests of a rock slope reinforced by pile-anchor structures were firstly performed subjected to a suite of real earthquake records with different intensity levels. Subsequently, the seismic responses from shaking table model tests were then gathered for the optimal analysis to determine the optimal engineering demand parameter (EDP) and earthquake intensity measure (IM) for seismic fragility analysis of slopes. At last, the identified most appropriate IM and EDP were selected to construct the scalar-valued fragility curves, vector-valued fragility surfaces and fragility surfaces considering acceleration elevation amplification effect by the proposed framework. The results indicate that using single IM to construct scalar-valued fragility functions may result in different failure probabilities depending on the chosen earthquake IM. The vector-valued fragility surfaces provide more information than ordinary fragility curves, and lead to more proper and realistic evaluations of seismic hazard of slopes. The methods and results presented in this paper also provide the basis for the probabilistic seismic risk and loss assessment of earthquake-induced slope geological disaster. • A novel seismic fragility analysis framework of slopes combined with the data from shaking table model tests is proposed. • The optimal EDP and earthquake IM for scalar- and vector-valued fragility analysis of slopes are identified. • The fragility surfaces of a slope considering acceleration elevation amplification effect are constructed. [ABSTRACT FROM AUTHOR]

Numerous interbedded rock (IR) slopes fail during the Wenchuan earthquake in the mountainous region of western China. Landslides are also triggered in IR slopes with a 60° layer inclination, which are generally stable in gravity-dominant environments. This study examines the effect of seismic motion on the response characteristics and failure patterns of IR slopes with steep layers to develop a landslide hazard assessment tool for earthquake-prone regions. First, we use a centrifuge shaking table test to model the failure process and acceleration responses of two IR slope models with stratigraphic dips of 60° and 80°, respectively, under different seismic intensities. Next, we adopt the Particle Flow Code to examine the crack propagation features and peak ground acceleration amplification effects for the IR slopes. We find that the seismic failure pattern of IR slopes depends largely on rock layer inclination: buckling failure is triggered when rock layers are parallel or nearly parallel to the slope surface, while toppling failure is triggered when the rock layer inclination is significantly higher than that of the slope surface. Following seismic excitation, the damage is mainly observed in the weak rock layers, creating lateral stress on adjacent strong rocks, which undergoes deformation and ultimate macroscopic failure. Further, displacement of the IR slope is negatively correlated to rock layer inclination. Rock layer thickness has a major influence on the damaged area inside the slope mass, while rock layer stiffness mainly affects the deformation distribution near the slope shoulder. • Interbedded slopes with steep layers under earthquake exhibit buckling or toppling. • Using centrifugal shaking table test to reproduce interbedded rock slope failures. • Non-uniform deformation is highly developed in interbedded slopes. • Rock layer inclination is a significant factor on failure type in interbedded slopes. [ABSTRACT FROM AUTHOR]

This paper presents the development of a shear-wave velocity (V S) model of reference site conditions in South Carolina for site response analysis and seismic hazard mapping. Model development involves creating contour maps of two geologic surfaces in the Coastal Plain with high contrasts in V S and combining the results with median V S profiles that extend to hard rock (V S ≥ 3000 m/s). Uncertainties associated with the contour maps of depth are captured in maps of the standard prediction error. Values of time-averaged V S in the top 30 m (V S,30) ranging from 1183 to 1538 m/s and 354 to 385 m/s characterize the reference site conditions in the Blue Ridge and Piedmont regions and the Coastal Plain region, respectively. The model is divided into 14 zones to facilitate site response analysis and to avoid over-smoothing resonances due to the impedance contrasts. This study is part of a larger effort to develop statewide probabilistic seismic hazard maps for the South Carolina Department of Transportation. • A geology-based, shear-wave velocity (V S) model of reference site conditions in South Carolina is developed. • Two depth contour maps are created: 1) Top of the Jurassic and older basement; and 2) Bottom of the Santee Limestone. • V S30 for reference outcrop site conditions are between 354 m/s and 385 m/s in Coastal Plain of South Carolina. • In the South Carolina Blue Ridge and Piedmont, V S30 for reference outcrop site conditions range from 1183 to 1518 m/s. • South Carolina is divided into 14 zones for the purposes of site response analysis. [ABSTRACT FROM AUTHOR]

The paper reports on the results of 1D and 2D site response analyses carried out in the Kathmandu Valley, Nepal, in order to investigate how site effects influenced the seismic response during the 2015 Mw 7.8 Gorkha earthquake sequence. The mainshock and a Mw 6.6 aftershock, for which recordings at both rock and soil sites are available, were considered. First, 1D analyses were carried out for the Pulchowk soft soil site, where a borehole was drilled. The shear wave velocity profile was defined using several 2D seismic array surveys carried out in the valley and constrained by noise measurements at Pulchowk site; the nonlinear soil behavior was characterized by means of cyclic simple shear tests carried out on undisturbed soil samples. Both equivalent and nonlinear approaches were adopted in the 1D analyses. Overall, the 1D model was capable to capture some relevant features shown by mainshock recordings such as the de-amplification of medium-to-high frequencies. On the contrary, the unusual high spectral amplification at long periods (3–6 s) recorded during the mainshock was better captured by the 2D finite element analyses carried out on a 20 km-large cross section of the entire valley, thus supporting the hypothesis of the occurrence of basin effects. The paper contributes to the understanding of site effects in Kathmandu Valley for the implementation of seismic risk mitigation strategies in the area. [ABSTRACT FROM AUTHOR]

The Newmark sliding block model is widely used in the evaluation of seismic slope stability. When using this model, the horizontal acceleration is usually considered to act parallel to the slope (downslope), and the vertical component of ground motions is commonly ignored. This study investigates the effect of vertical accelerations on the results of sliding displacement, and quantitatively compares the difference of the displacement D N calculated from the original Newmark's approach with the displacement D HV calculated from a more realistic case (horizontal and vertical accelerations properly considered). For this purpose, 4136 ground motion recordings are selected from the NGA-West2 database, and >2000 hypothetic slopes are simulated with critical accelerations a c ranging from 0 to 0.8 g. Comparison of D HV and D H shows that the incorporation of vertical accelerations only slightly increases the displacement. Besides, the calculated D HV / D N ratios are generally in the range of 1 to 5, indicating that D N significantly underestimates the slope displacement due to the simplification of seismic excitations. The D HV / D N ratios generally increase as a c increases. Simple predictive models are proposed to quantify the correction factor K D ( K D  =  D HV / D N ). They can be easily used in seismic landslide hazard assessment to improve the performance of the original Newmark's approach. [ABSTRACT FROM AUTHOR]

Earthquake-induced soil liquefaction can cause enormous life and economic losses. The joint distribution of the peak ground surface acceleration ( a max ) and the moment magnitude of an earthquake ( M w ) is crucial to the estimation of soil liquefaction-related hazards. In this paper, a new method is suggested to estimate the joint distribution of a max and M w . Compared with the existing method, the suggested method is easier to implement, and depends on fewer assumptions. The derived joint distribution is then used to assess the damage state of a building caused by soil liquefaction during a given exposure time. The method is illustrated with an example in which a reinforced concrete frame building is assumed to be at different locations and is subjected to different exposure time. It is found that, as the exposure time increases, the chance of higher degree of damage also increases. When the exposure time is the same, the damage state of the building may also be different when it is at different locations. The method suggested in this paper can be used to quantify the effect of ground condition as well as the seismic effects at a site on the damage state of a building, thus providing a more transparent understanding of the risk associated with the soil liquefaction hazard on structural damage. [ABSTRACT FROM AUTHOR]

In some geological settings, such as thick sedimentary basins, the deeper layers may also significantly contribute to the amplification of the ground motion during an earthquake, particularly at low frequencies. To properly quantify the influence of the deeper geology on the seismic ground motion at the surface, the use of indirect geophysical exploration techniques is particularly suitable, as it can produce realistic representations of the subsoil structure over a large range of depths. This paper focuses on the acquisition and processing of combined shear- (S) and compressional-wave (P) seismic reflection data to obtain an accurate geological/geophysical model with reliable seismic velocities for local-scale seismic site response studies, as needed for hazard assessment and calculation of earthquake scenarios. As an application case, we focus on a microzonation study performed for the Cavezzo municipality in Northern Italy. The investigated site is located in the epicentral area of the 2012 Emilia earthquake in the Po Valley, which is a typical deep sediment-filled basin. Specifically, in this study we were able to: (1) confirm the advantages of reflection seismic prospecting for soft-sediment basin characterization; (2) evaluate the resolution of S- and P-wave signals and their depth of penetration; (3) improve the knowledge on the geological structure of the Cavezzo area by measuring S- and P-wave interval velocities in Plio-Quaternary formations. • S- and P-wave signal resolution and penetration capabilities have been assessed. • Seismic reflection data were processed to directly evaluate the local seismic response. • An improved S- and P-wave interval velocities model for the Plio-Quaternary formations of the Po Plain are presented. [ABSTRACT FROM AUTHOR]

In this paper we present a new methodological approach which integrates geological and geophysical data into a 3D modelling process to be mainly employed in seismic hazard assessment studies of earthquake-prone areas around the world, as well as in applications for land use and urban planning. As a case study, the reconstruction of a geology-based 3D velocity model of the uppermost hundreds of metres of the Amatrice high-seismic-hazard area is described. The model was constructed using geological (e.g., maps, cross-sections and core-wells) and geophysical (e.g., down-hole, MASW, refraction, and seismic noise measurements) data, which were georeferenced and uploaded into 3D geological modelling software, where faults, stratigraphic boundaries, and geophysical attributes were digitised, checked, hierarchised, and modelled. The performed 3D geological model was parameterised with V s and V p velocities and, finally, the environmental noise (i.e., horizontal-to-vertical spectral ratio analysis, HVSR) recorded at some seismic stations was compared with the seismic responses modelled at some nearby control points. In the study area, the proposed geology-based 3D velocity model represents both a new potential geophysical prediction tool for areas devoid of geophysical measurements (i.e. HVSR curves) and a potential input-model for future ground-motion and seismic-wave-propagation simulations aimed at a more precise local seismic response assessment and, consequently, at the development of more realistic seismic hazard scenarios. The model here presented constitutes a first version of the 3D geological-geophysical model for the studied area, which will be improved with new data and more advanced algorithms available in the future. [Display omitted] • A new methodological approach for the 3D geophysical model reconstructing is proposed. • The first 3D geology-based velocity model of the Amatrice area was reconstructed. • The model integrates and completes the earlier studies on the Amatrice Basin. • The performed model represents a new geophysical prediction tool for the Amatrice area. [ABSTRACT FROM AUTHOR]

DInSAR data provide a powerful tool to recognize the Earth's surface where permanent deformations are concentrated and undergo the strongest ground motion during an earthquake, i.e., defining the epicentral area. We analyzed three recent seismic events in the Apennines belt in Italy related to extensional and contractional earthquakes and documented that the largest macroseismic intensity is recorded where the ground underwent the largest vertical component of motion and the stronger vertical component of the peak ground acceleration. Besides site amplification effects that may occur everywhere even outside the epicentral area, we infer that the vertical oscillations in the so-called near-field allow the horizontal shaking to be more effective, hence producing larger damage above the active crustal volume. The active volume is the one moving vertically and is contemporaneously crossed by seismic waves nucleated by the shear on the fault. The surrounding passive volume in the far field is only crossed by seismic waves. The near field areas are elliptical and cover areas of 300–600 km2 for a range of M w 6–6.9, and they should be the areas where the highest seismic hazard must be expected. Therefore, their area is too large to be neglected for seismic hazard assessment. • Coseismic vertical deformation is recorded for three Apenninic seismic sequences. • Vertical deformation coincides with the largest shaking recorded (PGA). • Maximum vertical DInSar data delimit the highest macroseismic and intensity measures. • The largest damage is recorded in the so-called active crustal volume. • The active volume is where vertical displacement occurs while crossed by seismic waves. [ABSTRACT FROM AUTHOR]

Prediction of surface ground motion based on advanced approaches is a non-trivial task at large area. In fact, advanced approaches require a detailed geological and geotechnical subsoil model based on geological, physical and mechanical properties retrieved from in situ investigations and laboratory data, and to perform time consuming numerical simulation of the seismic site response. Bearing in mind that there is a lack of available site investigations and laboratory data at large area, simplified methods based on "soft data" are often adopted. On this regard, a methodology to perform seismic site response analyses at national scale is proposed in this paper, based on the following assumptions: 1) homogeneous areas were identified at national scale based both on morpho-geological and lithological properties, in order to classify and analyse site-data archived in the Italian database of seismic microzonation studies; 2) geotechnical subsoil one-dimensional models of the Italian site conditions were defined according to site-data; 3) in accordance with the reference seismic hazard levels, pseudo-acceleration response spectra were selected as input motions for numerical simulations. About thirty million numerical simulations of seismic site response were performed, and national maps of seismic shaking amplification factors were produced in terms of ratio of response spectra referred to three intervals of period. This estimation of surface ground motion based on the proposed procedure is a supporting tool for large area territorial and emergency planning. • Numerical models of national site conditions were identified from Italian database. • Numerical models were extended down to the seismic bedrock. • Morpho-geological clusters were used to spatialize input and output data. • National maps of amplified seismic intensity were proposed. • The proposed procedure will support seismic risk mitigation at large area. [ABSTRACT FROM AUTHOR]

The 1999 Chi-Chi, Taiwan, earthquake (Mw 7.6) caused damage to 49 tunnels within 60 km of the epicenter, including the collapse of three mountain tunnels, ending the traditional perception that tunnels, and especially mountain tunnels, were earthquake-resistant. A series of studies that involved seismic damage investigation, physical model and numerical simulation experiments, numerical analyses and field monitoring have been carried out to investigate the damage mechanism, influencing factors and the seismic response of tunnels, and these are reviewed in this manuscript. The results of such studies over the last two decades demonstrate that tunnels suffer less seismic damage than surface structures. Most of seismic damage to tunnels involves failure of the adjacent ground or the movement of a fault that is crossed by the tunnel. Tunnel damage that is caused by ground motion occurs where the seismic effects are amplified, such as at the slope next to mined sections and portals, in shallow overburden tunnels, and at particular combinations of tunnel depth and geological conditions. Results of physical model-based tests using shaking tables and centrifuges reveal that tunnels exhibit rocking responses along with ovaling deformation or racking distortions under seismic excitation; also residual earth pressures remain on the tunnel side walls and internal forces of lining exist after shaking. The structural responses of a tunnel under non-uniform excitation are larger than that under uniform excitation. Beyond investigations of the damage mechanism and influencing factors, physical model-based tests provide experimental benchmarks for correcting numerical model settings and calibrating the characteristic parameters in numerical simulations, and validate the application of such models and simulations in the investigation of tunnel seismic responses. A fully dynamic analysis with complete descriptions of the topography and geological characteristics of a site, and the engineering characteristics of the tunnel and surrounding ground, is nowadays probably the most effective way for the seismic simulation of tunnels, and the design and evaluation of the same. An understanding of the state-of-the-art with respect to the seismic response of tunnels demonstrates the importance of three-dimensional geological models for tunnel seismic analysis; the evaluation of regional ground motion models that can capture the expected range of possible ground motions at tunnel site, and determinations of both the static and dynamic characteristics of geological conditions that are associated with site-scale vulnerabilities. Since residual earth pressures and internal forces remain on tunnel after an earthquake, the seismic damage to tunnels and loads and strain of the surrounding ground may accumulate. Methods for evaluating potential seismic damage to existing tunnels must be further developed, and more field monitoring of the seismic responses of tunnels is required. • Seismic response of tunnels revealed in past two decades is extensively reviewed. • Ground motion-induced tunnel damage occurs where the seismic effects are amplified. • Seismic damage to tunnels and loads and strain of surrounding ground may accumulate. • 3D geological models and more field monitoring are important for fully dynamic analyses. [ABSTRACT FROM AUTHOR]

Many existing earth dams have been designed and built worldwide before the establishment of a seismic code, so that it is of relevant interest to evaluate their seismic performance and post-seismic operational conditions. This requires an accurate geotechnical characterisation of the dam and foundation soils, a proper definition of the seismic scenarios at the site of the dam, the use of simplified procedures for screening-level seismic analyses and advanced non-linear dynamic analyses to study the most critical seismic scenarios. This process has been used for the evaluation of the seismic performance of a zoned earth dam located in a high seismic hazard area of Southern Italy. In this paper the available data of historical seismicity at the site of the dam and the results of a probabilistic seismic hazard analysis are first discussed and input ground motions are selected using compatibility criteria with the energy and frequency content of the expected target motion in a range of vibration period relevant for the non-linear response of the dam. Seismic performance of the dam is then evaluated through procedures based on Newmark-type computations, in which permanent displacements are related to ground motion characteristics and to the seismic resistance of the dam, the latter evaluated detecting the earthquake-induced plastic mechanisms and the corresponding critical accelerations. Also, an original improvement of the well-known Makdisi & Seed procedure, was proposed to better capture the actual influence of non-linear soil behaviour in the evaluation of horizontal acceleration and permanent displacements of the crest of the dam. The analysis results pointed out the relevant role of the earthquake-induced shear strength reduction on the dam permanent displacements. • The seismic performance of an earth dam is evaluated through displacement analyses. • A finite difference model was calibrated using laboratory and field data. • Input motions were defined through historical and probabilistic hazard analysis. • Permanent displacements were estimated using well-known and an original procedure. • The role of shear strength reduction occurring in the dam materials is highlighted. [ABSTRACT FROM AUTHOR]

Slopes should be regarded as stochastic dynamic systems when investigating the uncertainty of ground motions. In engineering, this involves the propagation and spatial evolution of physical quantities (e.g., acceleration) that can be assessed with some probability. More specifically, stochastic stability and amplification effects are important issues under random seismic inputs. In studying the above, a large-scale shaking-table test is carried out to investigate the slope seismic response under random earthquake ground motions, where 144 ground-motion samples are generated using a time-evolution power spectrum model. In this test, acceleration evolves in a way that its PDF changes with time and space. In the time domain, the PDF variability first increases and then decreases with the progression of the earthquake. In the spatial domain, the PDF variability increases upslope, indicating that the mean and standard deviation of amplification factors increase with elevation. Moreover, the stability of the upper part of the slope stochastic dynamic system is poor. Meanwhile, the de-amplification effect of the slope toe is identified under random seismic inputs. Finally, we discuss different amplification factors and consider applications of the research results in slope engineering. • Stochastic seismic response of a slope investigated in large-scale shaking-table test. • Random dynamic analysis framework for slopes was proposed for trial analysis. • Probability propagation and evolution of acceleration were obtained. • Amplification and de-amplification effects were identified under random inputs. [ABSTRACT FROM AUTHOR]

The evaluation of seismic site amplification and its relationship with geological structures play a vital role in earthquake engineering and hazard assessment. This study presents an analysis of topographic and geologic site amplification effects observed in the Qiaozhuang region, Sichuan Province, China. Seismic recordings at several monitoring stations installed on two slopes located at a distance of about 1 km in the study area after the 2008 Wenchuan earthquake provide evidence of strong and variable amplifications. To assess the combined effect of topographic and geological controls in slope response to seismic motions, we built 2D and 3D dynamic numerical models using monitoring data as inputs. Four layers with different P-wave velocities, based on geophysical survey in the study area, are considered in the numerical models. Models that only consider topography give an amplification factor of <3, which is much lower than the amplification factor of 5–6 from the monitoring data, while those models that considered both the topographic variation and four layers agree well with the monitoring data. The 3D modeling results show that the subsurface amplification factor is less than that on the slope surface even at the same elevation. This is also consistent with the monitoring data and further indicates that the combined topographic and geological amplification effect on the slope surface is more significant than the sole geological amplification effect inside the slope. Our numerical simulation results suggest that it is important to consider the combined topographic and geological amplification effects in the hazard assessment of seismically induced slope failures. • This study presents long-term monitoring data of slope seismic response. • Homogeneous numerical models show that topography alone only produces low amplification effects. • The results from three-layer heterogenous 3D model indicat the importance of geological amplification effect. [ABSTRACT FROM AUTHOR]