A 3D velocity model and 3D wave propagation code have been employed to simulate long-period ground motions in the upper Mississippi embayment. This region is exposed to seismic hazard in the form of large earthquakes in the New Madrid seismic zone and observational data are sparse, making simulation a valuable tool for predicting the effects of large events. These simulations were undertaken in order to estimate ground-motion characteristics and to investigate the influence of the 3D embayment structure and finite-fault mechanics. There are three primary fault zones in the New Madrid seismic zone, each of which was likely associated with one of the three main shocks of the 1811-1812 earthquake sequence. For this study, three simulations have been conducted on each major segment, evaluating the effects of different epicentral locations and rupture directions on ground motions. The full wave field up to a frequency of 0.5 Hz was computed on a 200 × 200 × 50 km3 volume, and up to a frequency of 1.0 Hz on a 100 × 100 × 50 km3 volume, using a staggered-grid finitedifference code. Peak horizontal velocity, bracketed durations, and pseudospectral accelerations were calculated at the free surface. Animations showing the evolution of peak horizontal velocity through time at the free surface were also generated. The New Madrid seismic zone simulations indicate that for the considered bandwidth, finite-fault mechanics such as fault proximity, directivity effect, and slip distribution exert the most control on ground motions. The 3D geologic structure of the upper Mississippi embayment also influences ground motion, with indications that the bedrock surface acts as a wave guide, trapping waves in shallow, low-velocity parts of the embayment.

Numerical modeling techniques commonly used to compute the response of soil and rock media under earthquake shaking are evaluated by analyzing the observations provided by instrumented borehole arrays. The NIED Kik-Net database in Japan is selected as the main source of borehole array data for this study. The stiffness of the site and the availability of high intensity motions are the primary factors considered towards the selection of appropriate Kik-Net borehole arrays for investigation. Overall, 13 instrumented vertical arrays are investigated using over 750 recorded ground motions characterized by low (less than 0.05 g) to high (greater than 0.3 g) recorded peak ground accelerations at the downhole sensor. Based on data from the selected borehole arrays, site response predictions using 1-D linear elastic (LE) analysis, equivalent linear (EQL) analysis, equivalent linear analysis with frequency-dependent soil properties (EQL-FD), and fully nonlinear analysis (NL) are compared with the borehole observations. Initially, the low intensity motions are used to evaluate common assumptions regarding 1-D site response analysis. First, we identify the borehole wavefield best simulating the actual boundary condition at depth by comparing the theoretical linear-elastic (LE) and observed responses. Then, we identify the best-fit small-strain damping profiles that can incorporate the additional in-situ attenuation mechanisms. Finally, we assess the validity of the one-dimensional modeling assumption. Our analyses indicate that the appropriate boundary condition for analysis of a borehole array depends on the depth of the borehole sensor and that, for most of the considered vertical arrays, the one-dimensional assumption reasonably simulates the actual wave propagation pattern. In the second part of this study, we evaluate the accuracy of the EQL, EQL-FD and NL site response methods by quantifying the misfit (i.e., residual) between the simulations and observations at different levels of shaking. The evaluation of the performance of the theoretical models is made both on a site-by-site basis and in an aggregated manner. Thereafter, the variability in the predicted response from the three site response methods is assessed. Comparisons with the observed responses indicate that the misfit of simulations can be significant at short periods and large strains. Moreover, all models seem to be characterized by the same level of variability irrespectively of the level of shaking. Finally, several procedures that can be used to improve the accuracy of the one-dimensional EQL, EQL-FD and NL site response analyses, are investigated. First, an attempt to take into account the shear strength of the soil materials at large shear strains is made. Additionally, several modifications to the EQL-FD approach are proposed. The proposed modifications are evaluated against recordings from the borehole arrays. Our analyses indicate that the accuracy of the theoretical models can be, partly, increased by incorporating the proposed modifications. / text

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第21461号 / 工博第4536号 / 新制||工||1707(附属図書館) / 京都大学大学院工学研究科建築学専攻 / (主査)教授 松島 信一, 教授 竹脇 出, 教授 池田 芳樹 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

The design earthquake selection problem is fundamentally probabilistic. Disaggregation of a probabilistic model of the seismic hazard offers a rational and objective approach that can identify the most likely earthquake scenario(s) contributing to hazard. An ensemble of time series can be selected on the basis of the modal earthquakes derived from the disaggregation. This gives a useful time-domain realization of the seismic hazard, to the extent that a single motion parameter captures the important time-domain characteristics. A possible limitation to this approach arises because most currently available motion prediction models for peak ground motion or oscillator response are essentially independent of duration, and modal events derived using the peak motions for the analysis may not represent the optimal characterization of the hazard. The elastic input energy spectrum is an alternative to the elastic response spectrum for these types of analyses. The input energy combines the elements of amplitude and duration into a single parameter description of the ground motion that can be readily incorporated into standard probabilistic seismic hazard analysis methodology. This use of the elastic input energy spectrum is examined. Regression analysis is performed using strong motion data from Western North America and consistent data processing procedures for both the absolute input energy equivalent velocity, (Vea), and the elastic pseudo-relative velocity response (PSV) in the frequency range 0.5 to 10 Hz. The results show that the two parameters can be successfully fit with identical functional forms. The dependence of Vea and PSV upon (NEHRP) site classification is virtually identical. The variance of Vea is uniformly less than that of PSV, indicating that Vea can be predicted with slightly less uncertainty as a function of magnitude, distance and site classification. The effects of site class are important at frequencies less than a few Hertz. The regression modeling does not resolve significant effects due to site class at frequencies greater than approximately 5 Hz. Disaggregation of general seismic hazard models using Vea indicates that the modal magnitudes for the higher frequency oscillators tend to be larger, and vary less with oscillator frequency, than those derived using PSV. Insofar as the elastic input energy may be a better parameter for quantifying the damage potential of ground motion, its use in probabilistic seismic hazard analysis could provide an improved means for selecting earthquake scenarios and establishing design earthquakes for many types of engineering analyses. / Ph. D.

Les mouvements forts du sol générés par les séismes peuvent être reliés aux caractéristiques de la source sismique (chute de contrainte, vitesse de rupture, etc.), et aux conditions frictionnelles des failles. Ces caractéristiques sont généralement étudiées via l’analyse des enregistrements accélérométriques des grands séismes à basses fréquences (≤ 1 Hz). L'amélioration des réseaux accélérométriques a permis l’enregistrement de nombreux séismes de magnitudes faibles et modérées (Mw ≤ 6,0). Ces enregistrements contiennent des informations dans la bande de hautes fréquences (1 Hz - 50 Hz), qui ne peuvent pas être exploitées avec les méthodes sismologiques classiques.Pour exploiter ces données de mouvements forts des séismes de faible intensité dans l’objectif d’étudier étudier les conditions frictionnelles de l'interface de subduction, nous explorons deux méthodes pour comparer le contenu fréquentiel des séismes: la première basée sur les rapports spectraux et la deuxième basée sur les équations de prédiction du mouvement du terrain (GMPE). Ces méthodes ont été utilisées pour étudier la variabilité spatiale et temporelle du contenu fréquentiel des séismes rompant l'interface de subduction au nord du Chili et au Japon. Leurs avantages et limitations respectifs ont été analysés. Ces comparaisons méthodologiques nous ont a permis de confronter et de valider les résultats, et ainsi de proposer une nouvelle méthodologie fiable basée sur l'analyse des résidus de GMPE pour analyser le contenu fréquentiel des séismes.L'analyse des résultats obtenus a montré une dépendance avec la profondeur du contenu en fréquence des séismes de l’interface de subduction, concordant avec les observations de grandes ruptures de subduction [Lay et al., 2012]. En même temps, des variations du contenu fréquentiel des séismes ont été détectées le long de la fosse de subduction, ce qui nous a conduit à décrire une segmentation latérale de l'interface de subduction. Cette segmentation a été comparée avec la distribution spatiale de la sismicité, à la géométrie de l'interface de subduction et à ses conditions de glissement. Finalement, avant le séisme d’Iquique de 2014 (Mw 8.1) au nord du Chili, nous avons détecté une évolution temporelle du contenu en fréquence des séismes associée à l’occurrence d'un glissement lent précurseur au choc principal. / The strong ground motions generated by earthquakes can be related to the characteristics of the earthquakes source (stress drop, rupture velocity etc..), and therefore to the frictional conditions of the faults. These characteristics are usually studied by analyzing the low frequency band (≤ 1Hz) of the strong motion records of large earthquakes. The improvement of strong motion networks has generated large datasets of records of moderate and low magnitude earthquakes (Mw ≤ 6.0). These records contain informations in the high frequency band (1 Hz – 50 Hz), which cannot be exploited using classical seismological methods.In order exploit the strong motion records of low magnitude earthquakes to study the subduction interface’s frictional conditions, we explore two methods for comparing the earthquakes frequency content: the first one based on spectral ratios, and the second one based on the Ground Motion Prediction Equations (GMPEs). These methods have been used to investigate the spatial and temporal variability of the frequency content of subduction interface earthquakes in North Chile and Japan. Their respective benefits and limitations have been analyzed. These methodological comparisons allowed us to cross compare and validate the results, and to propose a new, reliable methodology based on the analysis of GMPEs residuals to compare the earthquakes frequency content.The analysis of the results showed a depth dependency of the frequency content of subduction earthquakes in agreement with the one derived from large megathrust ruptures [Lay et al., 2012]. Additionally, variations of the earthquake frequency content along trench have been detected, which may drive to a lateral segmentation of the subduction interface. This segmentation has been compared to the spatial distribution of the seismicity, the geometry of the subduction interface and its slippage conditions. Finally, before the occurrence of 2014 Iquique Earthquake Mw 8.1 in North Chile, a temporal evolution of the frequency content of the foreshocks has been detected, associated to a precursory slow slip of the subduction interface.

This study examines the sensitivity of calculated site response in connection with alternative assumptions regarding input motions and procedures prescribed in the IBC 2000 building code, particularly the use of average shear wave velocity in the upper 30 meters as an index for engineering design response spectra. Site specific subsurface models are developed for four sites in and near Columbia, South Carolina using shear wave velocity measurements from cone penetrometer tests. The four sites are underlain by thin coastal plain sedimentary deposits, overlying high velocity Paleozoic crystalline rock. An equivalent-linear algorithm is used to estimate site response for vertically incident shear waves in a horizontally layered Earth model. Non-linear mechanical behavior of the soils is analyzed using previously published strain-dependent shear modulus and damping degradation models. Two models for material beneath the investigated near-surface deposits are used: B-C outcrop conditions and hard rock outcrop conditions. The rock outcrop model is considered a geologically realistic model where a velocity gradient, representing a transition zone of partially weathered rock and fractured rock, overlies a rock half-space. Synthetic earthquake input motions are generated using the deaggregations from the 2002 National Seismic Hazard Maps, representing the characteristic Charleston source. The U. S. Geological Survey (2002) uniform hazard spectra are used to develop 2% in 50 year probability of exceedance input ground motions for both B-C boundary and hard rock outcrop conditions. An initial analysis was made for all sites using an 8 meter thick velocity gradient for the rock input model. Sensitivity of the models to uncertainty of the weathered zone thickness was assessed by randomizing the thickness of the velocity gradient. The effect of the velocity gradient representing the weathered rock zone increases site response at high frequencies. Both models (B-C outcrop conditions and rock outcrop conditions) are compared with the International Building Code (IBC 2000) maximum credible earthquake spectra. The results for both models exceed the IBC 2000 spectra at some frequencies, between 3 and 10 Hz at all four sites. However, site 2, which classifies as a C site and is therefore assumed to be the most competent of the four sites according to IBC 2000 design procedures, has the highest calculated spectral acceleration of the four sites analyzed. Site 2 has the highest response because a low velocity zone exists at the bottom of the geotechnical profile in immediate contact with the higher velocity rock material, producing a very large impedance contrast. An important shortcoming of the IBC 2000 building code results from the fact that it does not account for cases in which there is a strong rock-soil velocity contrast at depth less than 30 meters. It is suggested that other site-specific parameters, specifically, depth to bedrock and near-surface impedance ratio, should be included in the IBC design procedures. / Master of Science

In this thesis it is aimed to compile the Turkish strong-motion database for its efficient use in earthquake engineering and strong-motion seismology related studies. Within this context, the Turkish strong-motion database is homogenized in terms of basic earthquake source parameters (e.g. magnitude, style-of-faulting) as well as site classes and different source-to-site distance metrics. As part of this objective, empirical relationships for different magnitude scales are presented for further harmonization of the database. Data processing of the selected raw (unprocessed) strong-motion accelerograms that do not suffer from non-standard problems are realized. A comparative study is also conducted between the peak ground-motion values of Turkish strong-motion database with the estimations computed from different ground-motion prediction models. This way the regional differences of Turkish database are evaluated by making use of global prediction models. It is believed that the main products of this thesis will be of great use for reliable national seismic risk and hazard studies.

High-pass filtering that is one of the most efficient methods in removing long-period noise of accelerograms is investigated for its effect on nonlinear oscillator deformation response. Within this context, uncertainty in filter cut-off periods that would significantly modify the low-frequency content of accelerograms come into prominence for obtaining reliable long-period displacement response. Analog and digital ground-motion records from recently compiled Turkish strong-motion database are used and these records are high-pass filtered with a consistent methodology by randomly generated filter cut-offs that represent different filter cut-off decisions of the analysts. The uncertainty in inelastic spectral and residual displacements (SDIE and SDR, respectively) due to variations in filter cut-offs is examined to derive the usable period ranges where the effect of high-pass filtering is tolerable. Non-degrading, stiffness degrading and stiffness and strength degrading oscillator behavior are considered in these analyses. The level of nonlinear behavior in single degree of freedom (SDOF) response is described by varying the yield strength (R, normalized yield strength) and displacement ductility (&micro / ) levels. The usable period ranges that depend on magnitude, recording quality, level of inelasticity and level of degradation are determined for SDIE through robust probabilistic methodologies.

Lors de séismes forts, la propagation des ondes sismiques dans les sols met en jeu des non linéarités de comportement qui se manifestent différemment selon le niveau de sollicitation. En effet, pour de faibles déformations (généralement <10^{-6}), une loi de comportement linéaire (i.e. module et amortissement indépendants du niveau de sollicitation) permet de reproduire les observations expérimentales sur site. En revanche, pour des déformations plus élevées, une loi de comportement non linéaire hystérétique est nécessaire pour décrire l'évolution de la rigidité et des dissipations énergétiques au cours de la sollicitation sismique. De plus, comme les séismes forts sont caractérisés par des amplitudes et des durées plus importantes, le rôle de la pression interstitielle ne peut pas être négligé pour les sols saturés sous fortes sollicitations (mobilité cyclique et liquéfaction). Ces phénomènes peuvent conduire à l'annulation des contraintes effectives et devenir cause de dommages sévères pour les structures et les ouvrages. L'analyse proposée élargit l'applicabilité des modèles de calcul actuels pour une analyse plus fine du risque sismique. En partant d'une formulation aux éléments finis décrivant la propagation des ondes sismiques suivant la direction verticale en prenant en compte le chargement en 3D (l'approche « 1D-3C» : une direction-trois composantes) dans les sols nonlinéaires secs, de nouvelles stratégies pour la prise en compte du rôle de l'eau sont développées. Le modèle de comportement est basé sur la relation entre la pression interstitielle et le travail de la contrainte de cisaillement. Ce modèle décrivant l'évolution de la pression interstitielle considère l'état de contrainte tridimensionnel du matériau. Le modèle est validé par comparaison avec des résultats expérimentaux. L'approche « 1D-3C » a été utilisée pour modéliser la réponse des sols pour 4 séismes réels : le séisme de Superstition Hills en 1987 aux États-Unis (M_{w}=6.7); le séisme de Tohoku en 2011 au Japon (M_{w}=9.1 ); le séisme de Kushiro en 1993 au Japon (M_{w}=7.8) et le séisme d'Emilie Romagne en 2012 en Italie (M_{w}=5.9). Pour les trois premiers séismes, des enregistrements en profondeur et en surface sont disponibles. L'étude de ces trois premiers cas rend possible la validation du modèle par comparaison des données enregistrées et calculées. Le modèle peut donc être considéré comme un outil fiable pour la prédiction de la réponse sismique des sols saturés / During strong earthquakes, the seismic wave propagation in soils involves nonlinear behaviors strongly depending on the strain level. Indeed, for small strain (typically <10^{-6}), a linear constitutive law (modulus and damping independent on the load level) can reproduce the experimental observations on site. However, for larger strains, a nonlinear hysteretic constitutive law is needed to describe the evolution of stiffness and energy dissipation during seismic loading. In addition, as strong earthquakes are characterized by larger amplitudes and durations, the role of pore pressure cannot be neglected for saturated soils. Indeed pore water pressure controls phenomena such as cyclic mobility and liquefaction due to the loss of soil strength. This can lead to a fast decrease of effective stresses and permanent deformations in the soil causing severe damage to structures. This work extends the applicability of existing calculation models for a more detailed analysis of seismic risk. Starting from a FEM approach describing the propagation of seismic waves in the vertical direction, taking into account 3D loading (so-called "1D-3C" approach: 1 direction - 3 components) in nonlinear dry soils, new strategies to consider the role of water are developed. The model is based on the relationship between the pore pressure and the shear work. The three-dimensional stress state of the material is considered. The model is validated by comparison with experimental results. The "1D-3C" approach was used to model the response of soils for four real earthquakes: the Superstition Hills earthquake in 1987 in the United States (M_{w}=6.7), the Tohoku earthquake in 2011 in Japan (M_{w}=9.1), the Kushiro earthquake in Japan in 1993 (M_{w}=7.8) and the Emilia Romagna earthquake in Italy in 2012 (M_{w}=5.9). For the first three earthquakes, records at depth and on the surface are available. The study of the first three cases makes possible the validation of the model by comparing the calculated accelerations on the surface with the available records. The model can then be considered as an advanced tool for the prediction of the seismic soil response

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第18801号 / 理博第4059号 / 新制||理||1584(附属図書館) / 31752 / 京都大学大学院理学研究科地球惑星科学専攻 / (主査)教授 岩田 知孝, 教授 平原 和朗, 准教授 久家 慶子 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM