Research performed on the subject of dynamic soil-structure interaction (SS) concerning railway bridges is presented in this thesis with the focus on simply supported railway bridges supported by shallow foundations in soil strata on bedrock. The research aims to obtain insight into the SSI of high-speed railway bridges and to provide recommendations on how to model the soil-bridge system from a design perspective. A three-dimensional (3D) simply supported soil-bridge model was first developed and the effects from model assumptions made on the soil-foundation system was evaluated in a 3D setting (paper I). The soil-foundation system was then refined and a model assumptions study was performed in order to evaluate the effects of model assumptions on impedance functions, including the influence of the permanent load acting on the soil-foundation system (paper II). Finally, a study of the assembled soil-bridge system was performed in an extensive parametric study including a set of 2D bridge models in combination with a set of shallow foundations in soil strata on bedrock (paper III). A supplementary section related to paper III was also added in this thesis, showing the effects of the substructure mass. The model assumptions made when creating the soil-foundation model and the soil-bridge model can be very important and must be made with care. The permanent load acting on the soil-foundation systems of shallow foundations may alter the impedance functions significantly. The substructure mass may alter the behavior of the soil-bridge system depending on its magnitude, and neglecting it gives inaccurate results. The 3D effects of SSI do not cause high vibrations due to modes other than the first bending mode, and assuming a 2D bridge model is generally acceptable. The effects of SSI on the soil-bridge systems with shallow soil strata are largely dependent on the ratio between the natural frequency of the bridge and the fundamental frequency of the soil. Depending on the value of this ratio, the effect of including SSI in bridge models may contribute to the bridge obtaining a negligible, conservative, or non-conservative response, as compared to the bridge with the assumption of non-flexible supports. / Forskning i syfte att utröna effekten av dynamisk jord–struktur-interaktion (SSI)på järnvägsbroar presenteras i denna avhandling med huvudfokus på fritt upplagdabroar med stöd av plattgrundlagda fundament i jordar på fast berggrund. Forsknin-gen syftar till att ge förståelse för interaktionen mellan jord och järnvägsbroar samtatt ge rekommendationer på hur systemet kan modelleras ur ett designperspektiv.En tredimensionell (3D) fritt upplagd jord–bromodell utvecklades först och effek-terna av modellantaganden gjorda på jord–grundläggningssystemet utvärderadesi en 3D miljö (artikel I). Jord–grundläggningssystemet förfinades och en studiegenomfördes för att utvärdera effekterna av modellantaganden på impedansfunk-tioner, inklusive påverkan av den permanenta belastningen som verkar på jord–grundläggningssystemet (artikel II). Slutligen utfördes en omfattande parametriskstudie av det sammansatta jord–brosystemet där en uppsättning tvådimensionella(2D) bromodeller kombinerades med en uppsättning jordar (artikel III). Ett kom-pletterande avsnitt relaterat till artikel III lades till i denna avhandling som visareffekterna av massan av underbyggnaden på jord–brosystemet.De modellantaganden som görs vid skapandet av jord–grundläggningsmodeller ochjord–bromodeller kan vara mycket viktiga och bör utföras med varsamhet. Den per-manenta belastningen som verkar på jord–grundläggningssystemet kan väsentligtförändra impedansfunktionerna. Massan av underbyggnaden kan vidare ändra re-sponsen i jord–brosystemet, beroende på dess storlek, och att försumma den kan gefelaktiga resultat. De 3D effekterna av SSI orsakar inte höga vibrationer på grundav andra moder än den första böjmoden, och att anta en 2D bromodell är såledesgenerellt sett motiverat.Effekterna av SSI på jord–brosystemet i grunda jordar beror till stor del av kvotenmellan brons naturliga frekvens och jordens fundamentala frekvens. Beroende påvärdet på denna kvot kan effekten av att inkludera SSI i bromodeller bidra till attbron får en försumbar, konservativ, eller icke-konservativ respons, i jämförelse medbron med antagandet om fasta upplag. / <p>QC 20200903</p>

L’interaction dynamique entre le sol et les structures (IDSS) a fait l’objet de nombreuses études sous l’hypothèse de l’élasticité linéaire, bien que les effets de l’IDSS puissent être différents entre un système élastique et un système inélastique. De fait, les méthodologies usuelles développées à partir des études élastiques peuvent ne pas être adaptées aux bâtiments conçus pour dissiper de l’énergie par de l’endommagement lors de séismes sévères. De plus, il est bien connu que la limite d’élasticité du sol est normalement atteinte même pour de séismes relativement faibles. En conséquence, si les effets inélastiques de l’IDSS sont négligés, les études d’endommagement sismique des bâtiments peuvent être très inexactes. L’objectif de ce travail est de développer une stratégie générale pour l’étude du problème de l’IDSS non-linéaire dans le contexte de l’analyse de la vulnérabilité sismique des bâtiments. Ainsi, des modèles d’éléments finis réalistes sont développées et appliquées à des problèmes d’IDSS non-linéaires. Les modèles couvrent une large gamme des conditions pour le sol et des typologies de bâtiments soumis à plusieurs bases de données sismiques. Une stratégie de modélisation a été développée et validée afin de réduire significativement le coût numérique. Pour cela, un modèle 2D équivalent a été développé, implanté dans GEFDyn et utilisé pour effectuer une importante étude paramétrique. De nombreux indicateurs de comportement non-linéaire de la structure et du sol ont été proposés pour synthétiser leur fonctionnement lors du chargement sismique. De surcroît, une stratégie d’évaluation de la vulnérabilité sismique basée sur l’information apportée par une base des données sismiques a été développée. De façon, générale, les résultats ont mis en évidence une réduction de la demande sismique lorsque les effets inélastiques de l’IDSS sont pris en compte. Cette réduction est liée fondamentalement à deux phénomènes : l’amortissement par radiation et l’amortissement hystérétique du sol. Ces deux effets ont lieu simultanément pendant le mouvement sismique. Il est alors très difficile d’isoler l’influence de ces deux phénomènes. En effet, le mouvement effectif transmis à la structure n’est pas le même que celui en champs libre du aux effets d’interaction, ainsi qu’à la modification locale du comportement du sol fortement lié aux poids du bâtiment. Une série de mesures de sévérité sismique et des mécanismes de dissipation d’énergie au niveau du sol et du bâtiment a été introduite dans le but d’analyser ces effets. Cependant, ces résultats sont en général très irréguliers et leur généralisation a été très difficile. Néanmoins, ces résultats mettent en évidence l’importance de la prise en compte des effets du comportement inélastique du sol. La plupart des cas étudiés ont montré un effet favorable de l’IDSS non-linéaire. Mais, en général, l’IDSS peut augmenter ou diminuer la demande sismique en fonction de la typologie de la structure, des caractéristiques du mouvement sismique et des propriétés du sol. Tout de même, il y a une justification économique pour étudier les effets du comportement non-linéaire du sol sur la réponse sismique. / The dynamic interaction of the soil with a superstructure (DSSI) has been the subject of numerous investigations assuming elasticity of both, superstructure and soil foundation behavior. Nevertheless, the effect of DSSI may differ between elastic and inelastic systems. Thus, the current interaction methodologies based on elastic response studies could not be directly applicable to structures expected to behave inelastically during severe earthquakes. Additionally, the soil is known to exhibit inelastic behavior even for relatively weak to moderate ground motions. Consequently, ignoring these characteristics in studying DSSI could lead to erroneous predictions of structural damage. The main purpose of this work is to develop a general strategy to address the full DSSI problem in the context of the seismic vulnerability analysis of structures. Thus, realistic Finite Elements models are constructed and applied in a practical way to deal with these issues. These models cover a large range of soil conditions and structural typologies under several earthquake databases. Some modelling strategies are introduced and validated in order to reduce the computational cost. Therefore, an equivalent 2D model is developed, implemented in GEFDyn and used in the large parametric study conducted. Several indicators for both structural and soil responses are developed in order to synthesize their behavior under seismic loading. Additionally, a vulnerability assessment strategy is presented in terms of measures of information provided by a ground motion selection. According to the investigation conducted in this work, there is in general a reduction of seismic demand or structural damage when non-linear DSSI phenomenon is included. This reduction can be associated fundamentally to two phenomena: radiative damping and hysteretic damping due to non-linear soil behavior. Both effects take place simultaneously during the dynamic load and it is extremely difficult to separate the contribution of each part in reducing seismic demand. Indeed, effective motion transmitted to the superstructure does not correspond to the free field motion because of the geometrical and inertial interactions as well as the local modification of soil behavior, specially due to the supplementary confinement imposed by the superstructure’s weight. A series of strong-motion severity measures, structural damage measures and energy dissipation indicators have been introduced and studied for this purpose. Nevertheless, results are erratic and consequently, generalization was extremely difficult. Despite these difficulties, the results illustrate the importance of accounting for the inelastic soil behavior. The major part of the studied cases show beneficial effects such as the decrease of the maximum seismic structural demand. However, the non-linear DSSI could increase or decrease the expected structural damage depending on the type of the structure, the input motion, and the dynamic soil properties. Furthermore, there is an economic justification to take into account the modification effects due to inelastic soil behavior.

The recent 2011 Tōhoku earthquake and tsunami served as a stark reminder of the destructive capabilities of such combined events. Civil Engineers are increasingly tasked with protecting coastal populations and infrastructure against more severe multi-hazard events. Whilst the protective measures must be robust, their deployment over long stretches of coastline necessitates an economical and environmentally friendly design. The dual row retaining wall concept, which features two parallel sheet pile walls with a sand infill between them and tie rods connecting the wall heads, is potentially an efficient and resilient system in the face of both earthquake and tsunami loading. Optimal use of the soil's strength and stiffness as part of the structural system is an elegant geotechnical solution which could also be applied to harbours or elevated roads. However, both the static equilibrium and dynamic response of these types of constructions are not well understood and raise many academic and practical challenges. A combination of centrifuge and numerical modelling was utilised to investigate the problem. Studying the mechanics of the walls in dry sand from the soil stresses to the system displacements revealed the complex nature of the soil structure interaction. Increased wall flexibility can allow more utilisation of the soil's plastic capacity without necessarily increasing the total displacements. Recognising the dynamically varying vertical effective stresses promotes a purer understanding of the earth pressures mobilised around the walls and may encourage a move away from historically used dynamic earth pressure coefficients. In a similar vein, the proposed modified Winkler method can form the basis of an efficient preliminary design tool for practice with a reduced disconnect between the wall movements and mobilised soil stresses. When founded in liquefiable soil and subjected to harmonic base motion, the dual row walls were resilient to catastrophic collapse and only accrued deformation in a ratcheting fashion. The experiments and numerical simulations highlighted the importance of relative suction between the walls, shear-induced dilation and regained strength outside the walls and partial drainage in the co-seismic period. The use of surrogate modelling to automatically optimise parameter selection for the advanced constitutive model was successfully explored. Ultimately, focussing on the mechanics of the dual row walls has helped further the academic and practical understanding of these complex but life-saving systems.

Seismic safety of underground structures (culvert, subway, natural gas and water sewage systems) plays a major role in sustainable public safety and urban development. Very few experimental data are currently available and there is not generally accepted procedure to estimate the dynamic pressures acting on underground structures. This study aims to enhance the state of prevalent information necessary in understanding the dynamic behavior of box culverts and the stresses acting under dynamic excitations through experimental analyses. For this purpose, a series of shaking table tests were conducted on box-type culverts buried in dry sand. To simulate the free-field boundary conditions, a laminar box was designed and manufactured for use in a 1-g shake table. Four culvert models having different rigidities were tested under various harmonic motions in order to examine the effect of flexibility ratio on dynamic lateral soil pressures. Based on the tests results, a simplified dynamic pressure distribution acting on sidewalls of the culvert model was suggested. Then, a dynamic lateral coefficient was defined for the proposed peak pressure value in the distribution. The values of this coefficient were obtained as a function of shear strain and relative stiffness between the soil and underground structure. Finally, a simplified frame analysis approach was suggested for the assessment of the forces on the structure, to help to carry out a preliminary design of box-type culverts. In this approach, it was assumed that the culvert was fixed at bottom and subjected to lateral stresses on sidewalls and shear stresses on the upper face. For the confirmation of the method, centrifuge tests were conducted on a box-type culvert model under the Seventh Framework Programme of European Union with Grant Agreement No.227887. Results show that the proposed simplified procedure can be used in reasonable accuracy as a practical approach for the preliminary assessment of box-type culverts buried in dry sand under seismic action.

Orientadores: Aloisio Ernesto Assan, Persio Leister de Almeida Barros / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Civil, Arquitetura e Urbanismo / Made available in DSpace on 2018-08-11T01:26:45Z (GMT). No. of bitstreams: 1 BobadillaGuadalupe_Ulises_D.pdf: 1151597 bytes, checksum: 005fbfba6e1ebd787b740be4f94cbf34 (MD5) Previous issue date: 2008 / Resumo: As características que determinam o comportamento de uma estrutura sob carregamento dinâmico são as massas dos vários elementos, a rigidez dos seus membros e a dissipação de energia. Para avaliar corretamente a resposta dinâmica de uma estrutura levando em conta os efeitos da interação, é necessário incorporar as propriedades dinâmicas do solo dentro da formulação matemática do modelo físico adotado. Referido à superestrutura, os efeitos da interação alteram a resposta estrutural final, devido à inter-relação dinâmica entre o movimento do solo e o movimento da base de fundação. Conseqüentemente, se primeiro se avalia o movimento da base de fundação [produto da interação solo-estrutura (SSI)], a resposta estrutural final poderá ser resolvida depois via análise modal da superestrutura. Esta conceituação é utilizada no presente trabalho. Aqui, todo o processo de análise é feito no domínio da freqüência. A resposta estrutural é avaliada através da chamada Transformada Implícita de Fourier (ImFT), implementando-se para isto um algoritmo computacional que avalia a resposta dinâmica utilizando a ImFT eficientemente. A ImFT é uma avaliação racional das matrizes envolvendo as transformadas discretas de Fourier (DFT), para num mesmo processo matricial achar a resposta dinâmica estrutural diretamente no domínio do tempo. Correntemente, para a análise no domínio da freqüência tem-se utilizado a FFT (Fast Fourier Transform); embora a FFT seja computacionalmente eficiente, apresenta-se aqui a ImFT, um outro processo computacional alternativo à FFT e bastante eficiente para certos tipos de carregamento tais como uma excitação sísmica, que é o carregamento utilizado nesta pesquisa / Abstract: The characteristics that determine the behavior of a structure under dynamic loading are the masses of various elements, the rigidity of its members and the dissipation of energy. To properly evaluate the dynamic response of a structure taking into account the effects of the interaction, it is necessary to incorporate the dynamic properties of the soil within the mathematical formulation of the physical model adopted. Referred to the superstructure, the effects of interaction modify the final structural response due to the dynamic interrelationship between the soil motion and the base of foundation motions. Consequently, if that first assesses the base of foundation motions [product of the soil-structure interaction effects (SSI)], the final structural response can be assessed later by modal analysis of the superstructure. This concept is used in this work. Here, the whole process of analysis is done in frequency domain. The structural response is evaluated by the so-called Implicit Fourier Transform (ImFT), implementing to this a computational algorithm that assesses the structural dynamic response using the ImFT efficiently. The ImFT is a rational assessment of matrices involving the Discrete Fourier Transform (DFT) for a same matrix process find structural dynamics response directly on time domain. Commonly, for the analysis in the frequency domain has been used the FFT (Fast Fourier Transform), although the FFT is efficiently, presents itself here the ImFT, another alternative computational algorithm and quite competent to certain types of loading such as a seismic excitation which is the loading used in this study / Doutorado / Estruturas / Doutor em Engenharia Civil

L'interaction dynamique sol-structure a été largement explorée en supposant le comportement linéaire du sol. Néanmoins, pour des séismes d'intensité modérée à forte, la contrainte de cisaillement maximale peut facilement atteindre la limite élastique du sol. Du point de vue de l'interaction sol-structure, les effets non linéaires peuvent modifier la rigidité du sol à la base de la structure ainsi que la quantité d'énergie dissipée dans le sol. En conséquence, ignorer les caractéristiques non linéaires du sol dans l'interaction dynamique sol-structure (IDSS) peut conduire à des prédictions erronées de la réponse de la structure. Le but de ce travail est d'implémenter dans un code numérique une loi de comportement non linéaire pour le sol afin d'examiner l'effet de la nonlinéarité du sol sur l'interaction dynamique sol-structure. De plus, différents aspects sont pris en compte tels que l'effet de la contrainte de confinement sur le module de cisaillement du sol, les conditions statiques initiales, les conditions d'interface entre le sol et la structure, etc. Durant ce travail, une méthode simple de couche absorbante basée sur une formulation de Rayleigh / Caughey pour l'amortissement, qui est généralement disponible dans les logiciels existants d'éléments finis, a également été développée. Les conditions de stabilité des problèmes de propagation d'onde sont étudiées et on montre que les comportements linéaire et non linéaire sont très différents en ce qui concerne la dispersion numérique. La règle habituelle de 10 points par longueur d'onde, recommandée dans la littérature pour les milieux élastiques, apparaît pas suffisante dans le cas non linéaire.Le modèle implémenté est d'abord vérifié numériquement en comparant les résultats avec ceux d'autres codes numériques connus. Après cela, une étude paramétrique est menée pour différents types de structures et des profils de sol variés afin de caractériser les effets non linéaires. Différentes caractéristiques de l'IDSS sont comparées à celles du cas linéaire: modification de l'amplitude et du contenu fréquentiel des ondes se propageant dans le sol, fréquence fondamentale, dissipation de l'énergie dans le sol et réponse du système sol-structure. A travers ces études paramétriques nous montrons qu'en fonction des propriétés du sol, le contenu fréquentiel de la réponse du sol peut changer significativement à cause des nonlinéarités de comportement. Les pics de la fonction de transfert entre le champ libre et le rocher affleurant se décalent vers les basses fréquences et l'amplification se produit dans cette gamme de fréquences. Une réduction de l'amplification pour les hautes fréquences et même une dé-amplification peuvent se produire pour un fort niveau des mouvements d'entrée. Ces changements influencent la réponse de la structure. Ce travail montre également que la proximité des fréquences fondamentales de la structure et du sol influence fortement l'interaction sol-structure. Enfin, l'effet du poids de la structure et du balancement de la superstructure peut être significatif. Finalement, le bassin de Nice est utilisé comme un exemple de propagation d'onde dans un milieu non linéaire hétérogène et d'interaction dynamique sol-structure. La réponse du bassin dépend fortement de la combinaison de la nonlinéarité du sol, des effets topographiques et du contraste d'impédance entre les couches de sol. Pour les structures et les profils de sol sélectionnés dans ce travail, les simulations numériques réalisées montrent que le décalage de la fréquence fondamentale n'est pas un bon indicateur pour distinguer le comportement linéaire du sol du comportement non linéaire / The interaction of the soil with the structure has been largely explored the assumption of material and geometrical linearity of the soil. Nevertheless, for moderate or strong seismic events, the maximum shear strain can easily reach the elastic limit of the soil behavior. Considering soil-structure interaction, the nonlinear effects may change the soil stiffness at the base of the structure and therefore energy dissipation into the soil. Consequently, ignoring the nonlinear characteristics of the dynamic soil-structure interaction (DSSI) this phenomenon could lead toerroneous predictions of structural response. The goal of this work is to implement a fully nonlinear constitutive model for soils into anumerical code in order to investigate the effect of soil nonlinearity on dynamic soil structureinteraction. Moreover, different issues are taken into account such as the effect of confining stress on the shear modulus of the soil, initial static condition, contact elements in the soil-structure interface, etc. During this work, a simple absorbing layer method based on a Rayleigh / Caughey damping formulation, which is often already available in existing. Finite Element softwares, is also presented. The stability conditions of the wave propagation problems are studied and it is shown that the linear and nonlinear behavior are very different when dealing with numerical dispersion. It is shown that the 10 points per wavelength rule, recommended in the literature for the elastic media is not sufficient for the nonlinear case. The implemented model is first numerically verified by comparing the results with other known numerical codes. Afterward, a parametric study is carried out for different types of structures and various soil profiles to characterize nonlinear effects. Different features of the DSSI are compared to the linear case : modification of the amplitude and frequency content of the waves propagated into the soil, fundamental frequency, energy dissipation in the soil and the response of the soil-structure system. Through these parametric studies we show that depending on the soil properties, frequency content of the soil response could change significantly due to the soil nonlinearity. The peaks of the transfer function between free field and outcropping responsesshift to lower frequencies and amplification happens at this frequency range. Amplificationreduction for the high frequencies and even deamplication may happen for high level inputmotions. These changes influence the structural response.We show that depending on the combination of the fundamental frequency of the structureand the the natural frequency of the soil, the effect of soil-structure interaction could be significant or negligible. However, the effect of structure weight and rocking of the superstructurecould change the results. Finally, the basin of Nice is used as an example of wave propagation ona heterogeneous nonlinear media and dynamic soil-structure interaction. The basin response isstrongly dependent on the combination of soil nonlinearity, topographic effects and impedancecontrast between soil layers. For the selected structures and soil profiles of this work, the performed numerical simulations show that the shift of the fundamental frequency is not a goodindex to discriminate linear from nonlinear soil behavior

Στην παρούσα εργασία, πραγματοποιείται η μελέτη του φαινομένου της δυναμικής αλληλεπίδρασης εδάφους-κατασκευής όταν αυτή θεμελιώνεται επί εύκαμπτων, επιφανειακών θεμελιώσεων με ή χωρίς πασσάλους. Η μέχρι σήμερα μελέτη της αλληλεπίδρασης εδάφους-κατασκευής έχει περιορισθεί κυρίως σε επιφανειακά, άκαμπτα ή εύκαμπτα θεμέλια και επιφανειακές ή βαθιές, άκαμπτες θεμελιώσεις επί πασσάλων. Ο συνδυασμός εύκαμπτων θεμελιώσεων, επιφανειακών ή βαθιών, και πασσαλώσεων, κοινή πρακτική σε πλείστες κατασκευές, δεν έχει μελετηθεί λόγω της περιπλοκότητας του φαινομένου και των σχετικών αβεβαιοτήτων όσον αφορά στα δεδομένα του προβλήματος. Συνεπώς, η κρισιμότητα των διαφόρων παραμέτρων, και των πολλαπλών αλληλεπιδράσεων τους, στη σεισμική απόκριση των κατασκευών παραμένει εν πολλοίς άγνωστη. Στόχος της εργασίας είναι η ανάπτυξη ενός διακριτού προσομοιώματος για τον εδαφικό ημιχώρο, το οποίο θα δύναται να χρησιμοποιηθεί για την δυναμική ανάλυση εύκαμπτων θεμελίων με ή χωρίς πασσάλους. Σε αυτή την μελέτη χρησιμοποιείται αποκλειστικά η μέθοδος των πεπερασμένων στοιχείων (ΜΠΣ), για την διακριτοποίηση της κατασκευής, του εδάφους και του συστήματος θεμελίωσης. Ειδικότερα, με τη χρήση του πακέτου πεπερασμένων στοιχείων ACS SASSI, διακριτοποιείται ο εδαφικός ημιχώρος και υπολογίζεται το πεδίο μετακινήσεων της επιφάνειας του ημιχώρου για διάφορες συχνότητες που αντιστοιχούν σε συνήθεις σεισμικές διεγέρσεις. Με γνωστό το πεδίο μετακινήσεων, υπολογίζονται οι δυναμικές δυσκαμψίες του ημιχώρου συναρτήσει της συχνότητας και της απόστασης από το σημείο φόρτισης και δημιουργείται ένα νέο διακριτό προσομοίωμα που αντικαθιστά το έδαφος, με βάση το μητρώο δυναμικών δυσκαμψιών. Τέλος με κατάλληλη αδιαστατοποίηση, προκύπτουν κλειστές εκφράσεις για τα στοιχεία του μητρώου δυναμικών δυσκαμψιών, που παρέχουν το επιπλέον πλεονέκτημα ότι είναι ανεξάρτητες της συχνότητας διέγερσης και έτσι γίνεται εφικτή η απευθείας ανάλυση συστημάτων αλληλεπίδρασης εδάφους – κατασκευής στο πεδίο του χρόνου. Το προτεινόμενο προσομοίωμα του ελαστικού ημιχώρου επεκτείνεται στην περίπτωση του μεμονωμένου πασσάλου σε εδαφικό ημιχώρο. Το διακριτό προσομοίωμα που αναπτύχθηκε προηγουμένως είναι κατάλληλο για εισαγωγή σε πρόγραμμα πεπερασμένων στοιχείων γενικού σκοπού ( όπως π.χ. ANSYS, SAP2000, ABAQUS ) και μπορεί να χρησιμοποιηθεί για κάθε είδος επιφανειακής εύκαμπτης ή άκαμπτης θεμελίωσης ανεξαρτήτου γεωμετρίας, υποκαθιστώντας τον ελαστικά γραμμικό εδαφικό ημιχώρο. Με αυτό τον τρόπο γίνεται εφικτή η επίλυση οποιουδήποτε συστήματος αλληλεπίδρασης εδάφους – κατασκευής. Η ακρίβεια και αποτελεσματικότητα του προτεινόμενου προσομοιώματος για τον εδαφικό ημιχώρο, γίνεται φανερή από μια σειρά συγκρίσεων με αποτελέσματα παλαιοτέρων δημοσιεύσεων για άκαμπτα και εύκαμπτα θεμέλια και κατασκευών με ή χωρίς πασσάλους. / During the last decades the problem of dynamic soil-structure interaction (SSI), has received considerable attention due to the large number of installations and structures sensitive to dynamic excitations such as multistory buildings, bridges, nuclear reactors, platforms etc. In fact every structure that lies on a deformable soil medium, experiences the effects of dynamic SSI. Although the phenomenon is thoroughly studied for a variety of soil-structure systems the main research effort has been focused on the dynamic behavior of rigid foundations. The assumption of foundation rigidity is so popular because it simplifies the solution procedure. In contrast, the dynamic analysis of flexible foundations adds more parameters to the problem, by requiring the discretization of both the soil medium and the foundation itself. In this work, the main goal is the study of the dynamic response of flexible foundations (e.g. mat foundations, raft foundations) and the development of a new discrete model for the analysis of soil-foundation-structure systems, which incorporates the effects of foundation flexibility. The problem under consideration consists of determining the dynamic response of the surface of the soil medium idealized as a linear elastic, isotropic half space. To this end the Finite element Method (FEM) is utilized throughout this work. The FEM computer code ACS SASSI is used to simulate the half space. Using the dynamic response of half space surface due to a unit harmonic load, calculated at a dense network of surface nodal points, the impedances of the half space are computed. Using these values a simplified, frequency independent discrete impedance matrix is constructed. Results of the above analysis are presented for a range of frequencies and half-space material properties, in easy to use graphs. This model is expanded to incorporate the case of a single flexible pile embedded on a homogenous half space. The above discrete model can easily be imported to any standard general purpose FEM code (e.g., ANSYS, SAP2000), where the solution of the soil-structure system can be obtained for a variety of soil-structure systems. The main advantages of this model are 1) Accurate, easy and fast discretization of the soil medium 2) No limitation exist due to foundation type or geometry 3) Analyses can be performed both in frequency and time domain.

The interaction of the soil with the structure has been largely explored the assumption of material and geometrical linearity of the soil. Nevertheless, for moderate or strong seismic events, the maximum shear strain can easily reach the elastic limit of the soil behavior. Considering soil-structure interaction, the nonlinear effects may change the soil stiffness at the base of the structure and therefore energy dissipation into the soil. Consequently, ignoring the nonlinear characteristics of the dynamic soil-structure interaction (DSSI) this phenomenon could lead toerroneous predictions of structural response. The goal of this work is to implement a fully nonlinear constitutive model for soils into anumerical code in order to investigate the effect of soil nonlinearity on dynamic soil structureinteraction. Moreover, different issues are taken into account such as the effect of confining stress on the shear modulus of the soil, initial static condition, contact elements in the soil-structure interface, etc. During this work, a simple absorbing layer method based on a Rayleigh / Caughey damping formulation, which is often already available in existing. Finite Element softwares, is also presented. The stability conditions of the wave propagation problems are studied and it is shown that the linear and nonlinear behavior are very different when dealing with numerical dispersion. It is shown that the 10 points per wavelength rule, recommended in the literature for the elastic media is not sufficient for the nonlinear case. The implemented model is first numerically verified by comparing the results with other known numerical codes. Afterward, a parametric study is carried out for different types of structures and various soil profiles to characterize nonlinear effects. Different features of the DSSI are compared to the linear case : modification of the amplitude and frequency content of the waves propagated into the soil, fundamental frequency, energy dissipation in the soil and the response of the soil-structure system. Through these parametric studies we show that depending on the soil properties, frequency content of the soil response could change significantly due to the soil nonlinearity. The peaks of the transfer function between free field and outcropping responsesshift to lower frequencies and amplification happens at this frequency range. Amplificationreduction for the high frequencies and even deamplication may happen for high level inputmotions. These changes influence the structural response.We show that depending on the combination of the fundamental frequency of the structureand the the natural frequency of the soil, the effect of soil-structure interaction could be significant or negligible. However, the effect of structure weight and rocking of the superstructurecould change the results. Finally, the basin of Nice is used as an example of wave propagation ona heterogeneous nonlinear media and dynamic soil-structure interaction. The basin response isstrongly dependent on the combination of soil nonlinearity, topographic effects and impedancecontrast between soil layers. For the selected structures and soil profiles of this work, the performed numerical simulations show that the shift of the fundamental frequency is not a goodindex to discriminate linear from nonlinear soil behavior

In the built environment, human activities such as railway and road traffic, constructionworks or industrial manufacturing can give rise to ground borne vibrations. Such vibrations become a concern in urban areas as they can cause human discomfort or disruption of vibration sensitive equipment in buildings. In Sweden, geological formations of soft clay soils overlying till and a high quality bedrock are encountered in densely populated areas, which are soil conditions that are prone to high levels of ground borne vibrations. Under such soil conditions, end-bearing piles are often used in the design of building foundations. The dynamic response of a building is governed by the interaction between the soil and the foundation. It is therefore essential that models used for vibration predictions are able to capture the dynamic soil-structure interaction of pile foundations. The purpose of this thesis is to experimentally and numerically investigate dynamic soil-structure interaction of an end-bearing pile group in clay by constructing a test foundation of realistic dimensions. The small-strain properties in a shallow clay deposit are estimated using different site investigation and laboratory methods. The results are synthesised into a representative soil model to compute the free-field surface response, which is validated with vibration measurements performed at the site. It is found that detailed information regarding material damping in the clay and the topmost soil layer both have a profound influence on the predicted surface response, especially with an increasing distance from the source. Dynamic impedances of four end-bearing concrete piles driven at the site are measured. Pile-soil-pile interaction is investigated by measuring the response of the neighbour piles when one of the piles in the group is excited. The square pile group is subsequently joined in a concrete cap and measurements of the impedances of the pilegroup and acceleration measurements within the piles at depth are performed. A numerical model based on the identified soil properties is implemented and validated by the measurements. A good agreement between the predicted and measured responses and impedances of the pile group foundation is found, establishing confidence in the ability to predict the dynamic characteristics of end-bearing pile foundations under the studied soil conditions. / Mänsklig verksamhet i urbana miljöer så som väg- och järnvägstrafik, byggnation eller maskindrift inom industri kan ge upphov till vibrationer som sprider sig via marken i närområdet. Dessa vibrationer kan ge upphov till kännbara vibrationer eller påverka vibrationskänslig utrustning i byggnader. I Sverige förekommer ofta mjuka lerjordar ovanpå berg, och inte sällan i tätbebyggda områden. Under sådana jordförhållanden används ofta spetsbärande pålar för grundläggning av byggnader. Det dynamiska verkningssättet för byggnader är beroende av interaktionen mellan jorden och byggnadens grund. Det är därför viktigt att modeller som används för vibrationsanalys i byggnader kan beskriva denna interaktion mellan jord och byggnadsfundament. Syftet med denna avhandling är att experimentellt och via numeriska modeller studera dynamisk jord-struktur-interaktion av ett spetsbärande pålfundament i lera. Jordensmekaniska egenskaper vid små töjningar utvärderas för en lerjord som är avsatt på morän och berg genom både fältförsök och laboratorieanalyser av prover. Informationen kombineras för att konstruera en lagerförd jordmodell av platsen för att beräkna jordens dynamiska respons till följd av en punktlast. Modellen valideras med vibrationsmätningar som utförts på platsen. Studien visar att detaljerad information angående lerans materialdämpning och de mekaniska egenskaperna av jordens översta lager har en stor inverkan på förutsägelser av jordens dynamiska respons vid ytan, speciellt vid stora avstånd från vibrationskällan. Experimentella tester utförs för att mäta dynamiska impedanser av fyra slagna spetsbärande betongpålar. Interaktionen mellan pålarna utvärderas genom att utföra mätningarav de omgivande pålarnas respons till följd av excitering av en påle. Pålgruppen sammanfogas därefter i ett betongfundament och impedanserna samt accelerationer inuti pålarna uppmäts. En numerisk modell baserad på de identifierade mekaniska egenskaperna av jorden upprättas och valideras genom mätningarna. De numeriska resultaten är i god överensstämmelse med de uppmätta vilket styrker användningen av numeriska modeller för att förutsäga interaktionen mellan jord och spetsbärandepålar under de studerade jordförhållandena. / <p>QC 20210302</p>

Thesis deals with problems of the soil-structure interaction. In the theoretical part is described the approach to mathematical modeling of structure-foundation-soil interaction. The subsoil models are further described in detail, including the models with piles (both static and dynamics models). In the next chapter there is described the dynamics theory of the systems with single or more degrees of freedom. There is also an analysis of propagation, reflection and refraction of mechanical one-dimensional waves (P-wave, S-wave) and spatial waves (P- wave, SV-wave, SH-wave) and waves in homogeneous half-space (R-wave L-wave). The numerical analysis is logically sorted from hand calculation of the parameter change influence on the modal characteristics to complex computational FEM model of the machine with a foundation on piles placed in the spatial block of soil. Numerical studies aim to determine the influence of the subsoil model on the modal characteristics and thus confirm the absolute necessity of the subsoil model in tasks of dynamics. The next goal is to determine the appropriate key parameters of the computational model: the size of finite element, suitable shape of subsoil model, suitable inclination of boundary condition and suitable boundary conditions. For creating of set of computational models was used language APDL in conjunction with ANSYS software interface. All used input files are listed in the Annex.