京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第21081号 / 工博第4445号 / 新制||工||1691(附属図書館) / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 杉浦 邦征, 教授 KIM Chul-Woo, 教授 清野 純史 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

No / The causal and physically realizable Biot hysteretic model proves to be the simplest linear model able to describe the nearly rate-independent behaviour of engineering materials. In this paper, the performance of the Biot hysteretic model is analysed and compared with those of the ideal and causal hysteretic models. The Laguerre polynomial approximation (LPA) method, recently proposed for the time-domain analysis of linear viscoelastic systems, is then summarized and applied to the prediction of the dynamic response of linear hysteretic systems to deterministic and random excitations. The parameters of the LPA model generally need to be computed through numerical integrals; however, when this model is used to approximate the Biot hysteretic model, closed-form expressions can be found. Effective step-by-step procedures are also provided in the paper, which prove to be accurate also for high levels of damping. Finally, the method is applied to the dynamic analysis of a highway embankment excited by deterministic and random ground motions. The results show that in some cases the inaccuracy associated with the use of an equivalent viscous damping model is too large.

No description available.

Apesar de fortes eventos sísmicos serem raros no Brasil, engenheiros estruturais brasileiros são frequentemente envolvidos em tal análise para os países latino-americanos vizinhos. Informações sobre históricos de aceleração sísmica, de natureza aleatória, não estão em geral disponíveis, devido, em parte, à falta de registros. Para contornar tal situação, os códigos de construção indicam o uso de acelerogramas artificiais, mas não fornecem metodologia para sua obtenção. A informação normalizada é o chamado espectro de resposta elástico, que fornece a aceleração de resposta máxima para um sistema linear de um grau de liberdade. Muitas pesquisas estão sendo desenvolvidas a fim de gerar acelerogramas artificiais compatíveis com os espectros de norma. Assim, neste trabalho se apresenta uma proposta para a geração de acelerograma artificial compatível com espectro de resposta regulamentar. Para exemplo de aplicação, é gerado um acelerograma artificial compatível com a Norma Brasileira NBR 15421: 2006 e aplicado à base de um edifício shear building de dez pavimentos e através de integração numérica por diferenças finitas passo-a-passo no domínio do tempo é calculado o deslocamento do último pavimento deste edifício. De forma semelhante, é gerado um acelerograma artificial compatível com a Norma Venezuelana COVENIN 1756: 2001 e aplicado à base de um reservatório d\'água sobre quatro pilares e estudado o seu comportamento elastoplástico perfeito. / Although strong seismic events are rare in Brazil, Brazilian structural engineers are often involved in such an analysis for neighboring Latin American countries. Information on seismic acceleration histories of a random nature is not generally available, due in part to the lack of records. To circumvent such a situation, building codes indicate the use of artificial accelerograms, but do not provide a methodology for obtaining them. The normalized information is the so-called elastic response spectrum, which provides the maximum response acceleration for a linear system of a degree of freedom. Many researches are being developed in order to generate artificial accelerograms compatible with the norm spectra. Thus, this paper presents a proposal for the generation of an artificial accelerogram compatible with a regulatory response spectrum. For an application example, an artificial accelerogram compatible with the Brazilian Standard NBR 15421: 2006 is generated and applied to the base of a ten-story shear building and through numerical integration by finite differences step-by-step in the time domain is calculated the displacement of the last floor of this building. Similarly, an artificial accelerogram is generated that is compatible with the Venezuelan Standard COVENIN 1756: 2001 and applied on the basis of a water reservoir on four pillars and studied its perfect elastoplastic behavior.

The seismic response of steel monopole wind turbine towers is investigated and their risk is assessed in the Canadian seismic environment. This topic is of concern as wind turbines are increasingly being installed in seismic areas and design codes do not clearly address this aspect of design. An implicit finite element model of a 1.65MW tower was developed and validated. Incremental dynamic analysis was carried out to evaluate its behaviour under seismic excitation, to define several damage states, and to develop a framework for determining its probability of damage. This framework was implemented in two Canadian locations, where the risk was found to be low for the seismic hazard level prescribed for buildings. However, the design of wind turbine towers is subject to change, as is the design spectrum. Thus, a methodology is outlined to thoroughly investigate the probability of reaching predetermined damage states under seismic loading for future considerations.

The seismic response of steel monopole wind turbine towers is investigated and their risk is assessed in the Canadian seismic environment. This topic is of concern as wind turbines are increasingly being installed in seismic areas and design codes do not clearly address this aspect of design. An implicit finite element model of a 1.65MW tower was developed and validated. Incremental dynamic analysis was carried out to evaluate its behaviour under seismic excitation, to define several damage states, and to develop a framework for determining its probability of damage. This framework was implemented in two Canadian locations, where the risk was found to be low for the seismic hazard level prescribed for buildings. However, the design of wind turbine towers is subject to change, as is the design spectrum. Thus, a methodology is outlined to thoroughly investigate the probability of reaching predetermined damage states under seismic loading for future considerations.

[pt] A indústria de geração elétrica de fonte nuclear baseia-se em princípios de segurança e, nos critérios de projeto, considera-se a hipótese de terremoto. Os sistemas necessários à segurança são projetados para resistir e manter a operabilidade durante e após eventos sísmicos postulados. Propõe-se um roteiro para a produção de espectros de resposta sísmica para projeto de sistemas secundários, SS, incluindo a influência do acoplamento e em base probabilística. O roteiro utiliza a ferramenta SASSI, pode ser utilizado em situações gerais de cálculo e fornece um conjunto de programas para considerar modelos tridimensionais e suas respostas para uma excitação genérica em 3 direções ortogonais; representar os efeitos de acoplamento entre o SS e o sistema principal, SP; incluir a influência dos deslocamentos relativos entre os nós de apoio do sistema secundário no sistema principal; utilizar os fatores de transposição entre espectros elásticos e inelásticos; permitir a análise probabilística e a obtenção de Espectros de Resposta Uniformemente Prováveis, acoplados ou não; incluir interfaces para a utilização de seus resultados com outros programas de utilização geral, como o MS-EXCEL. O elevado grau de automatização permite a produção de espectros de resposta com refinamentos de modelagem, alcançando uma análise mais realista, sem a necessidade de esforços adicionais aos já requeridos pela metodologia convencional. A metodologia proposta enquadra-se no encaminhamento para o contexto atual de análise sísmica de instalações nucleares, com a utilização de espectros de resposta de projeto de ameaça uniforme, específico para o sítio da instalação, e o projeto sísmico de risco consistente. / [en] The electric power reactor industry is based on rigid safety principles. The design criteria include seismic scenario. All safety related systems are designed to resist and to keep the operability during and after a postulated earthquake. It is suggested a procedure for the generation of in-structure seismic response spectra for secondary system design. A probabilistic approach is used and coupling effects between primary and secondary systems are taken into account. The proposed script uses SASSI system and can be used in general situations. A set of computer programs is developed to consider three-dimensional models and their responses for a generic base excitation, acting in 3 orthogonal directions; represent the coupling effect between primary and secondary systems, include the influence, on the response spectra, of the secondary system supports relative displacements; include approximated factors for transposition of elastic into inelastic response spectra; produce Uniformly Probable Response Spectra, including or not coupling effects; consider interfaces with other general programs, as the MS-EXCEL, for pos-processing purpose. The degree of automation, allows the production of response spectra including modeling refinements, reaching a more realistic analysis, without additional efforts beyond those already required by the conventional methodology. The proposed methodology is in the way of a site specific uniform hazard design response spectra, and of a consistent-risk seismic design.

Apesar de fortes eventos sísmicos serem raros no Brasil, engenheiros estruturais brasileiros são frequentemente envolvidos em tal análise para os países latino-americanos vizinhos. Informações sobre históricos de aceleração sísmica, de natureza aleatória, não estão em geral disponíveis, devido, em parte, à falta de registros. Para contornar tal situação, os códigos de construção indicam o uso de acelerogramas artificiais, mas não fornecem metodologia para sua obtenção. A informação normalizada é o chamado espectro de resposta elástico, que fornece a aceleração de resposta máxima para um sistema linear de um grau de liberdade. Muitas pesquisas estão sendo desenvolvidas a fim de gerar acelerogramas artificiais compatíveis com os espectros de norma. Assim, neste trabalho se apresenta uma proposta para a geração de acelerograma artificial compatível com espectro de resposta regulamentar. Para exemplo de aplicação, é gerado um acelerograma artificial compatível com a Norma Brasileira NBR 15421: 2006 e aplicado à base de um edifício shear building de dez pavimentos e através de integração numérica por diferenças finitas passo-a-passo no domínio do tempo é calculado o deslocamento do último pavimento deste edifício. De forma semelhante, é gerado um acelerograma artificial compatível com a Norma Venezuelana COVENIN 1756: 2001 e aplicado à base de um reservatório d\'água sobre quatro pilares e estudado o seu comportamento elastoplástico perfeito. / Although strong seismic events are rare in Brazil, Brazilian structural engineers are often involved in such an analysis for neighboring Latin American countries. Information on seismic acceleration histories of a random nature is not generally available, due in part to the lack of records. To circumvent such a situation, building codes indicate the use of artificial accelerograms, but do not provide a methodology for obtaining them. The normalized information is the so-called elastic response spectrum, which provides the maximum response acceleration for a linear system of a degree of freedom. Many researches are being developed in order to generate artificial accelerograms compatible with the norm spectra. Thus, this paper presents a proposal for the generation of an artificial accelerogram compatible with a regulatory response spectrum. For an application example, an artificial accelerogram compatible with the Brazilian Standard NBR 15421: 2006 is generated and applied to the base of a ten-story shear building and through numerical integration by finite differences step-by-step in the time domain is calculated the displacement of the last floor of this building. Similarly, an artificial accelerogram is generated that is compatible with the Venezuelan Standard COVENIN 1756: 2001 and applied on the basis of a water reservoir on four pillars and studied its perfect elastoplastic behavior.

The emphasis on seismic design and assessment of reinforced concrete (RC) frame structure has shifted from force-based to performance-based design and assessment to accommodate strength and ductility for required performance of building. RC frame structure may suffer different levels of damage under seismic-induced ground motions, with potentials for formation of hinges in structural elements, depending on the level of stringency in design. Thus it is required to monitor the seismic behaviour and performance of buildings, which depend on the structural system, year of construction and the level of irregularities in the structural system. It is the objective of the current research project to assess seismic performance of RC frame buildings in Canada, while developing fragility curves as analytical tools for such assessment. This was done through dynamic inelastic analysis by modelling selected building structures and using PERFORM-3D as analysis software, while employing incremental dynamic analysis to generate performance data under incrementally increasing seismic intensity of selected earthquake records. The results lead to probabilistic tools to assess the performance of buildings designed following the National Building Code of Canada in different years of construction with and without irregularities. The research consists of three phases; i) regular buildings designed after 1975, ii) regular buildings designed prior to 1975, and iii) irregular buildings designed prior to 1975. The latter two phases address older buildings prior to the development of modern seismic building codes. All three phases were carried out by selecting and designing buildings in Ottawa, representing the seismic region in eastern Canada, as well as buildings in Vancouver, representing the seismic region in western Canada. Buildings had three heights (2; 5; and 10-stories) to cover a wide range of building periods encountered in practice. The resulting fragility curves indicated that the older buildings showed higher probabilities of exceeding life safety and/or collapse prevention performance levels. Newer buildings showed higher probabilities of exceeding target performance levels in western Canada than those located in the east.

L'analyse sismique des structures de génie civil est une problématique majeure pour la sécurité des personnes et la pérennité des ouvrages. L'étude expérimentale permet de comprendre le comportement réel de la structure mais occasionne des problèmes de coût important et d'effet d'échelle souvent inévitable dû aux dimensions des structures. D'un autre côté, l'étude numérique propose une bonne approximation du comportement global mais la représentation précise des phénomènes locaux (fissuration, perte de matière, flambement, grands déplacements) dans les zones fortement endommagées est délicate et souvent insuffisante. Ce travail de thèse propose l'élaboration d'une technique de sous-structuration hybride pour coupler un modèle numérique à une plateforme expérimentale. Ainsi, la partie faiblement endommagée de la structure est modélisée numériquement tandis que la partie fortement endommagée est testée expérimentalement. Cette méthode permet de coupler le réalisme de l'expérimental avec le faible coût numérique sans toutefois perdre en précision. Après avoir élaboré une méthode de couplage hybride peu intrusive pour le code de calcul (Cast3m), un modèle d'endommagement anisotrope adapté aux chargement sismique (effet unilatéral, déformations permanentes) est développé dans le cadre de la thermodynamique des milieux continus. Afin de valider la méthode hybride, une étude expérimentale est menée sur une structure type en béton armé. La fissuration de la partie expérimentale est étudiée grâce à la corrélation d'images. Ce travail expose donc une alternative intéressante aux analyses classiques des structures importantes soumises à des sollicitations complexes. / The seismic analysis of civil engineering structures is a major problem for the safety of the persons and the sustainability of the structures. The experimental study allows to understand the real behavior of the structure but causes problems of important cost and often inevitable scale effect owed in dimension of the structures. On the other hand, the numerical study proposes a good estimate of the global behavior but the accurate modelling of the local phenomena (cracking, losses of material, buckling, large displacements) in the strongly damaged zones is delicate and often insufficient. This work of this thesis proposes the elaboration of a hybrid technique of sub-structuring to couple a numerical model with an experimental platform. So, the weakly damaged part of the structure is numerically modelled whereas the strongly damaged part is experimentally tested. This method allows to couple the precision and the realism of the experimental with the numerical moderate cost without losing however in precision. Having elaborated a few intrusive hybrid method of coupling for the code of calculation (Cast3m), a anisotropic damage model adapted for seismic load (unilateral effect, permanent strains) is developed within the framework of the thermodynamics of the continuous media. To validate the hybrid method, an experimental study is led on a typical reinforced concrete structure. The cracking of the experimental part is studied thanks to images correlation. This work thus exposes an interesting alternative to the classic analyses of the important structures subjected to complex loading.