Neste trabalho são realizadas análises estruturais de edifícios de paredes de concreto moldadas no local e de alvenaria estrutural considerando a interação solo-estrutura e a sequência construtiva. Com solução de fundação em estacas pré-moldadas, cada edifício piloto com 45 metros de altura, formado por 15 pavimentos de parede com pé-direito de 2,80m, apoiado sobre um pilotis de concreto armado de 3 metros de altura é modelado com base em 4 metodologias de análise: (i) O AI_AF, modelo clássico de referência, que admite apoios indeslocáveis e ações instantâneas; (ii) O AE_AF, modelo que inclui as ações evolutivas, incorporando o aumento gradativo de carregamento e rigidez; (iii) O AI_ISE, modelo que incorpora a interação com o solo a partir da aplicação instantânea de ações e (iv) o AE_ISE, modelo mais refinado, que considera a interação com o solo no tempo de construção. As paredes são modeladas em elementos finitos de casca, os pilares de concreto, vigas de transição, estacas e blocos em elementos finitos de barra e o maciço de solo em elementos finitos sólidos isoparamétricos, com o auxílio do software comercial DIANA®. O trabalho evidencia que o modelo clássico de referência, que desconsidera a interação com o solo, não alerta para a necessidade de aumentar a ductilidade das paredes dos pavimentos iniciais em ambos os sistemas construtivos. / This study consist of a structural analyses of concrete walls and masonry building including the soil-structure interaction and the construction process. Each pilot building is 45 meters high, consisting of 15 floors with 2.80 m high. Each floors are seated on pillars of reinforced concrete with 3 meters of height, which were modeled using 4 methodologies: (i) The AI_AF, classic reference model, which adopts fixed foundations and instantaneous action; (ii) The AE_AF is a model, that includes construction loads and incorporates a gradual increasing in load and stiffness; (iii) The AI_ISE model incorporates interaction with the soil and the instantaneous application of actions; and, (iv) AE_ISE, which consists of a more refined model with soil interaction and the time of construction. The concrete walls are modeled based on shell finite elements, the concrete pillars, transition beams, stakes and blocks are modeled based on bar finite element and the soil mass is modelled as an isoparametric solid finite elements. The numerical modelling is conducted using commercial software DIANA®. Results show that the classic reference model, in which the soil-structure interaction is not considered, does not attent to the need of increasing the ductility of the walls in initials floor.

Neste trabalho, é desenvolvido um programa de computador para a análise estática e tridimensional de problemas de interação solo-estrutura. O programa permite considerar várias camadas de solo, cada qual com características físicas diferentes. Sobre este solo, o qual pode conter estacas, podem ser apoiados diversos tipos de estruturas, tais como placas e até um edifício. Todos os materiais considerados são homogêneos, isotrópicos, elásticos e lineares. O solo tridimensional é modelado com o método dos elementos de contorno (MEC), empregando as soluções fundamentais de Kelvin e uma técnica alternativa na consideração do maciço não-homogêneo. Esta técnica, que é uma contribuição original deste trabalho, é baseada no relacionamento das soluções fundamentais de deslocamento dos diferentes domínios, permitindo que sejam analisados como um único sólido sem a necessidade de equações de equilíbrio e compatibilidade. Isso reduz o sistema de equações final e melhora a precisão dos resultados, conforme comprovado nos exemplos apresentados. Para reduzir o custo computacional sem prejudicar a precisão dos resultados, é utilizada uma malha de elementos de contorno infinitos (ECI) nas bordas da malha de ECs para modelar o comportamento das variáveis de campo em longas distâncias. A formulação do ECI mapeado utilizado é outra contribuição original deste trabalho, sendo baseado em um EC triangular. É demonstrado por meio de exemplos que tal formulação é eficiente para a redução de malha, contribuindo de forma significativa na redução do custo computacional. Todas as estruturas que interagem com o solo, incluindo as de fundação, são simuladas empregando o método dos elementos finitos (MEF). Cada estaca é modelada como uma linha de carga empregando um único elemento finito com 14 parâmetros nodais, o qual utiliza funções de forma do quarto grau para aproximar os deslocamentos horizontais, do terceiro grau para as forças horizontais e deslocamentos verticais, do segundo grau para as forças cisalhantes verticais e constantes para as reações da base. Este elemento é empregado em outros trabalhos, no entanto os autores utilizam as soluções fundamentais de Mindlin na consideração da presença da estaca no solo. Desta forma, a formulação desenvolvida neste trabalho com as soluções fundamentais de Kelvin pode ser considerada mais uma contribuição original. No edifício, que pode incluir um radier como estrutura de fundação, são utilizados dois tipos de EFs. Os pilares e vigas são simulados com elementos de barra, os quais possuem dois nós e seis graus de liberdade por nó. As lajes e o radier são modelados empregando elementos planos, triangulares e com três nós. Nestes EFs triangulares são superpostos efeitos de membrana e flexão, totalizando também seis graus de liberdade por nó. O acoplamento MEC/MEF é feito transformando as cargas de superfície do MEC em carregamentos nodais reativos no MEF. Além de exemplos específicos nos Capítulos teóricos, um Capítulo inteiro é dedicado a demonstrar a abrangência e precisão da formulação desenvolvida, comparando-a com resultados de outros autores. / In this work, a computer code is developed for the static analysis of three-dimensional soil-structure interaction problems. The program allows considering a layered soil, which may contain piles. This soil may support several structures, such as shells or even an entire building. All materials are considered homogeneous, isotropic, elastic and linear. The three-dimensional soil is modeled with the boundary element method (BEM), employing Kelvin fundamental solutions and an alternative multi-region technique. This technique, which is an original contribution of this work, is based on relating the displacement fundamental solution of the different domains, allowing evaluating them as an unique solid and not requiring compatibility or equilibrium equations. In such a way, the final system of equations is reduced and more accurate results are obtained, as demonstrated in the presented examples. In order to reduce the computational cost maintaining the accuracy, an infinite boundary element (IBE) mesh is employed at the BE mesh limits to model the far field behavior. The mapped IBE utilized, based on a triangular EC, is another original contribution of this work. In the presented examples it is demonstrated that this IBE formulation is efficient for mesh reduction, implying on a significant computational cost reduction. All structures that interact with the soil, including the foundations, are simulated with de finite element method (FEM). The piles are modeled using a one-dimensional 14 parameter finite element, with forth degree shape functions for horizontal displacement approximation, third degree shape functions for horizontal forces and vertical displacement, second degree shape functions for vertical share force, and constant for the base reaction. This element is employed in other works, however the authors utilize Mindlin fundamental solutions for the pile presence consideration in the soil. In such a way, the formulation developed in this work with Kelvin fundamental solutions may be considered one more original contribution. The building, which may include a radier as a foundation structure, is modeled using two types os FEs. Piles and beams are simulated using bar FEs with two nodes and six degrees of freedom per node. The radier and pavements are modeled employing plane triangular three-node FEs. In these FEs plate and membrane effects are superposed, totalizing six degrees of freedom per node. FEM/BEM coupling is made by transforming the BEM tractions in nodal reactions in the FEM. Even though specific examples are presented in the theoretical Chapters, a role Chapter is dedicated for demonstrating the formulation accuracy and coverage. In most examples, the results are compared with the ones obtained by other authors.

Este trabalho busca aplicar técnicas de confiabilidade ao problema de grupo de estacas utilizadas como fundação de estruturas correntes. Para isso, lança-se mão de um modelo tridimensional de interação estaca-solo onde estão presentes o Método dos Elementos de Contorno (MEC) e o método dos Elementos Finitos (MEF) que atuam de forma acoplada. O MEC, com as soluções fundamentais de Mindlin (meio semi-infinito, homogêneo, isotrópico e elástico-linear é utiliza), é utilizado para modelar o solo. Já o MEF é utilizado para modelar as estacas. Definido o modelo de funcionamento estrutural das estacas, parte-se para a aplicação de métodos trazidos da confiabilidade estrutural para avaliação da adequabilidade em relação aos estados limite de serviço e estados limites últimos. Os métodos de confiabilidade utilizados foram o Método de Monte Carlo, o método FOSM (First-Order Second-Moment) e o método FORM (First-Order Reliability Method). / This work seeks to apply reliability techniques to the problem of piles groups used as current structures foundation. For this, makes use of a three-dimensional model of pile-soil interaction with the boundary element method (BEM) and the finite element method (FEM) working coupled. The BEM, with Mindlin fundamental solutions (semi-infinite medium, homogeneous, isotropic and linear elastic) is used to model the soil. The MEF is used to model the piles. Defined the model of structural functioning of the piles, the aim goes to the application of structural reliability for assessing the adequacy of the serviceability limit states and ultimate limit states. Reliability methods used were the Monte Carlo method, the FOSM (First-Order Second-Moment) method and the FORM method (First-Order Reliability Method).

O presente trabalho apresenta uma formulação misto do MEC (Método dos Elementos de Contorno) e o MEF (Método dos Elementos Finitos). Nessa formulação, as estacas são modeladas através do MEF como elementos de barra e o solo através do MEC, como um meio contínuo, elástico linear, isótropo e homogêneo, utilizando as soluções fundamentais de MINDLIN (1936). Os sistemas de equações do solo e das estacas para elementos verticais são apresentados como uma combinação de ambos, originando um único sistema final de equações. Apresentam-se também as modificações necessárias para um sistema composto por estacas inclinadas. Após a resolução do sistema final, obtém-se os deslocamentos e as tensões de contato solo-estaca. A seguir, apresentam-se alguns exemplos numéricos obtidos a partir da formulação proposta e compara-se com modelos de outros autores. / This work presents a hybrid formulation of BEM (Boundary Elements Method) and FEM (Finite Elements Method). In that formulation, the piles are modeled through FEM as bar elements and the soil through BEM, as an isotropic, homogeneous, semi-infinite and linear-elastic continuum, using the fundamental solutions of MINDLIN (1936). The systems of equations of the soil and of the piles for vertical elements are presented as a combination of both, originating a single final system of equations. Some modifications are accomplished for the system of inclinated piles. After the resolution of the final system, the displacements and the contact tensions between soil and pile are obtained. Numeric examples are obtained starting from the proposed formulation and to proceed they are compared with other authors\' models.

As galerias enterradas são amplamente utilizadas em obras de arte corrente (OAC) na infraestrutura de rodovias e ferrovias. Visando disponibilizar mais alternativas na produção de galerias de concreto pré-moldado, a pesquisa apresenta uma análise comparativa de custos para seções transversais não usuais, direcionado a grandes profundidades de instalação. A primeira seção proposta é denominada modificada e é composta por uma base retangular com uma cobertura em arco e a segunda seção proposta é definida por três arcos com uma base plana. Essas seções além de possuir formato de geometria favorável à distribuição dos esforços solicitantes, mantêm o benefício de possibilitar uma compactação adequada do solo na lateral da galeria, como acontece em galerias retangulares convencionais, e isso garante o confinamento do solo e a redistribuição de pressões, graças à mobilização do efeito de arqueamento. Para considerar a interação solo-estrutura foram realizadas análises via elementos finitos com o pacote computacional GeoStudio® próprio para análises geotécnicas. Os resultados comprovaram a interferência do formato das geometrias no comportamento da interação solo-galeria e também evidenciaram que a redução da espessura das paredes da galeria mobiliza em maior intensidade a capacidade resistente do solo e isso reduz os esforços na estrutura. Quanto à análise dos dimensionamentos, com base na comparação dos custos evidenciou-se que, para as galerias modificadas a redução das taxas de armadura está relacionada com flecha do arco da cobertura, sendo que a economia verificada no custo total de materiais das seções estudadas variou de 4 a 29% comparado à galeria retangular. Para a galeria definida por três arcos, confirmou-se que quanto mais alongado for o seu formato, melhor é o comportamento à grandes profundidades de instalação, sendo que a economia verificada no custo total da galeria mais elíptica (DTA I-b) chegou a 50% em relação à galeria retangular e demais geometrias estudadas tiveram economia entre 17 e 42%. / Box culverts are often used in current drainage in the infrastructure of highways and railways. In order to provide more production\'s alternatives of precast concrete culvert, the research presents the comparative cost analysis for unusual cross sections, directed to large depths of installation. The first proposed cross section called modified culvert is composed of a rectangular base with an arc roof. The second is defined by three arcs with a flat base. These cross sections have geometric shapes favorable to the distribution of bending moment and shear forces. It maintains the benefit of adequate compaction of the backfill at the side of the culvert, as well as in conventional box culvert, which ensures ground confinement and pressure redistribution due to the mobilization of the arching effect. In order to consider the soil-structure interaction, finite element analyzes were performed using GeoStudio® software, which is a computational package specific for geotechnical analysis. The results showed that the geometries have influence on the behavior of the soil-culvert interaction, and that the reduction of the thickness of the culvert walls mobilizes the soil resistant capacity, reducing the stresses in the structure. Regarding the structural design, considering the comparison of costs, it was evidenced a reduction of steel reinforcements for the modified culvert, which is related to the arrow of the arc of the cover. The materials saving verified in the total cost of the studied sections ranged from 4% to 29% compared to the box culvert. For the culvert defined by three arches, it was evidenced that the more elongated its geometry, better the behavior for the great depths of installation. The highest materials saving was verified in the total cost of the most elliptical culvert (DTA I-b), which reached a value of 50%. Compared to the box culvert, the others geometries studied had savings between 17-42%.

Neste trabalho foi desenvolvida uma formulação alternativa para o acoplamento entre o método dos elementos de contorno (MEC) e o método dos elementos finitos (MEF) para análise não linear geométrica de estruturas reticuladas ligadas a meios contínuos bidimensionais heterogêneos, aplicado a problemas de interação solo-estrutura. O solo foi considerado com comportamento elástico linear e modelado via MEC por meio de uma formulação alternativa à clássica técnica de sub-região permitindo a consideração de múltiplas inclusões mais ou menos rígidas do que o material padrão e de linhas de carga internas aos domínios. Este código foi então acoplado ao programa AcadFrame, baseado no MEF posicional para análise não linear geométrica de pórticos com consideração de cinemática exata. O acoplamento numérico foi realizado por meio de uma formulação algébrica onde a matriz de rigidez do solo e a força de contato são condensadas e somadas à matriz e ao vetor de forças internas da estrutura a cada iteração no processo de Newton-Raphson. Em ambos os programas foi utilizada uma generalização do grau de aproximação dos elementos através dos polinômios de Lagrange, o que permite a utilização de elementos curvos de alta ordem. Foi utilizada ainda a técnica dos mínimos quadrados para reduzir as oscilações de forças de superfície no contato. Os resultados obtidos de forma geral são bastante satisfatórios e comprovam a eficiência da formulação. O trabalho permite a análise de problemas de edificações apoiadas sobre solos estratificados com múltiplas inclusões e linhas de carga. Permite tanto a análise de elementos apoiados diretamente sobre o solo (sapatas, radies) quanto de elementos internos e em qualquer direção, como no caso de estacas verticais ou inclinadas. Pode-se inclusive considerar as estacas passando por diferentes camadas de solo. A aplicação pode ser estendida ainda a outros problemas elásticos, acoplamento entre peças mecânicas e análise de materiais compostos. / This work presents an alternative coupling of the boundary element method (BEM) and the finite element method (FEM) to create a computer program for non linear geometric analysis of frames coupled to continuous domains, applied to soil-structure interaction. A linear elastic behavior is considered for the soil, modeled by BEM. An alternative formulation is adopted for the classic sub-region technique, allowing the consideration of multiple inclusions and load lines inside the soil domain. The BEM computational code is coupled to the AcadFrame software, based on positional FEM for non linear geometric analysis of frames, considering exact kinematics. The numerical coupling is made by an algebraic formulation where the soil stiffness matrix and contact forces are condensed and added to the structure matrix and internal forces for each iteration on Newton-Raphson process. On both programs it is adopted a generalization of the element degree assuming the Lagrange polynomials, which allows the use of curved high order elements. It was also implemented the least square method in order to obtains better and smoother results of surface forces in the contact interface. The obtained results are satisfactory and prove the formulation efficiency. The program allows the analysis of buildings supported by layered soils with multiples inclusions and load lines. It considers directly supported elements over the soil (footing foundations, radies) and internal elements in any direction, like vertical and diagonal piles. It can also consider piles going through different layers of the soil. This formulation can be applied to other elastic problems like coupling between mechanic pieces and composite material analysis.

Rapid growth in population along with positive trends in global economy over the past several decades has significantly contributed to an increased demand for various infrastructure needs worldwide. For this reason, the focus of this thesis has been directed towards extending the mechanics of unsaturated soils, which is an emerging geotechnical engineering field to investigate the behavior of two key infrastructure systems, namely pile foundations and energy pipeline systems. The mechanism of soil-pile foundations and soil-pipeline systems interaction behavior has several similarities. Both these infrastructure facilities require comprehensive understanding of the soil-structure interaction mechanism. Reliable estimation of mechanical properties of both the soil and the soil-structure interface is required for the rational interpretation the load-displacement behavior of pile foundations and pipeline systems. Currently, the design of systems is predominantly based on design codes and guidelines that use empirical procedures or employ the principles of saturated soil mechanics. In many scenarios, pile foundations extend either totally or partly in unsaturated soils as the groundwater table level in many regions is at a greater depth. Such scenarios are commonly encountered in semi-arid and arid regions of the world. In addition, pipeline systems are typically buried at shallow depths in unsaturated soil strata, which are susceptible to wetting and drying, freezing and thawing cycles or both, due to seasonal environmental changes. Capillary stress or matric suction in the unsaturated zone increases the effective stress contribution towards the shear strength and stiffness of soil and soil-structure interface. Extending saturated soil mechanics to design or analyze such structures may lead to erroneous estimation of pile foundation carrying capacity or loads transferred on pipeline body from the surrounding unsaturated soil. Experimental, analytical and numerical investigations were undertaken to study the behavior of single pile, pile group, and pipeline systems in saturated and unsaturated sands under static loading. The experimental program includes 40 single model pile and 2×2 pile group, and six prototype pipeline tests under saturated and unsaturated condition. The results of the experimental studies suggest that matric suction has significant contribution towards the mechanical behavior of both pile foundation and pipeline system. The axial load carrying capacity of single pile and pile group increased approximately 2 to 2.5 times and the settlement reduced significantly compared to saturated condition. The influence of matric suction towards a single pile is significantly different in comparison to pile group behavior. The cumulative influence of matric suction and stress overlap of pile group behavior in sandy soils result in erroneous estimation of pile group capacity, if principles of saturated soil mechanics are extended. Group action plays major role in changing the moisture regime under the pile group leading to incompatible stress state condition in comparison to single pile behavior. On the other hand, the peak axial load on the pipe is almost 2.5 folds greater in unsaturated sand that undergoes much less displacement in comparison to saturated condition. Such an increase in the external axial forces may jeopardize the integrity of energy pipeline systems and requires careful reevaluation of existing design models extending the principles of unsaturated soil mechanics. Two analytical design models to estimate the axial force exerted on pipeline body were proposed. The proposed models take account of matric suction effect and soil dilatancy and provide smooth transition from unsaturated to saturated condition. These models were developed since measurement of the unsaturated soil and interface shear strength and stiffness properties need extensive equipment that require services of trained professional, which are expensive and time consuming. The models utilize the saturated soil shear strength parameters and soil-water characteristic curve (SWCC) to predict the mechanical behavior of the structure in saturated and unsaturated cohesionless soils. The prototype pipeline experimental results were used to verify the proposed models. The predicted axial force on pipeline using the proposed models agrees well with the measured behavior under both saturated and unsaturated conditions. Moreover, numerical techniques were proposed to investigate the behavior of pile foundation and pipeline system in saturated and unsaturated sand. The proposed methodology can be used with different commercially available software programs. Two finite element analysis programs were used in this study; namely, PLAXIS 2D (2012) to simulate soil-pile foundation behavior and SIGMA/W (2012) to simulate soil-pipeline system behavior. The proposed techniques require the information of unsaturated shear strength and stiffness, which can be derived from saturated soil properties and the SWCC. The model was verified using pile and pipeline test results from this study and other research studies from the published literature. There is a good agreement between the measured behavior and the predicted behavior for both the saturated and unsaturated conditions. The methodology was further extended to investigate the behavior of rigid and flexible pipelines buried in Indian Head till (IHT) during nearby soil excavation activity. The simulation results suggest that excavation can be extended safely without excessive deformation to several meters without the need for supporting system under unsaturated condition. The studies summarized in the thesis provide evidence that the principles of saturated soil mechanics underestimate the pile foundations carrying capacity as well as the axial force exerted on pipelines in unsaturated soils. Such approaches lead to both uneconomical pile foundation and unsafe pipeline systems designs. For this reason, the pile and pile group carrying capacity and pipeline axial force should be estimated taking into account the influence of matric suction as well as the dilatancy of the compacted sand. The experimental studies, testing techniques along with the analyses of test results and the proposed analytical and numerical models are useful for better understanding the pile foundation and buried pipeline behaviors under both saturated and unsaturated conditions. The proposed analytical and finite element models are promising for applying the mechanics of unsaturated soils into conventional geotechnical engineering practice using simple methods.

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.

Tunnelling in urban areas causes short-term and long-term ground movements under existing buildings. Finite element analysis provides a useful option for assessing the likely extent of damage induced in these buildings. Although finite element analysis is suggested to be used in the later phases of the building damage assessment procedures employed in practice, only the effect of short-term ground movements is typically considered and there are no detailed guidelines on the specification and complexity of the modelling. This thesis addresses the tunnel-soil-building interaction problem and the effect of long-term consolidation, as well as demonstrating the application of 3D finite element analysis with appropriate simplifications for practical assessment purposes. Finite element models are developed to quantify the effect of shallow tunnelling on an example masonry building founded on strip footings, considering both single- and twin-tunnel scenarios in a typical London soil profile. Total stress and effective stress analyses are adopted with specific modelling procedures to focus on the short-term and long-term response respectively. The analyses use a non-linear model for the masonry, and allow slippage and gapping at the soil-footing interface. Two advanced constitutive models for the soil (the extended Mohr-Coulomb model and the modified two-surface kinematic hardening model) are implemented with customized stress update schemes. The finite element results present the interaction between the soil and the building by comparing with the greenfield ground response. The horizontal coupling between the foundation and the ground is shown to be relatively weak. The dominant deformation mode of the building varies with the tunnel configuration (i.e. single or twin tunnels) and the tunnel eccentricity. Strain localization is found around the explicitly modelled window and door openings. The long-term consolidation is sensitive to the permeability of the tunnel lining. The building response to long-term ground movements is further affected by the tunnel-tunnel interaction in the case of twin-tunnel configuration. Performing 3D analysis of a single facade and foundation provides useful damage predictions, without the need to model a complete building. The proposed result processing methods such as characteristic strain and damage bar chart are practical tools for assessment. The study highlights some limitations of the elastic beam assessment method, which is often adopted in the early phase of the damage assessment process.

Relatos da observação do comportamento de obras de engenharia em escala natural, especialmente no caso de edifícios apoiados em fundações profundas, são pouco freqüentes em nossa literatura, embora estimulados por vários autores e pela Norma Brasileira de Projeto e Execução de Fundações. Este trabalho apresenta resultados de medidas de carga e recalque em 20 pilares contíguos de uma estrutura em concreto armado pré-moldada, apoiada em fundações do tipo estaca cravada. Os recalques foram medidos por meio de nivelamento ótico de precisão, tendo sido determinado valores máximos variáveis entre 1,1 e 4,3 mm. Foram observados deslocamentos sob carga constante, fluência da fundação, com taxa variável entre 0,8 e 3,2 mm/log t. As solicitações normais nos pilares foram avaliadas indiretamente por meio da variação de seu comprimento, utilizando-se um extensômetro mecânico removível. Apresenta-se a metodologia para interpretação das medidas efetuadas pelo extensômetro mecânico, levando em conta as variações dos fatores ambientais e a reologia do concreto, a qual conduz a uma boa concordância entre os valores medidos e os fornecidos pelo cálculo estrutural. Os dados coletados são retroanalisados sob o enfoque da interação solo estrutura pela modelagem da superestrutura por meio de pórtico espacial apoiado em molas representativas das fundações por estacas. No cálculo das molas foi utilizada a integração numérica da solução de Mindlin para a modelagem do efeito de grupo do sistema formado pelas estacas e o maciço de solo. Verificou-se que a elevada variabilidade dos solos da formação geológica do local foi refletida no comportamento da obra, como mostra o resultado da retroanálise efetuada. Conclui-se que as variabilidades da formação geotécnica devem ser consideradas para que previsões de comportamento sejam mais realistas / Although encouraged by several authors and by the Brazilian Foundation Code, reports of actual column loads measurement in natural scale are not frequent in our technical literature, especially in the case of buildings supported by deep foundations. Results of load and settlement measurements at 20 contiguous columns of a structure built in pre-cast reinforced concrete and supported by driven piles are presented. Settlements were measured by means of optical level and a range of values between 1.1 to 4.3 mm were observed. Settlement under constant load were observed under variable creep rates from 0.8 to 3.2 mm/log t. Loads over columns were indirectly evaluated through column length variation, using a demountable mechanical extensometer. The methodology for interpretation of measurements made with the mechanical extensometer is described, considering corrections due to the variation of environmental conditions and to the concrete’s rheology; this methodology leads to a good agreement between measured values and those supplied by conventional structural design. The collected data is back analysed taking into account the soil structure interaction. The superstructure was modelled as a spatial frame supported by springs with the same rigidity of the pile foundation element. The equivalent spring parameter for each column support has considered the settlement group effect for all piles embedded in soil, using the numerical integration of Mindlin's equations. Results of this back analysis show a high variability, reflecting the high degree of variability of local subsoil conditions. Therefore, one can conclude that predictions, in order to be reliable, must consider these soil variations