Fatigue cracking is one of the primary modes of distress in asphalt pavements that has an important economic impact. Fatigue resistance characterization of an asphalt mixture is a complex issue due to: (i) composite nature of the material, (ii) gradation of aggregate particles, (iii) variation of asphalt film thickness, (iv) air voids distributions, (v) asphalt binder nonlinear viscoelastic behavior, (vi) effects of binder oxidative aging as a function of time, and (vii) micro crack healing during rest periods. Different methods to assess fatigue cracking in asphalt materials are available in the literature. However, there is no methodology to characterize fatigue cracking behavior of asphalt materials that is independent of the mode of loading (controlled-strain or controlled-stress). The objective of this research is to develop a new methodology to characterize fatigue cracking of the fine aggregate matrix (FAM) portion of asphalt mixtures using dynamic mechanical analyses (DMA). This is accomplished through different, but related, approaches. The first approach relies on identifying the various mechanisms of energy dissipation during fatigue cracking that are manifested in: (i) nonlinear viscoelastic deformation, (ii) fracture, and (iii) permanent deformation. Energy indices were derived to quantify each of these energy dissipation mechanisms and to quantify fatigue cracking irrespective of the mode of loading. The first outcome of the approach is a fatigue damage parameter (crack growth index) that provides comparable results for a given material even when tested under different modes of loading and different load (strain or stress) amplitudes. The developed fatigue characterization method has a lower coefficient of variation when compared to conventional parameters (number of load cycles to failure or cumulative dissipated energy). The crack growth index parameter was also qualitatively and quantitatively compared to three dissipated energy methods available in the literature. The second outcome of this research is a constitutive model that can describe both asphalt mixtures’ nonlinear viscoelastic response and fatigue damage in one formulation. Nonlinear viscoelastic as well as damage parameters were obtained for both modes of loading. This second approach has the advantage that the constitutive model can be implemented in a numerical framework to describe the response of asphalt mixtures under various boundary conditions.

The deleterious effect of moisture on the structural integrity of asphalt mixtures has been recognized as one of the main causes of early deterioration of asphalt pavements. This phenomenon, usually referred to as moisture damage, is defined as the progressive loss of structural integrity of the mixture that is primarily caused by the presence of moisture in liquid or vapor state. Moisture damage is associated with the development of different physical, mechanical, and chemical processes occurring within the microstructure of the mixture at different intensities and rates. Although there have been important advancements in identifying and characterizing this phenomenon, there is still a lack of understanding of the damage mechanisms occurring at the microscopic level. This situation has motivated the research work reported in this dissertation. The main objective of this dissertation is to formulate and apply a numerical micromechanical model of moisture-induced damage in asphalt mixtures. The model focuses on coupling the effects of moisture diffusion—one of the three main modes of moisture transport within asphalt mixtures—with the mechanical performance of the microstructure. Specifically, the model aims to account for the effect of moisture diffusion on the degradation of the viscoelastic bulk matrix of the mixture (i.e., cohesive degradation) and on the gradual deterioration of the adhesive bonds between the aggregates and the asphalt matrix (i.e., adhesive degradation). The micromechanical model was applied to study the role of some physical and mechanical properties of the constitutive phases of the mixtures on the susceptibility of the mixture to moisture damage. The results from this analysis suggest that the diffusion coefficients of the asphalt matrix and aggregates, as well as the bond strength of the aggregate-matrix interface, have the most influence on the moisture susceptibility of the mixtures. The micromechanical model was further used to investigate the influence of the void phase of asphalt mixtures on the generation of moisture-related deterioration processes. Two different probabilistic-based approaches were used to accomplish this objective. In the first approach, a volumetric distribution of air voids sizes measured using X-Ray Computed Tomography in a dense-graded asphalt mixture was used to generate probable void structures in a microstructure of an asphalt mixture. In the second approach, a stochastic modeling technique based on random field theory was used to generate probable air voids distributions of the mixture. In this second approach, the influence of the air voids was accounted for by making the physical and mechanical properties of the asphalt matrix dependent on probable voids distributions. Although both approaches take into consideration the characteristics of the air void phase on the mechanical response of the mixtures subjected to moist environments, the former explicitly introduces the air phase within the microstructure while the latter indirectly includes its effects by modifying the material properties of the bulk matrix. The results from these simulations demonstrated that the amount, variability and location of air voids are decisive in determining the moisture-dependent performance of asphalt mixtures. The results from this dissertation provide new information on the kinetics of moisture damage mechanisms in asphalt mixtures. In particular, the results obtained from applying the micromechanical model permitted identification of the relative influence of the characteristics of the constitutive phases of a mixture on its moisture-related mechanical performance. This information can be used as part of design methodologies of asphalt mixtures, and/or as an input in life-cycle analysis models and maintenance programs of road infrastructure.

Asphalt bound mixtures have been put to diverse uses. The complicated nature of the material and the demanding conditions under which it is used preclude complete solutions to questions on load bearing capability under field conditions. In proportion to the quantity of its usage and in acknowledgment of modeling complexity, the material has been interrogated by many researchers using a variety of mechanical tests, and a plethora of linear viscoelastic models have been developed. Most models are intended to account for specific classes of problems. This work addresses the conspicuous absence of systematic documentation of normal forces generated as a result of shear. The normal force generated during simple shear is a clear indication of the nonlinear nature of the material. The effect of fillers (hydrated lime and limestone), air voids, aggregate gradation, asphalt source and step loading on normal force generation during torsion is experimentally investigated. Based on experimental evidence, a non-linear thermomechanical model for sandasphalt mixtures based on the idea of multiple natural configurations is developed. The model accounts for the fact that the mixture has a natural configuration (stressfree configuration) which evolves as it is subjected to loads. Assumptions are made regarding the manner in which the material stores and dissipates energy. A key assumption is that among the various constitutive relations possible, the one that is chosen is the one that maximizes the rate of entropy production. The model that is developed accounts for the anisotropic nature of the response. The experimental results show that asphalt bound mixtures generate significant normal forces even at low rotation rates. The source of asphalt, aggregate gradation, fillers and air voids have a pronounced effect on normal stress generation. The model is corroborated against data from torsion experiments.

Esta pesquisa apresenta os resultados obtidos em análise laboratorial de misturas asfálticas de módulo elevado para aplicação em camada de base de pavimento. Historicamente, pode-se dizer que este tipo de mistura asfáltica em camadas de base resulta em melhor comportamento mecânico comparado com o de misturas asfálticas usinadas com CAP convencional. As misturas de módulo elevado apresentam vantagens em relação às convencionais com maior resistência à deformação permanente e maior vida de fadiga. Inicialmente, foram realizados ensaios laboratoriais de caracterização dos agregados e de três tipos de ligantes asfálticos: CAP A e CAP B, ambos ligantes duros com penetração entre 10 e 20dmm, e um CAP convencional, CAP 30/45. Foram dosadas três misturas asfálticas, com uma graduação preestabelecida, e com os três ligantes asfálticos estudados, pelos métodos Marshall e SUPERPAVE. Os teores de projeto pelo método Marshall foram de 0,2 a 0,3% superiores ao calculado pelo SUPERPAVE. Verificou-se o atendimento das especificações francesas quanto ao volume de vazios das misturas compactadas com os CAPs A e B pelo compactador giratório francês PCG). Foram determinados os danos por umidade induzida, tendo sido verificado que a mistura com CAP B requer um melhorador de adesividade. Determinaram-se os afundamentos por deformação permanente, as resistências à tração, os módulos de resiliência em dois equipamentos diferentes (MTS e UTM), os módulos dinâmicos, e a vida de fadiga por compressão diametral e por flexão em viga de 4 pontos, das três misturas asfálticas estudadas. As deformações permanentes para as misturas com CAP A e CAP B são menores que o limite máximo estabelecido pelas diretrizes europeias, e inferiores ao valor obtido com o mistura com CAP 30/45. Os módulos de resiliência mostraram rigidez distinta entre as misturas do módulo elevado, sendo que a mistura com CAP B é duas vezes mais rígida que a mistura com CAP A, que por sua vez é similar ao módulo com CAP 30/45. Obtiveram-se resultados similares de módulo de resiliência com equipamentos distintos. Os resultados dos ensaios de módulo dinâmico demonstraram proximidade de comportamento da mistura asfáltica com CAP A com misturas asfálticas com ligantes modificados; o módulo dinâmico da mistura com CAP B é similar àquele com CAP 30/45. Os resultados dos dois ensaios de fadiga empregados neste estudo demonstraram o melhor comportamento de misturas asfálticas compostas por ligantes duros e de módulo elevado, quando comparada com a mistura convencional. Todos os resultados demonstram que há benefícios de misturas de módulo elevado como camada de base em relação às misturas com CAP convencional, e que esta é uma solução com grande potencial para estruturas de pavimento para tráfego pesado. / This study presents the results obtained in laboratory analysis, about high modulus asphalt mixtures for use in base layer of pavement. Historically, it can be said that this type of mixture, when use in base layer, results in better mechanical behavior when compared with conventional asphalts. High modulus mixtures provide better behavior as it related layer`s permanent deformation and a higher fatigue life. Initially, laboratory tests for characterizing aggregates and three types of asphalt binders were carried out: Binder A and Binder B, both hard binders with penetration between 10 and 20dmm, and a conventional binder, namelly 30/45. Three types of asphalt mixtures were designed, with one predefined graduation and with those three asphalt binders, using Marshall and SUPERPAVE methods. Marshall`s binder contents were from 0.2 to 0.3% higher than SUPERPAVE. The compacted asphalt mixtures were in compliance with French specifications in terms of air void content of Binder A and Binder B using the French gyratory compactor (PCG). Moisture-induced damage test was performed and it was verified that Binder B needs an antistrip agent. Rutting, indirect tensile strength, resilient modulus in two different machines (MTS and UTM), dynamic modulus, and fatigue life by diametral compression and four point bending tests, were carried out with the three asphalt mixtures studied. The results of rutting for mixtures with Binders A and B were lower than the maximum limit of European specifications, and lower than the result with Binder 30/45. Resilient modulus showed different stiffness between the high modulus mixtures, where the asphalt mixture with Binder B was twice stiffer than the mixture with Binder A, which in turn is similar to the resilient modulus of the asphalt mixtures with Binder 30/45. Analogous results were observed in different resilient modulus machines. Dynamic modulus results showed similar behavior between the asphalt mixture with Binder A and asphalt mixture with modified binders; dynamic modulus of the asphalt mixture with Binder B was similar to the asphalt mixture with Binder 30/45. Results in the two fatigue tests were better for the asphalt mixtures with hard binders and high modulus in comparison with the conventional asphalt mixture. All the results demonstrate benefits of high modulus mixtures as base layer in pavements in comparison with asphalt mixtures with conventional binders, and the solution has a great potential in pavement structures for heavy traffic.

Ao longo de sua vida útil, com o acúmulo de tráfego e de solicitações ambientais, infraestruturas de transportes degradam-se em termos de desgaste dos seus componentes e deformações permanentes de suas camadas, acarretando na perda gradual da qualidade estrutural e funcional da via. A camada de sublastro, parte integrante da subestrutura ferroviária, possui contribuição significativa no comportamento mecânico global de uma via permanente. Suas principais funções estão relacionadas à drenagem, atenuação e distribuição de tensões, e separação e transição entre as camadas de lastro e subleito. O objetivo principal deste estudo foi comparar duas diferentes configurações de sublastro implementadas em uma ferrovia de carga brasileira através da medição de respostas mecânicas e parâmetros geométricos \"in situ\". O emprego de sublastro betuminoso constituiu-se como a primeira aplicação de campo deste tipo no País e ambas as seções foram monitoradas também por intermédio de sensores de temperatura e de tensão. Procedeu-se a caracterização física e mecânica dos materiais que compõem os trechos analisados em laboratório. Ademais, tensões de sucção foram medidas em diferentes locais e profundidades da camada de subleito, a fim de verificar a eficácia da camada betuminosa em impermeabilizar as camadas subjacentes. Os resultados mostraram uma eficiência global superior do trecho contendo sublastro betuminoso comparativamente à seção de referência (sublastro granular). Variações das tensões de sucção ao longo de duas estações climáticas mostraram que a mistura asfáltica protegeu satisfatoriamente o subleito quanto à infiltração de água oriunda das precipitações. Medidas de temperatura tomadas em diferentes pontos das camadas de sublastro mostraram que a mistura asfáltica se encontra protegida das variações térmicas e da incidência de radiação solar, uma vez que está isolada pela camada de lastro. Em termos de deslocamentos medidos com o equipamento DMD (Dispositivo para Medição de Deslocamentos), mostrou-se uma redução crescente no deslocamento vertical médio da seção com sublastro betuminoso em comparação com a seção com sublastro granular. Por consequência, o módulo de via, u, calculado para ambas as seções, revelou que o perfil contendo mistura asfáltica tende a defletir menos quando solicitado pela passagem de trens. Medidas de parâmetros geométricos utilizando o equipamento Trolley AMBER apontaram uma relação bidirecional entre o comportamento mecânico e a qualidade da geometria da via, observada antes e após intervenção mecanizada de socaria. / In the course of its service life, with traffic loading accumulation in addition to environmental effects, transportation infrastructures tend to degrade in terms of wear of their components and permanent deformation of their layers, resulting in the gradual loss of structural and functional quality of the railway. The subballast plays an important role as part of the railway substructure and it has a significant contribution to the overall mechanical behavior of the track. Its main functions are related to drainage, to load attenuation and distribution, and to the separation and transition between the ballast and the subgrade. The main objective of this study was to compare two different subballast configurations implemented in a freight railway profile in Brazil through measurements of mechanical and geometric parameters in field. The use of bituminous subballast was the first application of this type in the country and both sections were also monitored by means of temperature and pressure sensors. Laboratory tests for physical and mechanical characterization for all materials composing both experimental sections were performed. In addition, the water pressure in terms of suction were measured at different places and depths of the subgrade, in order to verify the effectiveness of the bituminous subballast in waterproofing the underlying layers. The results showed a superior overall efficiency of the section containing bituminous subballast compared to the reference section (granular subballast). The suction tension variation measured along two climatic seasons showed that the asphalt mixture satisfactorily protects the subgrade from the infiltration of water due precipitations. Temperature measurements taken at different points of the subballast layers showed that the asphalt mixture is protected from thermal and solar radiation peaks, since it is isolated by the ballast layer from atmospheric environmental effects. In terms of displacements acquired using the DMD (Displacement Measurement Device) it was observed an increasing reduction in the mean vertical displacement of the section with bituminous subballast compared to the section with granular subballast. Consequently, the track modulus, u, calculated for both sections revealed that the profile containing asphalt mixture tends to deflect less when requested by the train passage. Measurements of geometric parameters made with the Trolley AMBER device highlighted the bidirectional relationship between the mechanical behavior and the quality of the track geometry, observed before and after tamping procedure.

Asphalt concrete pavements on commercial airports in the United States are constructed according to the Federal Aviation Administration Advisory Circular 150/5370-10B, Item P-401, “Plant Mix Bituminous Pavements.” This specification does not provide guidance for using the Superpave gyratory compactor in the design of asphalt mixtures. This thesis describes a laboratory study of hot mix asphalt (HMA) mix design for airport pavements that uses the Superpave gyratory compactor. These recommendations are based on comparisons of volumetric property measurements of HMA mixtures compacted using Marshall compaction and Superpave gyratory compaction.

[pt] O fenômeno da fadiga é uma das falhas mais comuns em pavimentos asfálticos e se caracteriza pela aparição e posterior desenvolvimento de micro e macro trincas. As misturas asfálticas são compostas por agregados graúdos e por uma matriz circundante composta por ligante asfáltico, agregados miúdos e fíler, conhecida como Mistura Asfáltica Fina (MAF). Estudos indicam que os fenômenos que afetam as microfissuras se iniciam na mistura asfáltica fina. O presente trabalho investiga o comportamento em relação à fadiga das misturas asfálticas, utilizando como passo intermediário a análise de MAFs. Foram selecionadas três misturas asfálticas, compostas pelo mesmo ligante, porém com agregados distribuídos em três curvas granulométricas diferentes, variando seu tamanho máximo nominal. Para as MAFs, as granulometrias foram proporcionais às encontradas nas respectivas misturas asfálticas. As misturas e as MAFs seguiram programação experimental semelhante, com a caracterização viscoelástica linear e realização de ensaios dinâmicos uniaxiais de fadiga. Para interpretação dos resultados, foi utilizado o modelo simplificado de viscoelasticidade do dano contínuo (S-VECD) e os critérios de ruptura G(R) e D(R), que descrevem satisfatoriamente o comportamento em relação à fadiga das misturas e MAFs estudadas. Foi analisada a resistência ao dano das MAFs, bem como a influência da variação do tamanho máximo nominal no seu desempenho a fadiga. Estudou-se a relação entre as deformações encontradas nas misturas asfálticas e na parcela fina das misturas asfálticas. As misturas com maior presença de mistura asfáltica fina apresentaram melhor desempenho a fadiga e os fatores de proporcionalidade de deformação foram maiores para as misturas com tamanhos máximos nominais maiores. / [en] Fatigue is one of the most common problems in asphalt pavements and it is characterized by the appearance and subsequent development of micro and macro cracks. Asphalt mixtures are composed of large aggregates and a matrix that surrounds these aggregates, composed of asphalt binder, small aggregates and filler, known as Fine Aggregate Mixture (FAM). Studies indicate that the phenomena that affect micro cracks starts in the FAM. Therefore, this work aims to investigate the fatigue behavior of the Asphalt mixtures, using as intermediate means an analysis of the FAM. To achieve this objective, three asphalt mixtures were selected. They are composed of the same binder, but with aggregates distributed according to three different aggregate gradation curves, varying their maximum nominal size. The FAM had gradation curves proportional to that applied in asphalt mixtures. Mixtures and MAFs followed similar experimental programming with linear viscoelastic characterization and dynamic uniaxial fatigue tests. In order to interpret the results it was used the Simplified Viscoelastic Continuum Damage (S-VECD) model, G(R) and D(R) failure criteria, which presented a good prediction of the fatigue behavior of the Mixtures and MAFs studied. The MAF s tolerance to damage was analyzed as well as the influence of the maximum nominal size variation. A relation between strains in the asphalt mixtures and the strain in the fine portion of the asphalt mixtures was studied. The asphalt mixtures with higher content of fine asphalt mixture showed better fatigue performance and the strain proportionality factors were higher for mixtures with larger nominal maximum sizes.

Este trabalho tem como objetivo avaliar o comportamento à fadiga de algumas misturas asfálticas, à luz de resultados de ensaios de fadiga por compressão diametral à tensão controlada, considerando-se a variação de quatro fatores: tipo de agregado, tipo de ligante, tipo de moldagem e tipo de filler. Todos os corpos-de-prova foram moldados na mesma faixa granulométrica (centro da faixa C do DNER ), segundo os procedimentos dos métodos Marshall, ASTM D 4867 ou AASHTO T 283, variando o tipo de agregado (basalto, granito e gabro), tipo de ligante (CAP 20 e CAP 40) e o tipo de material de preenchimento (cal hidratada). São apresentados resultados dos ensaios de resistência à tração por compressão diametral, módulo de resiliência e de fadiga. Os modelos de fadiga são apresentados em função das diferenças de tensões e deformações. Para avaliar o desempenho relativo das misturas em serviço, foram simuladas três estruturas de pavimentos com a ajuda do programa ELSYM5, para as quais foram determinadas as tensões e deformações atuantes que foram utilizadas nos modelos de fadiga para que todas as misturas fossem avaliadas sob o mesmo referencial. Os resultados indicam que em relação ao tipo de moldagem, as misturas moldadas pelo método da AASHTO T 283 apresentaram melhor desempenho (maior vida de fadiga), sendo que a maioria das misturas moldadas com CAP 40 apresentou um desempenho melhor que as moldadas com CAP 20. Com relação ao tipo de agregado os materiais gábricos apresentaram o melhor desempenho, seguidos pelos basálticos e graníticos. As misturas moldadas com cal hidratada como filler também apresentaram melhor desempenho à fadiga. / The purpose of this research is to evaluate the fatigue behavior of some asphalt mixtures, by means of indirect tests with controlled stress, considering the variation of four factors: aggregate type, asphalt type, molded type and filler type. The specimens was molded in same gradation (center of DNER C gradation), according to Marshall, ASTM D 4867 or AASHTO T 283 design method, with the variation of three aggregates (basaltic, granite and gabbro), two asphalt (CAP 20 and CAP 40) and the presence of lime. Tests results of indirect tensile strength, resilient modulus and fatigue are presented. The fatigue models are constructed as a function of the differences of stress and strain. The fatigue behavior of asphalt mixtures was evaluated based on three structures which the stress and strains was estimated with aid of ELSYM5 program. The tests results indicated that all mixtures molded with AASHTO T 283 design method presented higher fatigue life than the other ones, and almost all asphalt mixtures molded with CAP 40 presented higher fatigue life than mixtures molded with CAP 20. Considering the type of aggregate, the specimens molded with gabbro presented higher fatigue life than basaltic and granite. The mixtures molded with presence of lime presented an increase in fatigue life.

The diploma thesis is focused on an experimental design of low-noise asphalt mixtures. The theoretical part describes the problems of noise and methods for noise measurements. There are also specified the characteristics of various low-noise asphalt mixtures. The practical part addresses the experimental design of SMA 8 LA. Subsequently, the mixture is exposed to functional testing and the results are afterwards compared with the conventional type of stone mastix asphalt SMA 8 S. The second part is focused primarily on an experimental mixture design of a specific asphalt mixture type - coated macadam. There is especially solved the issue ofbinder drainage and the proportion of cracked grains during the compaction. The advantage of this mixture is the lower price demands cause by lower binder content and a smaller amount of fine aggregate. This mixture should be used as a base layer for low-loaded (traffic) roads. In the conclusion are thereafter evaluated the knowledge and the experience with the testing and designing of individual mixtures gained during the testing.

Este estudo apresenta uma contribuição à compreensão do fenômeno da adesão e da sua perda que as misturas asfálticas apresentam em conseqüência da ação da água. Para tanto, foram testados diversos tipos de misturas asfálticas à luz dos processos preconizados pela ASTM e pela AASHTO para avaliação de suscetibilidade à umidade. Foram executados vários experimentos fatoriais, que consideraram as principais variáveis influentes no fenômeno da adesão, tais como: tipo de agregado (basalto, gabro e granito), tipo de asfalto (CAP 20 e CAP 40), processo de produção de misturas (Marshall, ASTM e AASHTO), volume de vazios (4 e 7%), teor de asfalto (3) e uso ou não de cal. As propriedades monitoradas foram a resistência à tração e o módulo de resiliência. Entre outras constatações, verificou-se nesta pesquisa que a cal melhorou a adesividade das misturas na maioria dos casos; o procedimento da AASHTO mostrou-se mais efetivo na detecção de suscetibilidade à ação de água nas misturas; ensaios de módulo de resiliência indicaram variações nas características mecânicas dos corpos-de-prova que os ensaios de resistência à tração não conseguiram detectar; as misturas com asfalto tipo CAP 20 apresentaram menos suscetibilidade à umidade que as com CAP 40; o aumento no volume de vazios das misturas redundou em maior perda de adesividade e maiores teores de asfalto levaram as misturas a menores sensibilidades à água. Dentre todas as misturas estudadas, as com granito foram as que exibiram maior sensibilidade aos efeitos da água, independentemente do tipo e do teor de asfalto utilizados. / This study presents a contribution to the understanding of the adhesion phenomenon and its loss when water acts in asphalt mixtures. Several kinds of asphalt mixtures were evaluated by means of ASTM and AASHTO moisture susceptibility tests. Experiments based on factorial design were performed considering some of the main factors that act on adhesion phenomenon, such as: aggregate (basalt, gabbro and granite), asphalt (CAP 20 and CAP 40), laboratorial mixture production procedure (Marshall, ASTM and ASHTO), air voids content (4 and 7%), asphalt content (3 levels) and the presence of lime. Mechanical properties of asphalt mixtures were evaluated by resilient modulus and tensile strength. The main conclusions are: lime improved adhesiveness of asphalt mixtures for all studied cases; AASHTO’s procedure was more effective to diagnostic the moisture susceptibility; resilient modulus test indicated changes in mechanical characteristics of specimens that the tensile strength did not detect; mixtures with CAP 20 exhibited less moisture susceptibility than mixtures with CAP 40; mixtures with high air voids presented higher loss of adhesiveness; mixtures with higher asphalt content tended to exhibit lower adhesiveness losses. In almost all cases, the asphalt mixtures with granite aggregate exhibited more water susceptibility, despite asphalt’s kind and content.