Esse trabalho trata da análise de corpos multi-fraturados utilizando o método dos elementos de contorno. Este método numérico é conhecido por ser robusto e preciso neste tipo de problema e também por requerer pequeno esforço computacional na criação da malha de elementos para o crescimento das fissuras. Duas metodologias para a análise do comportamento das fissuras são consideradas. A primeira é a já consagrada metodologia dual. Por meio desta técnica equações integrais distintas são aplicadas às faces da fissura. Estas equações integrais são escritas em termos de deslocamentos e forças de superfície. A segunda metodologia é a que emprega a formulação singular onde a fissura é considerada como um vazio no domínio sendo as faces da fissura separadas por uma pequena distância. No tocante ao crescimento das fissuras foi desenvolvido um procedimento especial para a determinação da direção de crescimento das fissuras o qual mostrou-se muito eficiente levando a resultados precisos. O crescimento das fissuras é efetuado considerando o fator de intensidade de tensão atuante na extremidade de cada fissura. Dessa forma, as fissuras mais solicitadas apresentam maior comprimento de propagação tornando a análise mais realista. Os fatores de intensidade de tensão são calculados por meio de duas técnicas. A primeira é a já conhecida técnica de correlação de deslocamentos a qual relaciona os deslocamentos atuantes nas faces da fissura. Uma técnica alternativa é também utilizada a qual emprega o campo de tensões presente na extremidade da fissura. Após a determinação dos fatores de intensidade de tensão quatro diferentes teorias de interação de modos podem ser utilizadas para a determinação do ângulo de propagação. Foram analisadas estruturas sendo os resultados comparados aos previstos analiticamente e também numericamente. As respostas obtidas foram satisfatórias validando assim a metodologia proposta neste trabalho / This work deals with analysis of multi-fractured bodies using boundary element method. This numerical method is known to be robust and accurate in this kind of problem and by small computational effort to create elements mesh of crack growth. Two methodologies to analyze of crack behavior are considerate. The first is consecrated dual methodology. Through this technique different kind of integral equations are applied to crack boundaries. These integrals equations are written in displacements and traction variables. Second methodology is singular formulation. Through this technique crack is represented like a hole in body and the crack boundaries is separated by a small gap. For crack growth was created a special proceeding to determination crack growth direction. This method is very efficient and your results are accurate. Crack growth is made through the stress intensity factor performed in crack tip. Then the cracks more requested going to propagate with a larger length’s growth turning this model very realistic. The stress intensity factors are calculated through two techniques in this work. First is the known correlation displacement technique which related displacement in crack boundaries. An alternative technique is also used which consider stress field in crack tip. After determination of stress intensity factors four different theories are used to calculate the crack growth angle. In this work were analyzed structures with results are compared with analytical and numerical answers. The results obtained went very satisfactory validating the methodology proposed

A austenita retida está sempre presente na microestrutura de camada cementada de aços, em maior ou menor quantidade. Como é uma fase dúctil comparada à martensita, sua presença tem sido alvo de muita controvérsia. Este trabalho apresenta um estudo sobre a influência da austenita retida na propagação de trincas curtas por fadiga em camada cementada de aço SAE 8620. Foram feitos ensaios de fadiga por flexão em quatro pontos, a temperatura ambiente, em corpos de prova sem entalhe com três níveis de amplitude de tensão e razão de tensões de 0,1. Através de diferentes ciclos de cementação e tratamentos térmicos, foram obtidas camadas cementadas com quatro níveis de austenita retida na microestrutura. O teor de austenita retida foi medido através da técnica de difração de Raios-X. Trincas superficiais foram monitoradas por meio da técnica de réplicas de acetato. Como resultados foram obtidos tamanho de trinca em função do número de ciclos e taxa de crescimento de trincas curtas. Corpos de prova com maiores níveis de austenita retida apresentaram maior vida em fadiga. / The retained austenite is always present in case carburized steel microstructure in small or high percentages. Since it is a ductile phase, its presence has long been a controversial subject. The influence of retained austenite on short fatigue crack propagation in case carburized SAE 8620 steel was studied in this work. Four-point-bend fatigue tests were carried out at room temperature in specimens without notch using three levels of stress range and a stress ratio of 0.1. Four different amount of retained austenite in the case carburized microstructure were obtained through different cycles of carburizing and heat treating. The retained austenite content was measured by X-ray technique, and the surface short crack growth was monitored by means of acetate replication technique. Crack length versus number of cycles and crack growth rate versus mean crack length were obtained as results. Specimens with higher levels of retained austenite in the carburized case showed longer fatigue life.

Neste trabalho desenvolveremos um modelo estatístico em uma escala micrométrica de interações entre as componentes do sistema que pretende descrever a propagação de trincas em materiais submetidos a tensões cíclicas. Apesar de sua extrema simplicidade, este modelo é capaz de reproduzir um resultado experimental bastante difundido entre engenheiros e especialistas, conhecido como lei de Paris, cujo enunciado estabelece que a taxa de crescimento de uma trinca sob carregamento cíclico é proporcional a uma potência da variação em seu correspondente fator de intensidade de tensões sendo largamente utilizada em aplicações práticas. Estamos particularmente interessados em estudar a introdução de desordem em determinados parâmetros associados ao material investigando as modificações impostas por este tipo de abordagem ao comportamento estatístico do modelo. Nossos principais resultados serão obtidos numericamente a partir de uma aproximação do tipo campo efetivo que ignora a correlação existente entre as diversas trincas que podem se formar ao longo do sistema durante o processo. Simulações numéricas do modelo serão igualmente consideradas ao analisarmos situações mais gerais do processo de propagação em que efeitos associados à regeneração de trincas podem desempenhar um importante papel na descrição do comportamento mecânico de um material. / In this work we consider a statistical model in a micrometric scale of interactions between the components of the system which intends to describe the failure of materials subjected to cyclic-load fatigue. Although quite simple, this model is able to reproduce an important experimental result widespread among engineers and experts, known as Paris law, which states that the growth rate of a crack at subcritical load is proportional to a power of the change in its stress-intensity factor and it is largely used in engineering practice. We are particularly interested to study the introduction of disorder in some parameters of the material investigating the modifications caused by this kind of approach in the statistical properties of the model. Our main results will be obtained numerically assuming an effective-field like approximation which neglects the correlation between the different cracks emerging throughout the system during the breaking process. Numerical simulations of the model are also performed in order to describe more general situations of propagation where the effects of crack self-healing can play an important role in the material strength.

O objetivo desta pesquisa foi verificar a influência de diferentes tamanhos de partículas de carga na tenacidade à fratura (KIc), nos parâmetros do crescimento subcrítico de trinca (n e f0) e de Weibull (m e 0), na longevidade estimada pelo diagrama tensão-probabilidade-tempo (SPT) e no limite de fadiga cíclica (LFC) de compósitos resinosos experimentais. Quatro compósitos foram preparados contendo 78% em massa (59% em volume) de conteúdo inorgânico, constituído por 67% de pó de vidro com diferentes tamanhos de partículas (d50 = 0,5; 0,9; 1,2 e 1,9 µm) e 11% de sílica pirogênica. Dados de KIc obtidos pelo método single-edge notched beam (25x5x2,8 mm; n=15) foram submetidos a ANOVA/teste de Tukey (p < 0,05). n e f0 foram determinados através do ensaio de fadiga dinâmica (10-2 a 102 MPa/s) utilizando um dispositivo de flexão biaxial (12x1,2 mm; n=10). Para determinar m e 0, mais 20 espécimes de cada compósito foram testados na taxa de 100 MPa/s. Os diagramas SPT foram obtidos a partir dos dados da fadiga dinâmica e análise de Weibull. No ensaio de fadiga cíclica, um dispositivo de flexão biaxial (12x1,2 mm) foi utilizado para se obter a resistência à flexão inicial (RFI; n=14) e o LFC (n=20). LFC foi obtido pelo método escada após 105 ciclos. Para todos os testes, os espécimes foram armazenados em água destilada a 37oC por 24h. Foi realizada a fractografia dos espécimes fraturados nas taxas 10-2 e 10-1 MPa/s da fadiga dinâmica e nos ensaios para determinação da RFI e LFC. Houve relação direta entre d50 e KIc (C0,5: 1,2±0,1b; C0,9: 1,3±0,1ab; C1,2: 1,3±0,1ab; C1,9: 1,4±0,2a, em MPa.m0,5). C0,5 (31,2±6,2a) e C1,9 (34,7±7,4a) apresentaram valores de n superiores a C0,9 (20,3±3,0b) e C1,2 (17,3±1,8b). C1,2 (166,42±0,01a) apresentou o maior valor de f0 (em MPa), seguido pelo C1,9 (159,82±0,02b), C0,9 (159,59±0,02c) e C0,5 (158,40±0,02d). Não houve diferença estatística entre os valores de m (6,6 a 10,6) e 0 (170,6 a 176,4 MPa) dos compósitos. As reduções na tensão de fratura para uma probabilidade de falha de 5% após 10 anos estimadas pelo diagramas SPT foram de aproximadamente 22% para C0,5 e C1,9 e 36% para C0,9 e C1,2. Não houve diferença estatística entre as médias de RFI (155,4 a 170,7 MPa). C0,5 (93,0±18,6a) apresentou o maior LFC (em MPa), seguido pelo C1,2 (91,8±11,1ab), C1,9 (87,2±3,0b) e C0,9 (82,5±8,0c). Defeitos sub-superficiais e superficiais foram as principais origens de fratura. A trinca se propagou pela matriz polimérica ao redor das partículas (deflexão de trinca) e todas as superfícies apresentaram características de fratura frágil. Como conclusão, compósitos com partículas maiores apresentaram maior KIC, enquanto que partículas menores contribuíram para um maior LFC. Compósitos com distribuição granulométrica mais ampla, independentemente do d50, apresentaram maior resistência ao SCG. Nos demais parâmetros e propriedades avaliados (m, 0 e RFI) não houve influência do tamanho das partículas. / The aim of this study was to verify the influence of different filler sizes in the fracture toughness (KIc), subcritical crack growth (n e f0) and Weibull (m e 0) parameters, longevity estimated by the strength-probability-time (SPT) diagram and cyclic fatigue limit (CFL) of experimental resin composites. Four composites were prepared, each one containing 78 w% (59 vol%) of inorganic content, in which 67 w% were glass powder with different filler sizes (d50 = 0.5; 0.9; 1.2 e 1.9 µm) and 11 w% were pyrogenic silica. KIc data was obtained by the single-edge notched beam test and submitted to ANOVA/Tukey tests (p < 0.05). n and f0 were determined by the dynamic fatigue test (10-2 a 102 MPa/s) using a biaxial flexural device (12x1.2 mm; n=10). 20 specimens of each composite were tested at 100 MPa/s to determine Weibull parameters. SPT diagrams were constructed using the dynamic fatigue and Weibull data. For the cyclic fatigue test, a biaxial flexural device (12x1.2 mm) was used to obtain the initial flexural strength (IFS; n=14) and CFL (n=20). CFL was determined by staircase method after 105 cycles. For all tests, the specimens were stored in distilled water at 37oC for 24h. It was done the fractography of the fractured specimens that was subjected to the 10-2 e 10-1 MPa/s rates of the dynamic fatigue and to the IFS and CFL tests. There was a direct relation between d50 e KIc (C0.5: 1.2±0.1b; C0.9: 1.3±0.1ab; C1.2: 1.3±0.1ab; C1.9: 1.4±0,2a, in MPa.m0,5). C0.5 (31.2±6.2a) and C1.9 (34.7±7.4a) presented higher n values than C0.9 (20.3±3.0b) and C1.2 (17.3±1.8b). C1.2 (166.42±0.01a) showed the highest f0 value (in MPa), followed by C1.9 (159.82±0.02b), C0.9 (159.59±0.02c) and C0.5 (158.40±0.02d). There were no statistical differences among the m (6.6 to 10.6) and 0 (170.6 to 176.4 MPa) values of the composites. The reductions in fracture stress at 5% failure probability for a lifetime of 10 years estimated by the SPT diagrams were approximately 22% for C0.5 and C1.9 and 36% for C0.9 and C1.2. There were no statistical differences among the IFS means (155.4 to 170.7 MPa). C0.5 (93.0±18.6a) showed the highest CFL (in MPa), followed by C1.2 (91.8±11.1ab), C1.9 (87.2±3.0b) and C0.9 (82.5±8.0c). Near-surface and surface flaws were the main fracture origins. The crack propagated by the polymeric matrix around the fillers (crack deflection) and all the fracture surfaces showed brittle fracture features. As conclusion, composites with large fillers presented the highest KIC, while the small fillers contributed to increase the CFL. Composites with broader granulometric size distribution, regardless of d50, showed higher resistance to SCG. There was no influence of the composites filler sizes in the others parameters and properties evaluated (m, 0 and IFS).

O objetivo deste trabalho foi realizar a caracterização microestrutural e das propriedades mecânicas de diferentes cerâmicas odontológicas para entender sua influência nos parâmetros de crescimento de trinca subcrítico (SCG), n (coeficiente de susceptibilidade ao SCG) e ?f0 (parâmetro escalar), determinados por diferentes métodos. Cinco cerâmicas foram avaliadas: porcelanas VM7 (Vita) e D (d.Sign, Ivoclar), vitrocerâmicas E1 (IPS Empress, Ivoclar) e E2 (IPS Empress 2, Ivoclar) e compósito IC (In-Ceram Alumina, Vita). Espécimes em forma de disco (12 mm x 1 mm) foram confeccionados segundo as recomendações dos fabricantes. A caracterização microestrutural dos materiais foi realizada. A tenacidade à fratura (KIc) foi determinada pelo método da fratura por indentação (IF) e indentation strength. Os parâmetros de SCG foram determinados por fadiga dinâmica (com diferentes taxas de tensão constante), fadiga estática (com diferentes níveis de tensão constante) e fadiga por IF. Os resultados mostraram que todas as cerâmicas constituíram-se de fase vítrea e cristalina, com exceção da VM7, composta somente por fase amorfa. D e E1 apresentaram partículas de leucita, em frações volumétricas iguais a 0,16 e 0,29, sendo que E1 apresentou distribuição homogênea destes cristais, enquanto que D apresentou distribuição heterogênea, formando aglomerados. E2 apresentou cristais alongados de dissilicato de lítio, em fração volumétrica de 0,58, que apresentaram tendência de alinhamento com seu maior eixo alinhado perpendicularmente à direção de compactação, que variou angularmente a partir do ponto de injeção. O compósito IC apresentou-se constituído de cristais de alumina parcialmente sinterizados (65% em volume), infiltrados por vidro de baixa fusão. Independentemente do método, IC apresentou maior tenacidade (2,81 MPa.m1/2), seguido por E2 (1,81 MPa.m1/2) e E1 (0,96 MPa.m1/2) e, por último pelas porcelanas D (0,84 MPa.m1/2) e VM7 (0,67 MPa.m1/2). A deflexão de trincas foi o principal mecanismo de tenacificação dos materiais com partículas cristalinas. Para o ensaio de fadiga dinâmica, o valor de n dependeu do material estudado, ficando entre 17,2 e 31,1. Com relação ao parâmetro ?f0, foram encontrados valores entre 48 e 384 MPa. Para materiais a base de silicato de alumínio e potássio (porcelanas VM7 e D e vitrocerâmica E1), a composição química das fases vítreas dos materiais estudados parece estar relacionada com as diferenças entre os valores de n. O parâmetro ?f0 apresentou correlação com KIc e com a microestrutura, tendendo a aumentar com o aumento da fração volumétrica de cristais. No ensaio de fadiga estática, realizado apenas com D, foram obtidos valores de n e ?f0 respectivamente iguais a 31,4 e 47 MPa, sendo considerados similares aos obtidos por fadiga dinâmica. Ao se compararem os valores de n determinados nos ensaios dinâmico e por IF, nota-se que os materiais com maior volume de fase vítrea (VM7, D e E1) apresentaram valores de n próximos para os dois métodos, ao passo que materiais com maior fração de cristais (E2 e IC), apresentaram maiores discrepâncias entre os valores de n. O ensaio de fadiga por IF também permitiu a determinação dos limites de fadiga (KI0), encontrando-se valores de entre 0,48 e 2,89 MPa.m1/2. / The objective of this study was to determine the mechanical properties and the microstructure of different dental ceramics in order to understand their influence on the slow crack growth (SCG) parameters, n (crack growth exponent) and ?f0, (scaling parameter), determined by different methods. The five ceramics tested were: porcelains VM7 (Vita) and D (d.Sign, Ivoclar), glass-ceramics E1 (IPS Empress, Ivoclar) e E2 (IPS Empress 2, Ivoclar) and composite IC (In-Ceram Alumina, Vita). Disc specimens (12 mm x 1 mm) were prepared according to manufacturers\' instructions. The microstructure of the materials was carried out. The fracture toughness (KIc) was determined by means of the indentation fracture technique (IF) and indentation strength. The slow crack growth parameters were determined by dynamic fatigue test (constant stress rate), static fatigue test (constant stress) and the indentation fracture method. The results showed that all ceramic materials were composed by glassy matrix and crystalline phases, except for VM7 (vitreous porcelain). D and E1 presented leucite particles, in volume fractions of 0.16 and 0.29. For E1, the leucite crystals were homogeneously distributed in the glassy matrix, while in D, leucite formed agglomerates. E2 presented lithium dissilicate crystals (58% in volume) that presented an alignment tendency, with their major axis oriented perpendicularly to the pressing direction, which varied angularly from the injection point. IC presented alumina crystals (65% in volume) partially sintered, infiltrated by a lanthanum glass. Regardless of the method, the fracture toughness values were higher for IC (2.81 MPa.m1/2), followed by E2 (1.81 MPa.m1/2) and E1 (0.96 MPa.m1/2), and were lower for the porcelains D (0.84 MPa.m1/2) and VM7 (0.67 MPa.m1/2). Crack deflection was the main toughening mechanism observed for the ceramics containing crystalline phases. Regarding the dynamic fatigue test, the n values depended on the material, ranging from 17,2 to 31,1. With respect to ?f0, the values obtained for this parameter ranged from 48 e 384 MPa. For the ceramics based on potassium and aluminum silicate (porcelains VM7 and D and glass-ceramic E1), the chemical composition of the glass matrix seems to be related to the differences observed in the n values. The ?f0 parameter presented a positive correlation with KIc and volume fraction of crystalline particles. For the static fatigue test, used only for porcelain D, the n and ?f0 values were, respectively, 31,4 e 47 MPa, considered similar to the ones obtained by the dynamic method. When comparing the crack growth exponents determined by the dynamic and indentation fracture tests, it can be noted that n values for the ceramics with high volume fraction of glassy phase (VM7, D and E1) were similar for both methods, but for ceramics with higher crystalline content (E2 and IC), large discrepancies were observed. The static fatigue limit (KI0) was also determined for the five materials, ranging from 0.48 e 2.89 MPa.m1/2.

[pt] O grande potencial exploratório em águas profundas motiva a busca de um conhecimento tecnológico necessário para viabilizar a exploração e produção em alto mar. Paralelamente à busca de novas tecnologias de exploração e produção, a diminuição do risco de falha estrutural em unidades de exploração do tipo offshore tornou-se uma preocupação constante do setor, uma vez que falhas estruturais podem significar elevados custos decorrentes da utilização parcial de equipamentos, manutenção extemporânea, parada de produção, perdas materiais e, principalmente, perdas humanas e danos ao ecossistema. Este trabalho apresenta um estudo sobre a influência da quantidade de martensita revenida sobre a cinética de propagação da trinca de fadiga em aço estrutural do tipo grau R4, largamente utilizado em componentes estruturais para sistemas de ancoragem de unidades flutuantes do tipo offshore, sendo fabricados por meio de soldagem por centelhamento. Corpos de prova do tipo CT foram usinados a partir do material de base e do material de solda e submetidos a tratamentos térmicos de têmpera em diferentes temperaturas de austenitização, sendo submetidos a tratamentos de revenido em temperatura única. A fim de promover uma comparação com a condição microestrutural industrial, corpos de prova também foram usinados de elos de amarras após procedimentos industriais de soldagem por centelhamento, têmpera e revenido. Posteriormente, num total de cinco condições microestruturais, os corpos de prova foram ensaiados sob condições cíclicas de carregamento, objetivando a determinação das curvas de crescimento de trinca (a versus N and da/dN versus DK curves). Os resultados do desempenho em fadiga do material demonstram que a diminuição da quantidade de martensita revenida resultou no aumento da vida útil em fadiga do material, que a condição industrial está associada com a menor resistência à fadiga observada e que a vida em fadiga do material depende da posição de retirada do corpo de prova. / [en] The great explor atory potential in deep waters has led companies in the oil sector to search for the necessary technological improvement to make the offshore exploration and production feasible. In parallel to searching for new technologies, the reduction in risks of structural failure has become the sector s constant practice, considering that such failures mean high costs due to partial use of equipments, extemporary maintenance, production stops, material losses and, mainly, damage to the ecosystem and loss of lives. A study has been made concerning the influence of the fraction of tempered martensite on the kinetics of fatigue crack growth in a grade R4 structural steel, largely used for fabricating offshore mooring chains by means of flash welding. CT specimens were machined from the base material as well as the welded joints and subjected to quenching from different austenizing temperatures and tempering at a given temperature. Aiming at the comparison with the material in industrial condition, CT specimens were also machined from chain links after flash welding, quenching and tempering under industrial conditions. After that, in a total of five microstructural conditions, the specimens were cyclically loaded in order to obtain the fatigue crack growth curves (a versus N and da/dN versus DK curves). The results show that the fatigue life of the material has increased when decreasing the fraction of tempered martensite, the industrial condition is associated with the smallest fatigue resistance and the fatigue life depends on the position where the fatigue specimens were taken form the chain like.

Despite the phenomena of fracture and fatigue having been the focus of academic research for more than 150 years, it remains in effect an empirical science lacking a complete and comprehensive set of predictive solutions. In this regard, the focus of the research in this thesis is on the development of new cohesive-zone models for fracture and fatigue that are afforded an ability to capture strain-rate effects. For the case of monotonic fracture in ductile material, different combinations of material response are examined with rate effects appearing either in the bulk material or localised to the cohesive-zone or in both. The development of a new rate-dependent CZM required first an analysis of two existing methods for incorporating rate dependency, i.e.either via a temporal critical stress or a temporal critical separation. The analysis revealed unrealistic crack behaviour at high loading rates. The new rate-dependent cohesive model introduced in the thesis couples the temporal responses of critical stress and critical separation and is shown to provide a stable and realistic solution to dynamic fracture. For the case of fatigue, a new frequency-dependent cohesive-zone model (FDCZM) has been developed for the simulation of both high and low-cycle fatigue-crack growth in elasto-plastic material. The developed model provides an alternative approach that delivers the accuracy of the loading-unloading hysteresis damage model along with the computational efficiency of the equally well-established envelope load-damage model by incorporating a fast-track feature. With the fast-track procedure, a particular damage state for one loading cycle is 'frozen in' over a predefined number of cycles. Stress and strain states are subsequently updated followed by an update on the damage state in the representative loading cycle which again is 'frozen in' and applied over the same number of cycles. The process is repeated up to failure. The technique is shown to be highly efficient in terms of time and cost and is particularly effective when a large number of frozen cycles can be applied without significant loss of accuracy. To demonstrate the practical worth of the approach, the effect that the frequency has on fatigue crack growth in austenitic stainless-steel 304 is analysed. It is found that the crack growth rate (da/dN) decreases with increasing frequency up to a frequency of 5 Hz after which it levels off. The behaviour, which can be linked to martensitic phase transformation, is shown to be accurately captured by the new FDCZM.

Built on the recent successes in understanding the crystallographic mechanism for short fatigue crack (SFC) growth across a grain boundary (GB) and developing an experimental method to quantify the GB resistance against short crack growth, a microstructure-based model was developed in this study to simulate the growth behaviors of SFCs in 3-D, by taking into account both the driving force and resistance along at each point along the crack front in an alloy. It was found that the GB resistance was a Weibull function of the minimum twist angle of crack deflection at the boundary in AA2024-T3 Al alloys. In the digital microstructure used in the model, the resistance at each GB that the short crack interacted with could be calculated, as long as the orientations of grains and the crack were known. In the model, an influence function accounting for the overlapping effect of the resistance from the neighboring grain boundaries was proposed, allowing for calculation of the total resistance distribution along the crack front. In order to overcome the time consuming problem for the existing equations to derive the distribution of stress intensity factor along the crack front under cyclic loading, an analytical equation was proposed to quantify the stress intensity factor distribution along an irregular shape planar crack. By introducing two shape-dependent factors, the fractured area and the perimeter of the crack front, the newly proposed equation could readily and accurately derive the stress intensity factor distribution along the crack front that had large curvatures and singularities. Finally, a microscopic-scale Paris’ equation was proposed that took into account both the driving force, i.e., stress intensity factor range, and the total resistance to calculate the growth rate at each point along crack front. The model developed in this work was able to incorporate microstructure, such as grain size and shape, and texture into simulation of SFC growth in 3-D. It was capable of simulating all the anomalous growth behaviors of SFCs, such as the marked scatters in growth rate measurement, retardation and arrest at grain boundaries, and crack plane deflection at grain boundaries, etc. The model was used to simulate the growth behaviors of SFCs initiated from prefractured constituent particles in order to interpret the multi-site fatigue crack initiation observed in AA2024-T351 Al alloys. Three types of SFCs were observed initiating from these particles, namely, type-I non-propagating cracks; type-II cracks which were arrested soon after propagating into the matrix; and type-III propagating cracks. To quantitatively study the 3-D effects of particle geometry and micro-texture on the growth behaviors of micro-cracks in these particles, rectangular micro-notches with different dimensions were fabricated using focused ion beam in the selected grains on the T-S planes in AA2024-T351 Al alloys, to mimic the pre-fractured particles in these alloys. Knowing the notch dimensions or particle shape, grain orientation and GB geometry, the simulated crack growth behaviors were consistent with the experimental observations, and the model was able to verify that the three types of cracks evolved from these particles were mainly associated with the thickness and width of the pre-fractured particles, though the particle geometry and grain orientation could also affect the behaviors of fatigue crack initiation at the particles. When the widths of the particles were less than 15 μm, like in most high strength Al alloys, the simulated results confirmed that the crack type was only associated with the particle thickness, consistent with the experimental results in AA2024-T351 alloys with a strong rolling texture. The lives for the SFCs to reach 0.5 mm in length were quantified with the model in the AA2024 alloy, revealing that there was a bimodal distribution in the life spectrum calculated, with the longer life peak being related to larger twist angles of crack deflection at the first GB the cracks encountered and the shorter life peak being associated with small twist angles (< 5°) at the first GB. The model further demonstrated the influence of grain structure on SFC growth by considering two different grain structures with the same initial short crack, namely, a layered grain structure with only the primary GBs perpendicular to the surface and the layered grains with both primary and secondary GBs. Depending on their positions and geometry, the secondary GBs could still exert a strong retarding effect on SFC growth on surface. The model was validated by matching to the growth rate measured on surface of a SFC in an AA8090 Al-Li alloy. Good consistency was achieved between the simulated and experimentally measured growth rates when both the primary and secondary GBs were considered in the model. The model developed in this study exhibits its potential applications to optimizing the microstructure and texture in alloys to enhance their fatigue resistance against fatigue crack growth, and to satisfactory life prediction of engineering alloys.

Welding, as one of the most effective joining methods for metals, has been extensively applied in engineering usage for a long time. When cracks occur in the vicinity of weldments, weld repairs are frequently considered for crack repair to extend service life. In order to evaluate to what extent the weld repair has improved the fatigue life of a cracked welded structure, it is necessary to be able to determine the residual life of the cracked welded joint, as well as the life of the weld repaired joint. Both these assessments require that the fatigue crack growth data be available. The determination of crack propagation rates of welded and weld repaired structures is thus of paramount importance to implement a damage tolerant approach to structural life extension. However, since most studies on welded joints so far have concentrated on fatigue life evaluation, at the present time only limited information is available on crack propagation rates in welded joints, and virtually none on fatigue behaviour and crack propagation in weld repaired joints. This thesis has focused on examination of fatigue and crack propagation behaviour in as welded and weld repaired aluminium alloy 5083, a weldable marine grade alloy extensively used in construction of high speed ferries and aerospace structures. Crack growth rates were measured during constant amplitude fatigue testing on unwelded, as-welded and weld repaired specimens of 5083-H321 aluminium alloy. A 3-D finite element analysis was conducted to determine the stress intensity factors for different lengths of crack taking into account the three-dimensional nature of the weld profile. The effects of crack closure due to weld residual stresses were evaluated by taking measurements of the crack opening displacements and utilised to determine the effective stress intensity factors for each condition. Metallurgical examinations and fractography of the fracture surface were conducted using an optical microscope and SEM. It was found that crack growth rates in welded plates are of the same order of magnitude as those of parent material when effective stress intensity factors were applied. However weld repaired plates exhibit higher crack growth rates compared to those of unwelded and once-only welded plates.

This thesis deals with some aspects of crack growth in the presence of cyclic loading, i.e. fatigue. The cyclic load cases studied here are primary of non-proportional mixed mode type. Under non-proportional loading the principal stress directions rotate and, generally, the ratio between the principal stresses vary. A new criterion has been presented for prediction of incipient crack path direction after changes in load from steady Mode I to non-proportional loading. The criterion is based on FE-simulations which are used to compute the actual elasto-plastic stress state in the vicinity of the crack tip. The predictions of the criterion capture several phenomena observed in the literature, which indicates that plasticity effects have to be included in a criterion for crack path predictions under non-proportional loading. The effects of Mode II overloads on subsequent Mode I crack growth have been studied relatively little in the literature. Also, the results deviates substantially. In the present thesis, this load case has been investigated in detail, both experimentally and analytically. The results show that the Mode I crack growth rate decreases after a single Mode II load, if the R-ratio is not as high as to keep the entire Mode I load cycle above the closure level. This is based on the fact, shown in this thesis, that the reduction is caused by crack closure due to tangential displacement of crack-surface irregularities. A new loading device is presented. With this device, it is possible to apply sequential loading in Mode I and Mode II in an automated way, without having to dismount the specimens. This loading device is used to study the influence of periodic Mode II loading on Mode I crack growth. The main parameters concerning the influence of periodic Mode II loading on Mode I crack growth are; (i) the Mode I R-ratio, (ii) the Mode II magnitude and (iii) the Mode II periodicity, M (number of Mode I loads for every Mode II load). The mechanisms involved are mainly RICC (Roughness-Induced Crack Closure) and a Mode II mechanism that increases the growth rate temporary at every Mode II load. Hence, the latter becomes more significant for low M-values. The higher the Mode I R-ratio the smaller is the reduction. / QC 20101004