The objective of this work is to develop an evolution equation for the ductile growth of a spherical void in a highly strain rate and temperature dependent material. The material considered in this work is stainless steel 304L at 982 °C. The material is characterized by a physically-based internal state variable model derived within consistent kinematics and thermodynamics — Evolving Microstructure Model for Inelasticity. Through this formulation, the degradation of the elastic moduli due to damage has been naturally acquired. An elastoviscoplasticity user material subroutine has also been developed and implemented into a commercially available finite element software ABAQUS. The subroutine utilizes a return mapping algorithm, where a purely elastic trial state (elastic predictor) is followed by a plastic corrector phase (return mapping). A conditionally stable fully-implicit scheme, derived from the backward Euler integration method, has been employed to calculate the values of the internal state variables in the elastoviscoplasticity integration routine. A repeating unit cell problem is set up by introducing a spherical void inside a matrix material that simulates a periodic array of voids in a component. Using finite element analysis, a database is generated by recording the responses of the unit cell under various combinations of loading conditions, porosity, and state variables. Functional forms of the void growth equations are constructed by utilizing normalization techniques to collapse all the data into master curves. The evolution equations are converted to a form consistent with the continuum damage variable in the complete thermal-elastic-plastic-damage version of the physically-based internal state variable model.

The ductile fracture process, which consists of the nucleation, growth and coalescence of microvoids, was extensively studied for materials deforming homogeneously. For materials with a non-homogeneous deformation behavior, such as those having hexagonal closed packed (HCP) crystal structure, experimental and numerical data is lacking. Therefore, the fracture properties of materials with such HCP structure, like titanium (used in aerospace and biomedical applications), zirconium (nuclear industry) and magnesium (manufacturing industry) are not well understood. The main research objective of this Ph.D. thesis is to better understand the mechanisms governing fracture in commercially pure (CP) titanium. In particular, the effect of grain orientation on void growth is investigated. The fracture process of CP titanium was visualized in model materials containing artificial holes. These model materials were fabricated using a femtosecond laser coupled with a diffusion bonding technique to obtain voids in the interior of titanium samples. Diffusion bonding was carried out either above or below the phase transformation temperature resulting in different microstructures. Changes in void dimensions during in-situ straining were recorded in three dimensions using x-ray computed tomography. Void growth obtained experimentally was compared with the Rice and Tracey model which predicted well the average void growth. However, a large scatter in void growth was observed experimentally and was explained in terms of differences in grain orientation which was confirmed by crystal plasticity simulations. It was also shown that grain orientation has a stronger effect on void growth than intervoid spacing and material strength. Intervoid spacing, however, appears to control whether the intervoid ligament failure is ductile or brittle. While this study showed a good agreement between experiments and simulations on average, there is no direct void growth comparison for particular grain orientations. In a follow-up study, an experimental approach was developed to directly relate the growth of a void to its underlying grain orientation. This is achieved by first annealing CP titanium samples below the α-β phase transformation temperature, then performing electron backscatter diffraction iii (EBSD) and finally diffusion bonding the samples together. Samples were then tested in x-ray tomography. This study showed the importance of the local state of strain on void growth. Crystal plasticity simulations that take into account the particular grain orientation and the local state of strain were found to predict well experimental void growth. Crystal plasticity simulations confirmed that the orientation of the voidcontaining grain is more important than the orientation of surrounding grains and more important than the volume fraction of voids, in order to determine void growth. This thesis on the growth and coalescence of voids is important to validate and improve the predictions of ductile fracture models and to design new materials with improved fracture properties.

<p>Void growth and coalescence in a single layer model material with holes were visualized by the environmental electron scanning microscope coupled with in situ tensile test. Single sheet model materials were manufactured with a line of laser drilled holes through thickness. In order to investigate the effect of shear localization, the line of holes were oriented with the misorientation angle <em>θ </em>= 0°, 15°, 30°, and 45°. The α-brass samples were studied to introduce the work hardening effect in comparison with the pure copper samples.</p> <p>By taking images at intervals with small strain increments, the void growth behaviors were visualized during the interrupted tensile testing. The void coalescence (defined consistent with Hosokawa et al (2011), as the point at which the voids stopped shrinking laterally) was successfully captured for the first time in the two dimensional studies. The evolutions of void shape change and void rotation during deformation were also studied quantitatively. The results showed that the higher work hardening behaviors can suppress the void coalescence. It also showed that the effect of local volume fraction dominated the coalescence event rather than the void spacing and shear localization. A comparison of the classic models with the experimental results were also made.</p> / Master of Science in Materials Science and Engineering (MSMSE)

The industrial application of high strength steels in structural components has increased the demand on understanding the ductile failure behavior of this type of materials. In practical situations the loading experienced on components made out of these materials can be very complex, which may affect the failure behavior. The objective of this work is to study the effect of stress state on ductile failure and the mechanisms leading to rupture in high strength steels. The stress state is characterized by the stress triaxiality T and the Lode parameter L, which is a deviatoric stress state parameter that discriminates between axisymmetric or shear dominated stress states. For this purpose experiments on two different specimen configurations are performed; a double notched tube (DNT) specimen tested in combined tension and shear and a round notched bar (RNB) specimen tested in uniaxial tension. The two specimens give rise to different stress states at failure in terms of T and L. The failure loci for the DNT specimen show an abrupt change in ductility, indicating a transition between the rupture mechanisms necking of intervoid ligaments and shearing of intervoid ligaments. A clear difference in ductility between the two specimen configurations is also observed, which is closely associated with the difference in stress state at failure. A micromechanical model is developed, which assumes that ductile material failure occurs when the deformation becomes highly non-linear and localizes into a band. The model, which is applied to analyze the experiments, consists of a band with a square array of equally sized cells, with a spherical void located in the center of each cell. The model, extended with a shear criterion, captures the experimental trend rather well. The model also shows that the effect of the deviatoric stress state (L) on void growth, void shape evolution and coalescence is significant, especially at low levels of T and shear dominated stress state. / QC 20100621

L'endommagement des matériaux ductiles est un processus impliquant trois étapes : la nucléation, la croissance et la coalescence de vides. La phase de croissance des vides a été largement étudiée dans la littérature. Il a été montré que, durant cette étape, la forme des vides joue un rôle fondamental sur le comportement macroscopique du matériau. Dans le cas de sollicitations dynamiques, les effets micro inertiels, qui résultent des accélérations subies par la matrice au voisinage du vide, influent eux aussi fortement sur la croissance des vides. Cependant, les travaux intégrant simultanément ces deux contributions (effets inertiels et forme) sont très rares. L'objectif de ce travail est de proposer un modèle de comportement pour les matériaux poreux qui prend en compte la forme des vides et les effets micro inertiels. Dans une première partie, un volume élémentaire représentatif défini par deux ellipsoïdes allongés confocaux est utilisé pour représenter le matériau poreux. La matrice est rigide viscoplastique. En se basant sur les travaux de Molinari et Mercier (2001), la contrainte macroscopique se décompose en une partie statique et une partie dynamique. La contrainte statique est décrite par le modèle de Gologanu et al. (1997). La contrainte dynamique est obtenue en adoptant le champ de vitesse de Gologanu et al. (1993). Avec cette modélisation, il est montré que la contrainte dynamique est liée de façon quadratique au tenseur des vitesses des déformations et de façon linéaire à sa dérivée par rapport au temps. Le modèle fait l'objet d'une validation sur la base de comparaisons avec des résultats de calculs par éléments finis. Différentes forme de vides et valeurs de la porosité ont été considérées. Dans une seconde partie, le cas de matériaux contenant des vides aplatis est abordé ; le volume élémentaire représentatif est défini par deux ellipsoïdes confocaux aplatis. La contrainte statique est toujours décrite par le modèle de Gologanu et al. (1997). La contrainte dynamique est obtenue en adoptant le champ de vitesse de Gologanu et al. (1994). La procédure de validation est identique à celle mise en œuvre dans le cas des vides allongés. Une bonne adéquation entre les résultats du modèle et les résultats de calculs par éléments finis est retrouvée. L'utilisation des surfaces d'écoulement permet de mettre en lumière les effets de la forme des vides sur le comportement du matériau poreux sous chargement dynamique. En fonction du chargement appliqué, certaines géométries de vide favorisent la déformation du matériau. Le cas particulier du vide sphérique est étudié comme limite des deux modèles. La continuité des deux modèles est démontrée. L'évolution de la porosité et de la forme des vides dans un matériau poreux sous chargement dynamique est analysée. Des comparaisons avec des résultats de simulations par éléments finis sont proposées. L'influence de la triaxialité et de la vitesse du chargement sur le comportement dynamique du matériau poreux est étudiée, ainsi que celle de la forme initiale du vide. Au final, il est démontré que le modèle développé dans cette thèse permet de retrouver les tendances fournies par les calculs éléments finis / The ductile fracture mechanism involves three stages: void nucleation, void growth and void coalescence. Under dynamic loading conditions, void growth is strongly affected by microinertia effects resulting from the local acceleration of the matrix material in the vicinity of the void. Several works devoted to quasi-static conditions also show that void shape has a strong impact on the behavior of porous ductile materials. However, there exist only few works considering the combined effect of these two contributions. In the present work, we propose an original, multi-scale constitutive model of porous materials, taking into account void shape and micro-inertia effects. In a first step, a representative volume element defined by two confocal prolate spheroids is used to represent the porous material. The matrix behavior is assumed to be rigid-viscoplastic. Based on the work of Molinari and Mercier (2001), the macroscopic stress is the sum of a static and a dynamic part. The static contribution is described by the Gologanu et al. model (1997). The dynamic stress is derived by choosing the trial velocity field proposed by Gologanu et al. (1993). With the present modeling, a link is established between the macroscopic dynamic stress, on the one hand and, the macroscopic strain rate tensor and its time derivative on the other hand. To validate the proposed model, finite element computations have been performed for different void geometries and void volume fractions. The influence of micro-inertia on the macroscopic flow surface is analyzed and a good agreement between modeling and simulations is observed. In a second step, a representative volume element defined by two confocal oblate spheroids is used to represent the porous material. For this configuration, the static contribution is also described by using the Gologanu et al. model (1997), while the derivation of the dynamic stress is based on the trial velocity field proposed by Gologanu et al. (1994). As for the prolate case, a good agreement is retrieved between model predictions and results of finite element computations. The spherical void configuration is investigated as the limit case for the oblate and prolate models. The continuity between the two models is established. Finally, the proposed models are combined to investigate the porosity and void shape evolutions in a porous solid under dynamic loadings. A parametric study has been performed by varying the stress triaxiality, the initial void shape and the loading rate. Significant void shape variations are observed for low triaxiality loadings. With the present modeling, the void can evolve from prolate to oblate shapes (and the reverse). Model predictions are compared to finite element computations

any experiments demonstrate that isotropic ductile materials used in engineering applications develop anisotropic damage and shows significant variation in elongation to failure. This anisotropic damage is manifest by material microstructural heterogeneities and morphological changes during deformation. The variation in elongation to the failure could be attributed to the uncertainties in the material microstructure and loading conditions. To study this deformation induced anisotropy arising from the initial material heterogeneities, we first performed uncertainty analysis using current form on an internal state variable plasticity and isotropic damage model (Bammann, 1984; Horstemeyer, 2001) to quantify the effect due to variations in material microstructure and loading conditions on elongation to failure. We extend the current isotropic damage form of theory into an anisotropic damage form for ductile material in which material heterogeneities are introduced based on damage distribution functions converted into a damage tensor of second rank. The outcome of this research is a physically motivated, uncertainty-based, anisotropic damage constitutive model that links microstructural features to mechanical properties. This was accomplished by pursuing three sub goals: (1) develop and quantify uncertainty related to material heterogeneities, (2) develop a methodology related to a higher order tensorial rank of damage for void nucleation and void growth, and (3) integrate thermodynamically constrained damage with a rate dependent plasticity constitutive material model. Later, we also proposed a new ISV theory that physically and strongly couples deformation due to damage-related internal defects to metal plasticity.

No description available.

Dans le cadre du développement de nuances d’aciers toujours plus performantes en termes de résistance à l’effort et à l’endommagement, les aciers Dual-Phase (DP) présentent un bon compromis résistance/ductilité. Cependant, il est nécessaire de disposer de meilleures connaissances concernant les mécanismes menant à la rupture de tels aciers. Les mécanismes d’endommagement ont ainsi été étudiés dans cette thèse à l’aide de la tomographie aux rayons X. Des essais de traction in-situ ont été réalisés sur plusieurs nuances d’aciers DP, un acier ferritique et un acier martensitique afin de caractériser chaque étape de l’endommagement ductile. Des observations qualitatives et des données quantitatives concernant la germination de l’endommagement, la croissance des cavités et la coalescence ont été recueillies lors de ces essais. Ces données quantitatives ont ensuite été utilisées pour le développement et/ou la validation de modèles d’endommagement. Une prédiction de la cinétique de germination a ainsi été proposée et la version du modèle de croissance de cavités de Rice et Tracey corrigée par Huang et prenant mieux en compte l’effet de la triaxialité a été validée expérimentalement. L’étape de coalescence des cavités menant à la rupture des matériaux a pour la première fois été caractérisée de façon quantitative dans un matériau industriel et des critères de coalescence ont été appliqués localement sur les couples de cavités présentes dans le matériau. L’utilisation de ces modèles analytiques a permis une meilleure compréhension des propriétés agissant sur les phénomènes mis en jeu. L’effet de la part cinématique de l’écrouissage sur la germination et la croissance de l’endommagement a notamment été souligné et validé par des essais de chargements complexes. / As part of the current context of requiring ever more efficient grades of steels in terms of resistance to stress and to damage, the Dual-Phase steels (DP) present an acceptable strength/ductility compromise. It is nevertheless necessary to have a better understanding of the mechanisms leading to the fracture of such steels. Damage mechanisms were studied in this PhD using X-ray tomography. In-situ tensile tests were carried out on several grades of DP steel, a ferritic steel and a martensitic steel in order to characterize each step of ductile damage. Qualitative observations and quantitative data on the nucleation of damage, the void growth and the coalescence of cavities were collected during these tests. This quantitative data was then used for the development and/or the validation of damage models. A prediction of the kinetic of nucleation was proposed and the Huang’s correction of the void growth model of Rice and Tracey accounting for the triaxiality was experimentally validated. For the first time, the step of void coalescence leading to fracture of materials was quantitatively characterized in an industrial material and coalescence criteria were locally applied on couples of neighboring cavities present in the studied specimen. The use of analytical models enabled a better understanding of the properties influencing the studied damage phenomena. The effect of the kinematic part of the strain hardening on void nucleation and void growth was notably emphasized and validated by performing complex loading tests.

The failure mechanism in ductile materials involves void nucleation, their growth and subsequent coalescence to form the fracture surface. The voids are generated due to fracture or debonding of second phase particles or at slip band intersections. The triaxial stress field prevailing around a crack tip and in the necking region strongly influences the growth of these voids. In the initial stages of deformation, these microscale voids are often sufficiently small so that they exist entirely within a single grain of a polycrystalline material. Further, single crystals are used in high technology applications like turbine blades. This motivates the need to study void growth in a single crystal while investigating ductile fracture. Thus, the objectives of this work are to analyze the interaction between a notch tip and void as well as the growth and coalescence of a periodic array of voids under different states of stress in ductile FCC single crystals. First, the growth of a cylindrical void ahead of a notch tip in ductile FCC single crystals is studied. To this end, 2D plane strain finite element simulations are carried out under mode I, small scale yielding conditions, neglecting elastic anisotropy. In most of these computations, the orientation of the FCC single crystal is chosen so that notch lies in the (010) plane, with notch front along the [101] direction and potential crack growth along [101]. This orientation has been frequently observed in experimental studies on fracture of FCC single crystals. Three equivalent slip systems are considered which are deduced by combining three pairs of 3D conjugate slip systems producing only in-plane deformation. Attention is focused on the effects of crystal hardening, ratio of void diameter to spacing from the notch on plastic flow localization in the ligament connecting the notch and the void as well as their growth. The results show strong interaction between slip shear bands emanating from the notch and angular sectors of single slip forming around the void leading to intense plastic strain development in the ligament. However, the ductile fracture processes are retarded by increase in hardening of the single crystal and decrease in ratio of void diameter to spacing from the notch. In order to examine the effect of crystal orientation, computations are performed with an orientation wherein the three effective slip systems are rotated about the normal to the plane of deformation. A strong influence of crystal orientation on near-tip void growth and plastic slip band development is observed. Further, in order to study the synergistic, cooperative growth of multiple voids ahead of the notchtip, an analysis is performed by considering a series of voids located ahead of the tip. It is found that enhanced void growth occurs at higher load levels as compared to the single void model. Next, the growth and coalescence of a periodic array of cylindrical voids in a FCC single crystal is analyzed under different stress states by employing a 2D plane strain, unit cell approach. The orientation of the crystal studied here considers [101] and [010] crystal directions along the minor and major principal stress directions, respectively. Three equivalent slip systems, similar to those in the notch and void simulations are taken into account. Fringe contours of plastic slip and evolution of macroscopic hydrostatic stress and void volume fraction are examined. A criterion for unstable void growth which leads to onset of void coalescence is established. The effects of various stress triaxialities, initial void volume fraction and hardening on void growth and coalescence is assessed. It is observed that plastic slip activity around the void intensifies with increase in stress triaxiality. The macroscopic hydrostatic stress increases with deformation, reaches a peak value and subsequently decreases rapidly. An increase in stress triaxiality enhances the macroscopic hydrostatic stress sustained by the unit cell and promotes void coalescence. The stress triaxiality also has a profound effect on the shape of the void profile. The values of critical void volume fraction and critical strain, which mark onset of void coalescence, decrease within crease in stress triaxiality. However, the onset of void coalescence is delayed by increase in hardening and decrease initial void volume fraction.

An understanding of crack tip fields and fracture mechanisms in single crystals can help in developing better polycrystalline alloys and manufacturing processes. To this end, the effects of loading rate, material inertia and strain rate sensitivity on crack tip fields and their influence on fracture mechanisms in FCC single crystals are examined in this work by performing finite element analysis. It is shown that, in the absence of inertial effects, high loading rates elevate the stresses ahead of a crack tip and decrease the plastic strains in rate dependent single crystals. Also, it is found that the quasi-static near-tip stress field can be adequately characterized by the energy release rate J and a constraint parameter Q. Similar two-parameter characterization is possible even under dynamic loading. It is observed that if a suitable reference solution is used, the role of inertia manifests as a loss of constraint with increasing loading rate irrespective of strain rate sensitivity and lattice orientation. Thus, at very high loading rates, inertial effects oppose the role of rate sensitivity and cause a decrease in stresses near the tip. The relative influence of these two factors depends on rate sensitivity index. For a mildly rate dependent single crystal, the predicted cleavage fracture toughness remains constant up to a certain loading rate and thereafter increases sharply. On the other hand, for a strongly rate dependent single crystal, fracture toughness drops initially up to a certain loading rate beyond which it increases marginally. The loss of crack tip constraint is found to retard the ductile fracture mechanisms of void growth and coalescence. However, this is dependent on lattice orientation. In-situ experimental observation of void growth near a notch tip also shows strong orientation dependence. In addition, 3D finite element results indicate though-thickness dependence of equivalent plastic slip and hydrostatic stress leading to variations in void growth along the thickness direction of the specimens. The predicted load-displacement curves, lattice rotation, slip traces and void growth using finite element analysis are found to be in good agreement with the experimental observations. Thus, the present study has provided an understanding of the role of several factors such as constraint level, rate sensitivity, material inertia, lattice orientation and 3D effects on the mechanics of fracture of ductile single crystals.