Nous étudions numériquement la solidification directionnelle d'un alliage binaire à base de succinonitrile. Pour cela, nous développons un code s'appuyant sur le formalisme du champ de phase adapté au cas de la croissance dans des lames minces. Les résultats numériques obtenus sont comparés qualitativement et quantitativement avec les observations expérimentales. Une bonne confirmation des lois expérimentales et de nouvelles informations sur la dynamique des microstructures sont obtenues.La direction de croissance est généralement limitée par deux axes : l'axe cristallin principal et la direction du gradient thermique. Une première partie de la thèse porte sur l'étude des effets de la désorientation de l'axe cristallin sur la direction de croissance des structures et sur leurs morphologies. Nos résultats sont directement comparés à la loi expérimentale qui donne la réponse en orientation des microstructures sur l'ensemble de leur domaine d'existence en fonction du nombre de Péclet. Nous obtenons un accord très satisfaisant entre simulation et expérience. Dans la seconde partie de la thèse, une instabilité oscillante (mode 2λ − O) est étudiée en se basant sur le diagramme de stabilité expérimental. Dans ce mode deux cellules voisines oscillent en opposition de phase en largeur et en hauteur. Nos simulations reproduisent ce mode oscillant dans des lames minces et permettent une comparaison quantitative avec les expériences. Le régime des oscillations forcées est notamment exploré pour obtenir des informations sur la réponse en fréquence du système. / We report on a numerical study of directional solidification in thin samples of succinonitrile-based dilute alloy. This thesis is based on 3D phase-field simulations. Numerical results are compared qualitatively and quantitatively with experimental observations. The comparison gives a good confirmation of the experimental laws, while providing new information on the dynamics of microstructures. Growth direction of the microstructure is constrained by two axes : the main crystal axis and the direction of the thermal gradient. Simulations allow us to test the variations of the growth direction and the microstructure stability at various misorientation angles. Our results are directly compared with the experimental law that gives the microstructure orientation response in a large domain of Péclet numbers. We obtain a good agreement, both on qualitative and quantitative grounds, between experiments and 3D simulations.In the second part of this manuscript, an oscillatory instability (2λ − O mode) is numerically studied. This mode involves oscillations of both cell width and cell tip position. This instability is reproduced in numerical simulations with the aim of allowing a fine and relevant comparison with experiments of the domain of existence and the periods of oscillation. In particular, the forced oscillation regime is explored to obtain information on the frequency response of the system.

Dans les alliages multiphasés, la cohérence des interfaces entre des phases en désaccord paramétrique génère des champs élastiques internes à longue distance et généralement anisotropes. L'interaction de ces champs affecte fortement la cinétique des transformations de phase diffusives, et influence la forme et l'arrangement spatial des précipités. Dans la microstructure des superalliages monocristallins à base de nickel, obtenue par précipitation de la phase γ’ ordonnée L12 dans la matrice CFC γ, l'élasticité conduit à la formation d'alignements quasi-périodiques des précipités γ’ cuboïdaux. La microstructure γ/ γ’ possède cependant des défauts systématiques d'alignement des précipités: des branches, des macro-dislocations et des motifs en chevrons. Nous nous intéressons à l'origine de ces défauts d'alignement. Nous conduisons des analyses de stabilité de l'arrangement périodique de précipités en interactions élastiques. Contrairement à la stabilité attendue, les calculs semi-analytiques ont révélé l'instabilité de la distribution périodique de précipités γ’ cubiques, vis-à-vis de certains modes de perturbation. Les principales instabilités sont le mode longitudinal [100] et le mode transverse [110], et leur domaine d'instabilité est analysé vis-à-vis de l'anisotropie élastique. Le développement de ces modes instables est étudié par une méthode de champ de phase classique, en simulant l'évolution de microstructures périodiques soumises à des légères perturbations initiales. Nous montrons que l'expression des instabilités d'arrangement procède essentiellement par l'évolution de la forme des précipités, et conduit à la formation de motifs qui ont pu être reliés à des microstructures expérimentales. En particulier, le mode transverse [110] conduit à la formation de motifs en chevrons. Nous étudions l'influence du taux de phase γ’ et de l'inhomogénéité du module élastique C’, et nous montrons le rôle qu'ils jouent dans la stabilisation de l'arrangement périodique. Dans des simulations réalisées dans des études antérieures, la dynamique des défauts est analysée au moyen de paramètres topologiques issus de la phénoménologie des structures hors-équilibre. Au cours d'un recuit isotherme, nous observons que les branches et les macro-dislocations migrent dans la microstructure selon des mécanismes de montée et de glissement. Nous utilisons ensuite une nouvelle formulation des modèles de champ de phase, intrinsèquement discrète, dans laquelle les interfaces sont résolues essentiellement avec un pas de grille sans friction de réseau et avec une invariance par rotation précise. Cette approche, appelée Sharp Phase Field Method (S-PFM), est implémentée sur une grille CFC, et avec une description des quatre variants de translation des précipités γ’. Nous montrons que la S-PFM permet la modélisation de microstructures à grande échelle, avec plusieurs milliers de précipités à deux et trois dimensions, et donne ainsi accès à des informations statistiques sur l'évolution de la microstructure et sur la dynamique des défauts d'alignement. Nous discutons finalement la perspective de modéliser l'évolution de la microstructure γ/γ’ à une échelle supérieure par une description de la dynamique des défauts d'alignement des précipités. / In multiphase alloys, internal elastic fields often arise as a result of a coherently adjusted misfit between the lattices of coexisting phases. Given their long-range and usually anisotropic nature, the interaction of these fields is known to significantly alter the kinetics of diffusion-controlled phase transformations, as well as influence the shapes and spatial arrangement of the misfitting precipitates. In the microstructure of single-crystal nickel-base superalloys, obtained by precipitation of the L12-ordered γ’ phase in the FCC γ matrix, elasticity leads to the formation of nearly periodic alignments of the cuboidal γ’ precipitates. However, the γ/γ’ microstructure systematically displays defects in the precipitate alignment: branches, macro-dislocations and chevron patterns. We first address the question of the origin of these alignment defects. Stability analyses of the periodic arrangement of elastically interacting precipitates are carried out. Contrary to the expected stability, the semi-analytical calculations revealed the periodic distribution of cubic γ‘ precipitates to be unstable against specific perturbation modes. The main instabilities are the [100] longitudinal mode and the [110] transverse mode, and their instability range is analyzed with respect to the elastic anisotropy. The consequences of these unstable modes are investigated using a classic phase field method, by modeling the evolution of periodic microstructures undergoing small initial perturbations. We show the expression of the instabilities mainly proceeds by the evolution of the precipitate shapes, and leads to the formation of patterns which were related to experimental microstructures. Specifically, the [110] transverse instability is responsible for the formation of chevron patterns. The effects of the volume fraction and of an inhomogeneity on the C’ shear modulus on the stability of the arrangement are studied, and we show the role they play in the partial stabilization of the periodic distribution, though the [100] longitudinal mode always remains unstable. In phase field calculations carried out in previous studies, the dynamics of alignment defects are analyzed by means of topological parameters derived from pattern formation theory. During annealing, branches and macro-dislocations were observed to migrate in the microstructure according to climbing and gliding mechanisms. We then use a new formulation of phase field models, intrinsically discrete, in which the interfaces are resolved with essentially one grid point with no pinning on the grid and an accurate rotational invariance. This approach, known as the Sharp Phase Field Method (S-PFM), is implemented on a FCC grid and accounts for the four translational variants of the γ’ precipitates. We show that the S-PFM allows for the modeling of large-scale microstructures, with several thousand precipitates both in two and three dimensions, and provides access to statistical information on the microstructure evolution and on the the dynamics of alignment defects. We finally discuss the perspective of modeling the evolution of the γ/γ’ microstructure at the macroscale by means of a description of the defect dynamics in the precipitate alignments.

La croissance dendritique est calculée en utilisant un modèle champ de phase avec adaptation automatique anisotrope et non structurées d’un maillage éléments finis. Les inconnues sont la fonction champ de phase, une température adimensionnelle et une composition adimensionnelle, tel que proposé par [KAR1998] et [RAM2004]. Une interpolation linéaire d’éléments finis est utilisée pour les trois variables, après des techniques de stabilisation de discrétisation qui assurent la convergence vers une solution correcte non-oscillante. Afin d'effectuer des calculs quantitatifs de la croissance dendritique sur un grand domaine, deux ingrédients numériques supplémentaires sont nécessaires: un maillage adaptatif anisotrope et non structuré [COU2011], [COU2014] et un calcul parallèle [DIG2001], mis à disposition de la plateforme numérique utilisée (CimLib) basée sur des développements C++. L'adaptation du maillage se trouve à réduire considérablement le nombre de degrés de liberté. Les résultats des simulations en champ de phase pour les dendrites pour une solidification d'un matériau pur et d’un alliage binaire en deux et trois dimensions sont présentés et comparés à des travaux de référence. Une discussion sur les détails de l'algorithme et le temps CPU sont présentés et une comparaison avec un modèle macroscopique sont faite. / Dendritic growth is computed using a phase-field model with automatic adaptation of an anisotropic and unstructured finite element mesh. Unknowns are the phase-field function, a dimensionless temperature and a dimensionless composition, as proposed by [KAR1998] and [RAM2004]. Linear finite element interpolation is used for all variables, after discretization stabilization techniques that ensure convergence towards a correct non-oscillating solution. In order to perform quantitative computations of dendritic growth on a large domain, two additional numerical ingredients are necessary: automatic anisotropic unstructured adaptive meshing [COU2011], [COU2014] and parallel implementations [DIG2001], both made available with the numerical platform used (CimLib) based on C++ developments. Mesh adaptation is found to greatly reduce the number of degrees of freedom. Results of phase-field simulations for dendritic solidification of a pure material and a binary alloy in two and three dimensions are shown and compared with reference work. Discussion on algorithm details and the CPU time are outlined and a comparison with a macroscopic model are made.

This work describes two phase field models for the simulation of the interface evolution between a LiCoO2 cathode (LCO) and a Li7La3Zr2O12 solid electrolyte (LLZO) in a Li-metal/LLZO/LCO battery during high temperature sintering. In these conditions atomic species tend to diffuse into the opposing material, creating an intermediate layer of mixed composition which resists the movement of lithium ions. This undesired effect prevents the resulting solid-state battery to achieve its theoretical performances and needs to be avoided. The first model is an adaptation of the work of J. M. Hu et alii [1] for a similar interface problem encountered between yttria-stabilized zirconia electrolytes (YSZ) and lanthanum-strontium-manganite cathodes (LSM) in solid oxide fuelcells (SOFC), while the second is based on the work of D. A. Cogswell [2][3] for phase separation in metal alloys, extended to include electrostatic effects due to internal charge unbalances and externally applied electric fields. Animplementation of the latter is however lacking, and the interested reader is encouraged to build one up on the theoretical framework presented in this paper. In the conclusion section it is possible to find insights on how to prevent the interfacial diffusion between LCO and LLZO with reference to experimental attempts and simulations, as well as future directions for the development of the models.

La rupture macroscopique d’un matériau intervient généralement lorsque des micro-défauts coalescent, plutôt que par la propagation catastrophique d’une seule fissure. Il est donc souhaitable d’étudier des configurations de rupture où de multiples fissures interagissent. Les paires de fissures en-passant (EP), où deux fissures parallèles croissent l’une vers l’autre, sont particulièrement intéressantes d’un point de vue applicatif. Cette configuration de rupture se retrouve aussi bien dans des situations naturelles (os, dorsales océaniques,…) qu’industrielles (génie civil, pièces métalliques,…). Malgré la diversité de tailles et de matériaux dans lesquels ces fissures existent, leurs trajectoires ont une forme typique en crochet quasi-universelle dont l’origine, résultant de l’interaction fissure-fissure répulsive puis attractive, est mal comprise. En particulier, le comportement répulsif initial semble mettre à mal la mécanique élastique linéaire de la rupture (MELR). Dans cette thèse, nous avons d’abord étudié les fissures EP dans le cadre de la MELR. L’étude de l’angle initial de déviation et la simulation de trajectoires a montré contre toute attente que la MELR permet de reproduire qualitativement la forme en crochet. Prédire précisément certaines caractéristiques, comme l’intensité de la phase répulsive, nécessite plus de finesse au niveau de la représentation du comportement matériau. Nous avons ensuite utilisé un modèle par champ de phase pour enrichir le modèle matériau. Les nouvelles trajectoires simulées étant fortement influencées par la longueur caractéristique du champ de phase, il est possible d’obtenir un modèle plus juste quantitativement. Une perspective intéressante reste de relier cette longueur à la microstructure du matériau / Macroscopic failure of a material happens generally through the coalescence of micro-defects rather than the catastrophic propagation of a single crack. It is therefore advisable to study fracture problems in which many cracks interact. The case of en-passant crack pairs (EP-cracks), two parallel and offset cracks approaching each other by propagating through their inner tips, presents a marked interest as these cracks can be found in various natural (bones, oceanic rifts,..) or industrial (civil engineering,…) situations. Despite the large variety of scales and materials in which these cracks are observed, their trajectories present a remarkably self-similar hook-shape. This shape result from the crack-crack interaction, first repulsive before becoming attractive, and its origin is poorly understood. In particular, the initial repulsive behaviour seems to question the validity of linear elastic fracture mechanics (LEFM). In this thesis, we first studied EP-cracks in the LEFM framework. The study of the initial kink angle and the simulation of crack paths showed against all expectations that LEFM is able to reproduce qualitatively the hook-shaped paths. Precise predictions of specific characteristics, such as the magnitude of repulsion, requires a more refined model of the material behaviour. We then used a phase-field model to augment the material representation. As they are strongly influenced by the characteristic length scale of the phase-field, the new simulated trajectories indicate that it is possible to develop a more quantitatively correct model. An attractive prospect is to link this characteristic length to the material microstructure

Avec le développement des techniques d’imagerie telles que la tomographie par rayons X au cours des dernières années, il est maintenant possible de prendre en compte la microstructure réelle dans les simulations des matériaux composites. Cependant, la complexité des composites tels que des fibres inclinées et brisées, les vides, exige un grand nombre des données à l’échelle microscopique pour décrire ces détails et amène ainsi des problèmes difficiles en termes de temps de calcul et de mémoire lors de l’utilisation de méthodes de simulation traditionnelles comme la méthode Eléments Finis. Ces problèmes deviennent encore plus sérieux dans la simulation de l’endommagement, comme la propagation des fissures. Par conséquent, il est nécessaire d’étudier des méthodes numériques plus efficaces pour ce genre de problèmes à grande échelle. La méthode Multigrille (MG) est une méthode qui peut être efficace parce que son coût de calcul est proportionnel au nombre d’inconnues. Dans cette thèse, un solveur de MG efficace pour ces problèmes est développé. La méthode MG est appliquée pour résoudre le problème d’élasticité statique basé sur l’équation de Lamé et aussi le problème de la propagation de fissures basé sur une méthode de champ de phase. La précision des solutions MG est validée par une solution analytique classique d’Eshelby. Ensuite, le solveur MG est développé pour étudier le processus d’homogénéisation des composites et ses solutions sont comparées avec des solutions existantes de la littérature. Après cela, le programme de calcul MG est appliqué pour simuler l’effet de bord libre dans les matériaux composites stratifiés. Une structure stratifiée réelle donnée par tomographie X est d’abord simulé. Enfin, le solveur MG est encore développé, combinant une méthode de champ de phase, pour simuler la rupture quasi-fragile. La méthode MG présente l’efficacité à la fois en temps de calcul et en mémoire pour résoudre les problèmes ci-dessus. / With the development of imaging techniques like X-Ray tomography in recent years, it is now possible to take into account the microscopic details in composite material simulations. However, the composites' complex nature such as inclined and broken fibers, voids, requires rich data to describe these details and thus brings challenging problems in terms of computational time and memory when using traditional simulation methods like the Finite Element Method. These problems become even more severe in simulating failure processes like crack propagation. Hence, it is necessary to investigate more efficient numerical methods for this kind of large scale problems. The MultiGrid (MG) method is such an efficient method, as its computational cost is proportional to the number of unknowns. In this thesis, an efficient MG solver is developed for these problems. The MG method is applied to solve the static elasticity problem based on the Lame's equation and the crack propagation problem based on a phase field method. The accuracy of the MG solutions is validated with Eshelby's classic analytic solution. Then the MG solver is developed to investigate the composite homogenization process and its solutions are compared with existing solutions in the literature. After that, the MG solver is applied to simulate the free-edge effect in laminated composites. A real laminated structure using X-Ray tomography is first simulated. At last, the MG solver is further developed, combined with a phase field method, to simulate the brittle crack propagation. The MG method demonstrates its efficiency both in time and memory dimensions for solving the above problems.

Surface energy plays a major role in many phenomena that are important in technological and industrial processes, for example in wetting, grain growth and sintering. In this thesis, such surface-energy driven processes are studied by means of the phase-field method. The phase-field method is often used to model mesoscale microstructural evolution in materials. It is a diffuse interface method, i.e., it considers the surface or phase boundary between two bulk phases to have a non-zero width with a gradual variation in physical properties such as energy density, composition and crystalline structure. Neck formation and coarsening are two important diffusion-controlled features in solid-state sintering and are studied using our multiphase phase-field method. Inclusion of Navier-Stokes equation with surface-tension forces and convective phase-field equations into the model, enables simulation of reactive wetting and liquid-phase sintering. Analysis of a spreading liquid on a surface is investigated and is shown to follow the dynamics of a known hydrodynamic theory. Analysis of important capillary phenomena with wetting and motion of two particles connected by a liquid bridge are studied in view of important parameters such as contact angles and volume ratios between the liquid and solid particles. The interaction between solute atoms and migrating grain boundaries affects the rate of recrystallization and grain growth. The phenomena is studied using a phase-field method with a concentration dependent double-well potential over the phase boundary. We will show that with a simple phase-field model it is possible to model the dynamics of grain-boundary segregation to a stationary boundary as well as solute drag on a moving boundary. Another important issue in phase-field modeling has been to develop an effective coupling of the phase-field and CALPHAD methods. Such coulping makes use of CALPHAD's thermodynamic information with Gibbs energy function in the phase-field method. With the appropriate thermodynamic and kinetic information from CALPHAD databases, the phase-field method can predict mictrostructural evolution in multicomponent multiphase alloys. A phase-field model coupled with a TQ-interface available from Thermo-Calc is developed to study spinodal decomposition in FeCr, FeCrNi and TiC-ZrC alloys. / QC 20100622

Establishing process-structure-property relationships is an important objective in the paradigm of materials design in order to reduce the time and cost needed to develop new materials. A method to link phase field (process-structure relations) and microstructure-sensitive finite element (structure-property relations) modeling is demonstrated for subsolvus polycrystalline IN100. A three-dimensional (3D) experimental dataset obtained by orientation imaging microscopy performed on serial sections is utilized to calibrate a phase field model and to calculate inputs for a finite element analysis. Simulated annealing of the dataset realized through phase field modeling results in a range of coarsened microstructures with varying grain size distributions that are each input into the finite element model. A rate dependent crystal plasticity constitutive model that captures the first order effects of grain size, precipitate size, and precipitate volume fraction on the mechanical response of IN100 at 650°C is used to simulate stress-strain behavior of the coarsened polycrystals. Model limitations and ideas for future work are discussed.

L'estudi de la formació de microstructures en processos de solidificació té importants aplicacions científiques i tecnològiques. L'aparició d'aquestes estructures determina les propietats elèctriques i mecàniques del material solidificat, i té per tant un important interès tecnològic. La majoria d'aquestes estructures tenen el seu origen en una desestabilització morfològica de la interfase sòlid-líquid que es produeix a mesura que el front avança. Per aquest motiu, l'estudi del comportament dinàmic de la interfase resulta essencial per entendre els mecanismes que intervenen en la creació d'aquests patrons. Els processos de solidificació solen descriure's mitjançant problemes de contorn mòbil. Aquestes formulacions consten d'equacions per a la difusió del calor i de massa en les fases sòlida i líquida, que s'han de resoldre imposant l'acompliment de diverses condicions de contorn mòbils a la interfase. Els problemes de contorn mòbil, malgrat contenir tots els elements que intervenen en la dinàmica i ser de molta utilitat en l'àmbit de l'enginyeria, requereixen un cost computacional que no permet simular sistemes reals en règims interfacials complexos. Els mètodes de camp de fase (phase-field methods), van aparèixer a principis dels anys vuitanta com una eina computacional que permetia l'estudi de fenòmens interfacials de caire general. Aquests mètodes descriuen la forma de la interfase mitjançant un camp continu que pren valors diferents i constants en les dues fases. La dinàmica d'aquest camp és llavors acoblada al camp de difusió de calor o massa que determina l'avanç del front de solidificació. Un dels avantatges d'aquests mètodes és que la seva simulació no requereix d'algorismes de seguiment de la interfase (front tracking algorithsms). És ben conegut que les característiques principals de les microestructures en solidificació, es determinen durant els transitoris inicials en els que els corrents de massa i calor s'adapten a la evolució dinàmica del front. Un dels objectius en aquesta tesi és el de fer servir mètodes de camp de fase per descriure de forma quantitativa aquests transitoris. Per comprovar la validesa del nostre procediment, es realitza una comparació quantitativa entre els resultats numèrics obtinguts i diferents prediccions analítiques derivades del problema de contorn mòbil. Per un altra banda, la desestabilització del front es veu afectada per la presència de fluctuacions al sistema. Aquestes pertorbacions microscòpiques poden tenir el seu origen a les fluctuacions termodinàmiques internes, o bé ser conseqüència de imperfeccions experimentals que actuen com a font externa de soroll. El segon objectiu d'aquesta tesi és la introducció de fluctuacions en mètodes de camp de fase, de forma que es pugui estudiar l'amplificació dinàmica de les pertorbacions microscòpiques que acaben donant lloc a estructures macroscòpiques. Per finalitzar, analitzem el problema de la selecció en solidificació direccional. Estudiem els règims lineal i no-lineal, tot determinant les condicions, el moment i la forma en que apareixen les estructures dendrítiques i cel·lulars. / Crystal growth is a non-equilibrium process which involves physical mechanisms at very different scales. When a solidification front advances, mass and heat diffusion processes are combined with interfacial phenomena like capillarity or kinetic attachment. A complex interplay between these mechanisms gives rise to complex interfacial structures like snowflakes or cellular patterns. The formation of microstructures in solidification has both a scientific and a technological interest. On one hand, the study of the different interfacial structures constitutes a fundamental problem in the field of non-equilibrium pattern-forming systems. On the other side, from a technological point of view, the presence of microstructures determines the final mechanical and electrical properties of the processed material. Directional solidification is a controlled solidification technique which reproduces the conditions occurring in some important metallurgical processes like material casting or zone melting refining procedures. In a directional solidification experiment, the alloy sample is pulled at a constant velocity towards the cold region of an externally-imposed temperature gradient. Depending on the growth conditions, a morphological destabilization of the solid-liquid interface occurs during early transient stages. These initial transients are associated to a solute redistribution process due to the adaptation of the concentration field to the forced motion of the sample. The main objective of this thesis is to study the dynamical evolution of the morphological deformations of the front from these initial transients to the final stages where the properties of the interfacial pattern are determined. An important point in this process is that the internal fluctuations of the system play the role of an initiation mechanism for the morphological deformations of the front. During the initial transients, some of these microscopic perturbations are amplified by several orders of magnitude, and a range of wavelengths becomes morphologically unstable. The interfacial deformations of the front can be then characterized by means of power spectrum techniques. In order to study the dynamical evolution of the solidification front in directional solidification, we have used both theoretical and computational approaches: The main computational technique used in this thesis is the phase-field approach, which is a powerful method to simulate complex interfacial phenomena. The model equations describe the evolution of a continuous field , which takes different constant values at the solid and liquid bulks of the system. This field is then coupled with equations for the mass diffusion, and allows performing numerical studies without simulating the standard Stefan-like moving boundary problem. The phase-field method provides a diffuse interface description in which the transition from solid to liquid happens in a region of a certain thickness. The interface thickness introduces a new length in the simulations which must be taken into account to recover quantitative results. One major point in this thesis concerns with the introduction of fluctuations in phase-field methods. In the particular case of variational phase-field formulations -in which the model equations can be derived from a single free energy functional for the whole system-, the introduction of fluctuations can be done by applying the Fluctuation-Dissipation theorem. Variational formulations, however, although their appealing structure, present a poor computational efficiency and cannot be used to obtain quantitative results. To this extent, we have derived a general approach which does not relay in the Fluctuation-Dissipation assumption and permits to introduce fluctuations in both variational and non-variational phase-field formulations. Well-established analytical techniques like boundary integral methods for the transient front position and linear stability analysis of the interface during the transient have been used as theoretical predictions for the computational results. The dynamical evolution of the solidification front can be divided in two stages: A linear regime where the initial noise is amplified, and a non-linear coarsening process where the final properties of the interfacial pattern are selected. We have studied these different stages of the solidification process by using the phase-field approach, and good agreement is obtained when comparing with well-established theoretical and experimental predictions.

Multiphase flow is a familiar phenomenon from daily life and occupies an important role in physics, engineering, and medicine. The understanding of multiphase flows relies largely on the theory of interfaces, which is not well understood in many cases. To date, the Navier-Stokes-Korteweg equations and the Cahn-Hilliard equation have represented two major branches of phase-field modeling. The Navier-Stokes-Korteweg equations describe a single component fluid material with multiple states of matter, e.g., water and water vapor; the Cahn-Hilliard type models describe multi-component materials with immiscible interfaces, e.g., air and water. In this dissertation, a unified multiphase fluid modeling framework is developed based on rigorous mathematical and thermodynamic principles. This framework does not assume any ad hoc modeling procedures and is capable of formulating meaningful new models with an arbitrary number of different types of interfaces. In addition to the modeling, novel numerical technologies are developed in this dissertation focusing on the Navier-Stokes-Korteweg equations. First, the notion of entropy variables is properly generalized to the functional setting, which results in an entropy-dissipative semi-discrete formulation. Second, a family of quadrature rules is developed and applied to generate fully discrete schemes. The resulting schemes are featured with two main properties: they are provably dissipative in entropy and second-order accurate in time. In the presence of complex geometries and high-order differential terms, isogeometric analysis is invoked to provide accurate representations of computational geometries and robust numerical tools. A novel periodic transformation operator technology is also developed within the isogeometric context. It significantly simplifies the procedure of the strong imposition of periodic boundary conditions. These attributes make the proposed technologies an ideal candidate for credible numerical simulation of multiphase flows. A general-purpose parallel computing software, named PERIGEE, is developed in this work to provide an implementation framework for the above numerical methods. A comprehensive set of numerical examples has been studied to corroborate the aforementioned theories. Additionally, a variety of application examples have been investigated, culminating with the boiling simulation. Importantly, the boiling model overcomes several challenges for traditional boiling models, owing to its thermodynamically consistent nature. The numerical results indicate the promising potential of the proposed methodology for a wide range of multiphase flow problems. / text