This thesis concerns the mathematical modelling of phase transformations with a special emphasis on martensitic phase transformations and their application to the modelling of steels. In Chapter 1, we develop a framework that determines the optimal transformation strain between any two Bravais lattices and use it to give a rigorous proof of a conjecture by E.C. Bain in 1924 on the optimality of the so-called Bain strain. In Chapter 2, we review the Ball-James model and related concepts. We present some simplification of existing results. In Chapter 3, we pose a conjecture for the explicit form of the quasiconvex hull of the three tetragonal wells, known as the three-well problem. We present a new approach to finding inner and outer bounds. In Chapter 4, we focus on highly compatible, so called self-accommodating, martensitic structures and present new results on their fine properties such as estimates on their minimum complexity and bounds on the relative proportion of each martensitic variant in them. In Chapter 5, we investigate the contrary situation when self-accommodating microstructures do not exist. We determine, whether in this situation, it is still energetically favourable to nucleate martensite within austenite. By constructing different types of inclusions, we find that the optimal shape of an inclusion is flat and thin which is in agreement with experimental observation. In Chapter 6, we introduce a mechanism that identifies transformation strains with orientation relationships. This mechanism allows us to develop a simpler, strain-based approach to phase transformation models in steels. One novelty of this approach is the derivation of an explicit dependence of the orientation relationships on the ratio of tetragonality of the product phase. In Chapter 7, we establish a correspondence between common phenomenological models for steels and the Ball-James model. This correspondence is then used to develop a new theory for the (5 5 7) lath transformation in low-carbon steels. Compared to existing theories, this new approach requires a significantly smaller number of input parameters. Furthermore, it predicts a microstructure morphology which differs from what is conventionally believed.

Le contexte de cette thèse est la théorie de l'élasticité non linéaire, appelée également "élasticité finie". On y présente des résultats concernant la propagation d'ondes d'amplitude finie dans des matériaux élastiques non linéaires soumis à une grande déformation statique homogène. Bien que les matériaux considérés soient isotropes, lors de la propagation d'ondes un comportement anisotrope dû à la déformation statique se manifeste. <p><p>Après un rappel des équations de base de l'élasticité non linéaire (Chapitre 1), on considère tout d'abord la classe générale des matériaux incompressibles. Pour ces matériaux, on montre que la propagation d'ondes transversales polarisées linéairement est possible pour des choix appropriés des directions de polarisation et de propagation. De plus, on propose des généralisations des modèles classiques de "Mooney-Rivlin" et "néo-Hookéen" qui conduisent à de nouvelles solutions. Bien que le contexte soit tri-dimensionnel, il s'avère que toutes ces ondes sont régies par des équations d'ondes scalaires non linéaires uni-dimensionelles. Dans le cas de solutions du type ondes simples, on met en évidence une propriété remarquable du flux et de la densité d'énergie. <p><p>Dans les Chapitres 3 et 4, on se limite à un modèle particulier de matériaux compressibles appelé "modèle restreint de Blatz-Ko", qui est une version compressible du modèle néo-Hookéen. <p><p>En milieu infini (Chapitre 3), on montre que des ondes transversales polarisées linéairement, faisant intervenir deux variables spatiales, peuvent se propager. Bien que la théorie soit non linéaire, le champ de déplacement de ces ondes est régi par une version anisotrope de l'équation d'onde bi-dimensionnelle classique. En particulier, on présente des solutions à symétrie "cylindrique elliptique" analogues aux ondes cylindriques. Comme cas particulier, on obtient aussi des ondes planes inhomogènes atténuées à la fois dans l'espace et dans le temps. De plus, on montre que diverses superpositions appropriées de solutions sont possibles. Dans chaque cas, on étudie les propriétés du flux et de la densité d'énergie. En particulier, dans le cas de superpositions il s'avère que des termes d'interactions interviennent dans les expressions de la densité et du flux d'énergie. <p><p>Finalement (Chapitre 4), on présente une solution exacte qui constitue une généralisation non linéaire de l'onde de Love classique. On considère ici un espace semi-infini, appelé "substrat" recouvert par une couche. Le substrat et la couche sont constitués de deux matériaux restreints de Blatz-Ko pré-déformés. L'onde non linéaire de Love est constituée d'un mouvement non atténué dans la couche et d'une onde plane inhomogène dans le substrat, choisies de manière à satisfaire aux conditions aux limites. La relation de dispersion qui en résulte est analysée en détail. On présente de plus des propriétés générales du flux et de la densité d'énergie dans le substrat et dans la couche. <p><p><p>The context of this thesis is the non linear elasticity theory, also called "finite elasticity".<p>Results are obtained for finite-amplitude waves in non linear elastic materials which are first subjected to a large homogeneous static deformation. Although the materials are assumed to be isotropic, anisotropic behaviour for wave propagation is induced by the static deformation. <p><p>After recalling the basic equations of the non linear elasticity theory (Chapter 1), we first consider general incompressible materials. For such materials, linearly polarized transverse plane waves solutions are obtained for adequate choices of the polarization and propagation directions (Chapter 2). Also, extensions of the classical Mooney-Rivlin and neo-Hookean models are introduced, for which more solutions are obtained. Although we use the full three dimensional elasticity theory, it turns out that all these waves are governed by scalar one-dimensional non linear wave equations. In the case of simple wave solutions of these equations, a remarkable property of the energy flux and energy density is exhibited.<p><p>In Chapter 3 and 4, a special model of compressible material is considered: the special Blatz-Ko model, which is a compressible counterpart of the incompressible neo-Hookean model. <p><p>In unbounded media (Chapter 3), linearly polarized two-dimensional transverse waves are obtained. Although the theory is non linear, the displacement field of these waves is governed by a linear equation which may be seen as an anisotropic version of the classical two-dimensional wave equation. In particular, solutions analogous to cylindrical waves, but with an "elliptic cylindrical symmetry" are presented. Special solutions representing "damped inhomogeneous plane waves" are also derived: such waves are attenuated both in space and time. Moreover, various appropriate superpositions of solutions are shown to be possible. In each case, the properties of the energy density and the energy flux are investigated. In particular, in the case of superpositions, it is seen that interaction terms enter the expressions for the energy density and the energy flux. <p><p>Finally (Chapter 4), an exact finite-amplitude Love wave solution is presented. Here, an half-space, called "substrate", is assumed to be covered by a layer, both made of different prestrained special Blatz-Ko materials. The Love surface wave solution consists of an unattenuated wave motion in the layer and an inhomogeneous plane wave in the substrate, which are combined to satisfy the exact boundary conditions. A dispersion relation is obtained and analysed. General properties of the energy flux and the energy density in the substrate and the layer are exhibited. <p><p><p><p><p> / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Die Least-Squares Finite-Elemente-Methoden (LSFEMn) basieren auf der Minimierung des Least-Squares-Funktionals, das aus quadrierten Normen der Residuen eines Systems von partiellen Differentialgleichungen erster Ordnung besteht. Dieses Funktional liefert einen a posteriori Fehlerschätzer und ermöglicht die adaptive Verfeinerung des zugrundeliegenden Netzes. Aus zwei Gründen versagen die gängigen Methoden zum Beweis optimaler Konvergenzraten, wie sie in Carstensen, Feischl, Page und Praetorius (Comp. Math. Appl., 67(6), 2014) zusammengefasst werden. Erstens scheinen fehlende Vorfaktoren proportional zur Netzweite den Beweis einer schrittweisen Reduktion der Least-Squares-Schätzerterme zu verhindern. Zweitens kontrolliert das Least-Squares-Funktional den Fehler der Fluss- beziehungsweise Spannungsvariablen in der H(div)-Norm, wodurch ein Datenapproximationsfehler der rechten Seite f auftritt. Diese Schwierigkeiten führten zu einem zweifachen Paradigmenwechsel in der Konvergenzanalyse adaptiver LSFEMn in Carstensen und Park (SIAM J. Numer. Anal., 53(1), 2015) für das 2D-Poisson-Modellproblem mit Diskretisierung niedrigster Ordnung und homogenen Dirichlet-Randdaten. Ein neuartiger expliziter residuenbasierter Fehlerschätzer ermöglicht den Beweis der Reduktionseigenschaft. Durch separiertes Markieren im adaptiven Algorithmus wird zudem der Datenapproximationsfehler reduziert. Die vorliegende Arbeit verallgemeinert diese Techniken auf die drei linearen Modellprobleme das Poisson-Problem, die Stokes-Gleichungen und das lineare Elastizitätsproblem. Die Axiome der Adaptivität mit separiertem Markieren nach Carstensen und Rabus (SIAM J. Numer. Anal., 55(6), 2017) werden in drei Raumdimensionen nachgewiesen. Die Analysis umfasst Diskretisierungen mit beliebigem Polynomgrad sowie inhomogene Dirichlet- und Neumann-Randbedingungen. Abschließend bestätigen numerische Experimente mit dem h-adaptiven Algorithmus die theoretisch bewiesenen optimalen Konvergenzraten. / The least-squares finite element methods (LSFEMs) base on the minimisation of the least-squares functional consisting of the squared norms of the residuals of first-order systems of partial differential equations. This functional provides a reliable and efficient built-in a posteriori error estimator and allows for adaptive mesh-refinement. The established convergence analysis with rates for adaptive algorithms, as summarised in the axiomatic framework by Carstensen, Feischl, Page, and Praetorius (Comp. Math. Appl., 67(6), 2014), fails for two reasons. First, the least-squares estimator lacks prefactors in terms of the mesh-size, what seemingly prevents a reduction under mesh-refinement. Second, the first-order divergence LSFEMs measure the flux or stress errors in the H(div) norm and, thus, involve a data resolution error of the right-hand side f. These difficulties led to a twofold paradigm shift in the convergence analysis with rates for adaptive LSFEMs in Carstensen and Park (SIAM J. Numer. Anal., 53(1), 2015) for the lowest-order discretisation of the 2D Poisson model problem with homogeneous Dirichlet boundary conditions. Accordingly, some novel explicit residual-based a posteriori error estimator accomplishes the reduction property. Furthermore, a separate marking strategy in the adaptive algorithm ensures the sufficient data resolution. This thesis presents the generalisation of these techniques to three linear model problems, namely, the Poisson problem, the Stokes equations, and the linear elasticity problem. It verifies the axioms of adaptivity with separate marking by Carstensen and Rabus (SIAM J. Numer. Anal., 55(6), 2017) in three spatial dimensions. The analysis covers discretisations with arbitrary polynomial degree and inhomogeneous Dirichlet and Neumann boundary conditions. Numerical experiments confirm the theoretically proven optimal convergence rates of the h-adaptive algorithm.