Les sources intrinsèques d'héterogénéité cellulaire, comme l'expression stochastique des gènes, sont de plus en plus reconnues comme jouant un rôle important dans la dynamique des tissus, tumeurs, communautés microbiennes... Cependant, elles sont souvent ignorées ou représentées de manière simpliste dans les modèles théoriques de populations de cellules. Dans cette thèse, nous proposons une approche cellule-centrée (chaque cellule est représentée de manière individuelle), multi-échelle (les décisions cellulaires sont placées sous le contrôle de voies de signalisation biochimiques simulées dans chaque cellule) pour modéliser la dynamique de populations de cellules. La nouveauté principale de cette approche réside dans la prise en compte systématique (pour toutes les protéines modélisées) des fluctuations du niveau des protéines résultant de l'expression stochastique des gènes. Cela permet d'étudier l'effet combiné des causes intrinsèques et environnementales d'héterogénéité cellulaire sur la dynamique de la population de cellules. Un élément central de notre approche est une stratégie parsimonieuse pour attribuer les paramètres de modèles d'expression stochastique des gènes. Nous appliquons cette approche à deux cas d'étude. Nous considérons en premier la resistance à l'agent anti-cancer TRAIL, qui peut induire l'apoptose sélectivement dans les cellules cancéreuses. Nous construisons d'abord un modèle 'cellule unique' de l'apoptose induite par TRAIL et le comparons à des données existantes quantitatives et 'cellules uniques'. Le modèle explique la mort fractionnelle (le fait que seul une fraction des cellules meurent à la suite d'un traitement) et prédit correctement l'héritabilité transiente du destin cellulaire ainsi que l'acquisition transiente de résistance, deux propriétés observées mais hors de portée des modèles pré-existants, qui ne capturent pas la dynamique de l'héterogénéité cellulaire. Dans une seconde étape, nous intégrons ce modèle dans des simulations multi-cellulaires pour étudier la résistance à TRAIL dans des scénarios virtuels intermédiaires entre les études classiques in-vitro et la réponse de tumeurs in-vivo. Plus précisément, nous considérons la réponse en temps long de sphéroides multi-cellulaires à des traitements répétés de TRAIL. L'analyse de nos simulations permet de proposer une explication originale et méchanistique de l'acquisition transiente de résistance, impliquant la dégradation ciblée des protéines activées et un différentiel dans le renouvellement des protéines pro- et anti- apoptotiques. Nous appliquons aussi notre approche à un système synthétique de création de motifs développé dans des levures par des collaborateurs. Nous nous concentrons d'abord sur un circuit senseur d'une molécule messager pour lequel nous construisons un modèle cellule unique qui capture de manière fine la dynamique de réponse du circuit telle qu'observée par cytométrie en flux. Nous intégrons ensuite ce modèle dans des des simulations multi-cellulaires et montrons que la réponse de micro-colonies organisées spatialement et soumises à des gradients de molécule messager est correctement prédite. Finalement, nous incorporons un modèle d'un circuit de mort et comparons les motifs prédits de cellules mortes/vivantes avec des données expérimentales, nous permettons de mieux comprendre comment les paramètres du circuit se traduisent en phénotypes d'organisation multi-cellulaire. Notre approche peut contribuer à l'obtention de modèles de populations de cellules de plus en plus quantitatifs, prédictifs et qui englobent l'échelle moléculaire. / Cell-intrinsic, non-environmental sources of cell-to-cell variability, such as stochastic gene expression, are increasingly recognized to play an important role in the dynamics of tissues, tumors, microbial communities... However, they are usually ignored or oversimplified in theoretical models of cell populations. In this thesis, we propose a cell-based (each cell is represented individually), multi-scale (cellular decisions are controlled by biochemical reaction pathways simulated in each cell) approach to model the dynamics of cell populations. The main novelty compared to traditional approaches is that the fluctuations of protein levels driven by stochastic gene expression are systematically accounted for (i.e., for every protein in the modeled pathways). This enables to investigate the joint effect of cell-intrinsic and environmental sources of cell-to-cell variability on cell population dynamics. Central to our approach is a parsimonious and principled parameterization strategy for stochastic gene expression models. The approach is applied on two case studies. First, it is used to investigate the resistance of HeLa cells to the anti-cancer agent TRAIL, which can induce apoptosis specifically in cancer cells. A single-cell model of TRAIL-induced apoptosis is constructed and compared to existing quantitative, single-cell experimental data. The model explains fractional killing and correctly predicts transient cell fate inheritance and reversible resistance, two observed properties that are out of reach of previous models of TRAIL-induced apoptosis, which do not capture the dynamics of cell-to-cell variability. In a second step, we integrate this model into multi-cellular simulations to study TRAIL resistance in virtual scenarios constructed to help bridging the gap between standard in-vitro assays and the response of in-vivo tumors. More precisely, we consider the long-term response of multi-cellular spheroids to repeated TRAIL treatments. Analysis of model simulations points to an novel, mechanistic explanation for transient resistance acquisition, which involves the targeted degradation of activated proteins and a differential turnover between pro- and anti- apoptotic proteins. Second, we apply our approach to a synthetic spatial patterning system in yeast cells developed by collaborators. Focusing first on a sensing circuit responding to a messenger molecule, we construct a single-cell model that accurately capture the response kinetics of the circuit as observed in flow cytometry data. We then integrate this model into multi-cellular simulations and show that the response of spatially-organized micro-colonies submitted to gradients of messenger molecules is correctly predicted. Finally, we incorporate a model of a killing circuit and compare the predicted patterns of dead or alive cells with experimental data, yielding insights into how the circuit parameters translate into multi-cellular organization phenotypes. Our modeling approach has the potential to accelerate the obtention of more quantitative and predictive models of cell populations that encompass the molecular scale.

Der transformierende Wachstumsfaktor-β (TGFβ) übt verschiedene Wirkungen auf die Regulierung zahlreicher biologischer Prozesse aus. Insbesondere die zytostatische Wirkung von TGFβ ist wichtig, um die Homöostase in Geweben aufrechtzuerhalten und proliferative Störungen, wie in Krebs, zu verhindern. Frühere Studien zur Regulation des Zellzyklus mit TGFβ wurden auf Populationsebene, oft durch physikalische oder chemische Synchronisation durchgeführt. Dabei wird die Heterogenität auf zellulärer Ebene vernachlässigt und die Anfälligkeit gegen potenzielle Artefakte erhöht. Um zu verstehen, wie einzelne Zellen TGFβ-Signale entschlüsseln und diese in die Entscheidung zur Zellproliferation integrieren, wurden sowohl die Dynamik der TGFβ-Signale als auch die Zellzyklusprogression in asynchronen Zellen durch „Live Cell Imaging“ quantifiziert. In Kombination von experimentellen und theoretischen Studien wurde gezeigt, dass TGFβ einen „Alles-oder-Nichts-G1- Stillstand“ auslöst, der sowohl dosisabhängig als auch phasenabhängig ist. Wenn die Zellen während der S / G2 / M-Phase TGFβ ausgesetzt werden, erfahren sie in der darauf folgenden G1-Phase einen ererbten, verzögerten Stillstand. Zusätzlich sind die Zellen nach einem TGFβ-Stimulationsimpuls für weitere TGFβ-Behandlungen unempfindlich. In Anbetracht der Bedeutung von Einzelzellinformationen und den Herausforderungen bei der automatischen Zellverfolgung wurde ein Rahmenkonzept von „Population to Single Cell“ (P2S-Framework) erarbeitet, um von der Populationsdynamik auf die Abstammung einzelner Zellen zu schließen. Zusammengefasst bietet diese Arbeit neue Einblicke in Strategien zur Kontrolle der Zellproliferation durch Manipulation der TGFβ-Signalgebung. / The transforming growth factor-β (TGFβ) exerts diverse effects on regulating numerous biological processes. Especially, the cytostatic effect of TGFβ is important for maintaining tissue homeostasis and preventing proliferative disorders, like cancer. Previous studies on the regulation of cell cycle by TGFβ were conducted at the population level, and often through physical or chemical synchronization, which neglected cellular heterogeneity and might introduce artifacts. To understand how individual cells decode and integrate TGFβ signals into cell proliferation decisions, we quantitatively characterized both TGFβ signaling dynamics and cell cycle progression in asynchronous cells by live cell imaging. Combining experimental and theoretical studies, we demonstrated that TGFβ triggers all-or-none G1 arrest, which is both dose-dependent and phase- dependent. When exposed to TGFβ during S/G2/M phase, cells undergo an inherited, delayed arrest at the next G1 phase. In addition, after one pulse of TGFβ stimulation, cells are refractory to further TGFβ treatments. Considering the importance of single cell information and challenges in automatic cell tracking, we proposed a Population to Single cell framework (P2S framework) to infer single- cell lineages from population dynamics. Taken together, this work provides new insight into strategies to control cell proliferation by manipulating TGFβ signaling.

Au cours des prochaines années, neuf centrales nucléaires de type UNGG (Uranium Naturel Graphite Gaz) devront être démantelées en France. Ces centrales utilisent le graphite comme modérateur et réflecteur de neutrons. Pendant leur exploitation, celui-ci est activé. Leur démantèlement conduira à 23000 tonnes de déchets de graphite irradiés à gérer. Ce travail focalise sur deux radionucléides contenus dans ces déchets : le 36Cl et le 3H. Le 36Cl a l'une des demi-vies les plus longues (301 000 ans). Par contre, le 3H a une demi-vie plus courte (12 ans), mais contribue beaucoup à l'activité initiale des déchets. Différentes données expérimentales suggèrent que le 36Cl et le 3H sont piégés à différents endroits du graphite, comme les boucles de dislocation, les surfaces ou les joints de grains. Le seul mécanisme de migration des radionucléides est le relâchement. Pour cette raison, il est important de comprendre quels sont les pièges et les différentes conditions du relâchement.Le graphite UNGG a une structure complexe, hétérogène et multi-échelle qui diffère du monocristal idéal du graphite. Cependant, pour comprendre les données macroscopiques, les études théoriques à l'échelle nanoscopique et microscopique sont des outils importants, même si elles reposent sur des modèles plus simples. Dans cette thèse, une approche multi-échelle a été utilisée afin d’étudier les interactions des radionucléides avec le graphite ainsi que les mécanismes de diffusion et de piégeage à l'échelle du nm-μm.Les interactions du 3H et du 36Cl avec différents défauts du graphite ont été étudiées dans le cadre de la théorie fonctionnelle de la densité (DFT). L'hydrogène forme une liaison covalente avec le graphite massique ainsi qu'avec ses surfaces (001), (100) et (110). Plusieurs reconstructions de surface ont été explorées. Les résultats montrent que les hypothèses existantes sur le piégeage de l'hydrogène doivent être affinées. Le comportement du Cl est plus complexe. Sa chimisorption est observée sur les surfaces (100) et (110). Cependant, sur la surface (001), le Cl interagit par transfert de charge. Le Cl2 n'interagit que par interactions de van der Waals avec celle-ci. Le Cl2 se dissocie dans le graphite massique.Les diffusions du H et du Cl dans le graphite irradié ont été étudiées en effectuant des simulations de dynamique moléculaire. Les résultats ab initio ont été utilisés pour développer des potentiels de type « bond order » afin de modéliser l'interaction des radionucléides avec la matrice de graphite, qui possède des contributions à court et à long portée. Pour le Cl, un nouveau potentiel a été paramétré qui reproduit toutes les données obtenues au niveau DFT. Pour les interactions 3H-graphite, les potentiels AIREBO/M, pour les interactions C-H, et LCBOP, pour les interactions C-C, ont été utilisés.Pour évaluer l'influence de la structure complexe du graphite UNGG sur le comportement des radionucléides, plusieurs modèles atomiques ont été utilisés pour rendre compte de cette diversité, tels que les surfaces, les joints de grains et les nanopores.Pour le Cl, des simulations d'irradiation ont été réalisées pour une gamme d’énergie allant de 1 à 10 keV et une gamme de température de 200 à 500ºC. Les dépendances à la température et à la direction d'irradiation ont été étudiées. D’une façon générale, les dommages causés par l'irradiation perpendiculaire aux surfaces augmentent avec la température. L'irradiation à des angles d’incidence <90º aux surfaces peut causer plus ou moins de dommages par rapport à l'irradiation perpendiculaire selon le type de surface.Les diffusions du H et du Cl montrent que tous les bords de cristallites avec des liaisons pendantes sont des pièges. Pour le Cl, la diffusion dans le graphite nanoporeux a révélé deux emplacements préférés: les bords des cristallites où le Cl forme une liaison covalente et les coins des microfissures où le Cl interagit par transfert de charge. / In the upcoming years, nine nuclear UNGG (Uranium Naturel Graphite Gaz) power plants will have to be dismantled in France. In these power plants, nuclear graphite was used as a neutron moderator and reflector, and was activated during operation. The dismantlement will lead to 23000 tons of irradiated graphite waste, which will have to be managed. The graphite is classified as a nuclear waste containing radionuclides with low activity and long half-life. Two radionuclides are the focus of this work: 36Cl and 3H. 36Cl has one of the longest half-lives (about 301000 years) among the waste's radionuclides. 3H has a shorter half-life (12 years), but contributes significantly to the waste’s initial activity. Previous experiments suggest that both, 36Cl and 3H, are mainly fixed at different traps in graphite, which are defective structures, such as dislocation loops, surfaces, or grain boundaries. Since the only significant migration mechanism of these radionuclides is release, it is important to understand where the traps are located and the conditions of the release.UNGG graphite has a complex heterogeneous multi-scale structure which differs substantially from an ideal monocrystal of graphite. However, in order to understand macroscopic data, theoretical studies at the nano- and microscopic scale are an important tool to explain underlying phenomena even though they rely on simpler models due to the limitations of computation power. A multi-scale approach was therefore applied to study the local interactions of the radionuclides with graphite as well as diffusion and trapping mechanisms on the nm-μm length scale.First, the interaction of 3H and 36Cl with defects in graphite was studied with density functional theory (DFT). Hydrogen interacts covalently with bulk graphite as well as with the studied surfaces (001), (100), and (110). Several surface reconstructions were investigated: arch-type reconstructions and in-plane reconstructions. The results show that the existing hypothesis on the trapping of hydrogen needs to be refined. The behavior of Cl is more complex. On the (100) and (110) surface chemisorption is observed. However, on the (001) surface a strong charge transfer interaction is observed for Cl. In contrast to that, Cl2 only interacts via weak van der Waals interactions with this surface. In bulk graphite Cl2 dissociates.The diffusion of H and Cl in irradiated graphite has been investigated by performing molecuar dynamics simulations. The ab initio results were used to develop bond order potentials to model the interaction of radionuclides and the graphite matrix, which attributes for short and long range interactions. For Cl, a new potential has been parameterized which is able to describe all aspects obtained with DFT. For the 3H-graphite interactions, the bond order potential AIREBO/M was used for C-H interactions. For C-C interactions the LCBOP potential was used.To evaluate the influence of the complex heterogeneous structure of the UNGG graphite on the radionuclide's behavior, several different atomic models were studied to account for this diversity such as surfaces, grain boundaries and nanopores.For Cl, irradiation simulations of different systems were performed up to an energy of 10 keV for the primary knock-on atom (PKA), and in a temperature range of 200 to 500ºC. The dependence on temperature and irradiation direction was investigated. In general, direct irradiation damage increases with temperature. Irradiation at incident angles <90º can create more or less damage compared to the perpendicular one depending on the surface type.Diffusion of H and Cl along surfaces shows that all crystallite edges with dangling bonds can serve as traps. For Cl, diffusion in nanoporous graphite revealed two preferred locations : First, the crystallite edges where Cl forms strong covalent; second, the corners of microcracks where Cl interacts via charge transfer.

Over the last few decades, rechargeable lithium-ion batteries have been extensively used in portable instruments due to their high energy density and low self-discharge rate. Recently, lithium-ion batteries have emerged as the most promising candidate for electric vehicles and stationary energy storage. However, the maximum energy that lithium-ion batteries can store decreases as they are used because of various irreversible degradation mechanisms. Lithium-ion batteries are complex systems to understand, and various processes and their interactions are responsible for the degradation over time. The mechanical integrity and stability of the electrode layers inside the battery highly influence the battery performance, which makes it a necessity to characterize the mechanical behavior of electrode active layers for mesoscopic and macroscopic level modeling. In papers 1 and 2, the macroscopic mechanical behavior of active layers in the electrodes is investigated using U-shape bending tests. The active layers are porous and a different tensile and compressive behavior is captured by performing tests on single side coated dry specimens. The experiments reveal that the active layer is stiffer in compression as compared to tension. The compressive stiffness increases with bending strain whereas the tensile stiffness is almost independent of the bending strain. A very low value of modulus of the active layer (1-5 GPa) is measured in comparison to the metal foils (70-110 GPa) and the active particles (50-200 GPa) which shows that the electrode properties are governed majorly by the binders present in the active layers.  The time-dependent and hysteresis effects are also captured by the method which circumvents the flaws of many other test methods presented in the literature.   In paper 3, we present a multiscale homogenization method that couples mechanics and electrochemistry at the particle, electrode, and battery scales. The active materials of lithium-ion battery electrodes exhibit volume change during lithium intercalation or deintercalation. A lithium concentration gradient develops inside particles, as well as inside the active layer. The developed stress due to deformations further affects solid diffusion.  We utilized models that have already been developed to couple particle and electrode layer levels. The mechanical coupling between the electrode and the battery level is achieved by homogenization of the layered battery using three-dimensional laminate theory.  By application of the model, we demonstrate that the stresses on all considered scales can be predicted from the homogenized model. It is furthermore demonstrated that the effects of external battery loadings like battery stacks, casings, and external pressure can be captured by the model. The model can also capture the effect of various electrochemical cycling rates and mechanical parameters like layer thicknesses, stiffnesses, and swelling properties. The presented multi-scale model is fast, accurate and the efficiency of the method is demonstrated by comparisons to detailed finite element computations where each layer is individually modeled.

Liver resection causes marked perfusion alterations in the liver remnant both on the organ scale (vascular anatomy) and on the microscale (sinusoidal blood flow on tissue level). These changes in perfusion affect hepatic functions via direct alterations in blood supply and drainage, followed by indirect changes of biomechanical tissue properties and cellular function. Changes in blood flow impose compression, tension and shear forces on the liver tissue. These forces are perceived by mechanosensors on parenchymal and non-parenchymal cells of the liver and regulate cell-cell and cell-matrix interactions as well as cellular signaling and metabolism. These interactions are key players in tissue growth and remodeling, a prerequisite to restore tissue function after PHx. Their dysregulation is associated with metabolic impairment of the liver eventually leading to liver failure, a serious post-hepatectomy complication with high morbidity and mortality. Though certain links are known, the overall functional change after liver surgery is not understood due to complex feedback loops, non-linearities, spatial heterogeneities and different time-scales of events. Computational modeling is a unique approach to gain a better understanding of complex biomedical systems. This approach allows (i) integration of heterogeneous data and knowledge on multiple scales into a consistent view of how perfusion is related to hepatic function; (ii) testing and generating hypotheses based on predictive models, which must be validated experimentally and clinically. In the long term, computational modeling will (iii) support surgical planning by predicting surgery-induced perfusion perturbations and their functional (metabolic) consequences; and thereby (iv) allow minimizing surgical risks for the individual patient. Here, we review the alterations of hepatic perfusion, biomechanical properties and function associated with hepatectomy. Specifically, we provide an overview over the clinical problem, preoperative diagnostics, functional imaging approaches, experimental approaches in animal models, mechanoperception in the liver and impact on cellular metabolism, omics approaches with a focus on transcriptomics, data integration and uncertainty analysis, and computational modeling on multiple scales. Finally, we provide a perspective on how multi-scale computational models, which couple perfusion changes to hepatic function, could become part of clinical workflows to predict and optimize patient outcome after complex liver surgery.

No description available.

No description available.

<p>This thesis is concerned with particle subgrid scale (SGS) modeling in large-eddy simulation (LES) of particle-laden turbulence. Although most particle-laden LES studies have neglected the effect of the subgrid scales on the particles, several particle SGS models have been proposed in the literature. In this research, the approximate deconvolution method (ADM), and the stochastic models of Fukagata et al. (2004), Shotorban and Mashayek (2006) and Berrouk et al. (2007) are analyzed. The particle SGS models are assessed by conducting both a priori and a posteriori tests of a periodic box of decaying, homogeneous and isotropic turbulence with an initial Reynolds number of Re=74. The model results are compared with particle statistics from a direct numerical simulation (DNS). Particles with a large range of Stokes numbers are tested using various filter sizes and stochastic model constant values. Simulations with and without gravity are performed to evaluate the ability of the models to account for the crossing trajectory and continuity effects. The results show that ADM improves results but is only capable of recovering a portion of the SGS turbulent kinetic energy. Conversely, the stochastic models are able to recover sufficient energy, but show a large range of results dependent on Stokes number and filter size. The stochastic models generally perform best at small Stokes numbers. Due to the random component, the stochastic models are unable to predict preferential concentration.</p> / Master of Applied Science (MASc)

Every moment of every day, at least hundreds of thousands of tires roll across a surface throughout the world. Tires are indisputably important in our daily life. The tire's primary component is rubber, which consumes energy when it rotates on a substrate due to the viscoelastic material's internal friction: a phenomenon referred to as rolling resistance. The interaction between the tire and the road surface is one of the most intricate and crucial phenomena in an automobile, because it is responsible for creating forces, moments, and deformation in the tire. Additionally, the road's roughness interacts with the tire and contributes significantly to its performance. This dissertation aims to develop a comprehensive physics-based model for predicting the rolling resistance of a viscoelastic material due to dynamic deformations caused by tire rotation using an analytical approach. The model was developed by proposing a Gaussian wave function propagating across a tire circumference's viscoelastic medium. The wave function was selected to describe the displacement field produced by tire-road interaction. Additionally, by adopting a multi-scale modeling technique, the model was upgraded to estimate rolling resistance while taking into account surface roughness at all length scales, from macroscopic to microscopic. Additionally, another mathematical model was developed using the Fourier series approach to evaluate the steady-state stress response and energy dissipation for any harmonic and non-harmonic periodic strain signals. Additionally, the dissertation strove to build a continuum damage mathematical model using a combined testing/modeling methodology to predict the aging of Styrene-Butadiene Rubber (SBR) after continuous exposure to the atmosphere. The obtained model was developed through the implementation of optimization techniques while formulating a mathematical model, which was then combined with a physics-based model to predict rolling resistance while taking into account rubber aging. Calibration of hyperelastic and viscoelastic material models with testing data was performed using an optimization technique that yielded sufficient results. The results of all mathematical models obtained in this dissertation are reported subsequently. The stress response of a viscoelastic material under harmonic and non-harmonic strain input yielded good agreement with the FEA model obtained using ABAQUS. The rolling resistance behavior under various operating conditions, including texture and aging effects, was reported, and the results aligned with the experimental results found in the literature. / Doctor of Philosophy / Every moment of every day, hundreds of thousands of automobile tires roll across a surface somewhere in the world. A tire is an undeniably important part of everyday life. Rubber is the tire's main component, and when it rotates on a surface, it loses energy, resulting in a force that resists motion, known as rolling resistance force. The contact between the tire and the road is one of the most complicated and important phenomena that happens in an automobile because it is responsible for the vehicle's dynamic performance in areas such as acceleration, stopping distance, and stability. Another factor that affects tire and car performance and should be taken into account is the road's roughness. This dissertation used an analytical method to come up with an accurate physics-based model for predicting the rolling resistance force of a viscoelastic material caused by tire rotation. The model was developed by assuming a Gaussian wave function would move across the tire circumference. Additionally, using a multi-scale modeling technique, the model was improved so that it could calculate the value of rolling resistance force considering surface roughness in all lengths of scale. This project also developed an additional mathematical model using the Fourier series method to determine how the stress response and energy dissipation would behave for any harmonic and nonharmonic periodic strain signals. Additionally, the dissertation presents the developing of a continuum damage mathematical model that could predict the material property of styrene-butadiene rubber (SBR) after being exposed to the air for a long time (i.e., aged). The model was developed based on experimental data and optimization techniques. This model was then combined with a physics-based model to predict rolling resistance force while taking aging into account. The material models were defined using an optimization method that yielded good results. The stress response of a viscoelastic material when it was subjected to harmonic and non-harmonic strain was in good agreement with the Finite Element Analysis (FEA) model made with ABAQUS. Rolling resistance behavior was observed, and the results were consistent with those found in the literature.

Les outillages de mise en forme en acier martensitique de type AISI H11 sont des pièces critiques dont le comportement en service est étroitement lié à leurs structures internes et à leur évolution. Les conditions des sollicitations lors de la mise en oeuvre du procédé est souvent à l'origine de modifications microstructurales en surface, à savoir la morphologie des lattes de martensite, les orientations cristallographiques, l'état d'écrouissage interne ou encore le profil de surface. Ces aspects peuvent éventuellement altérer les performances mécaniques de l'acier AISI H11. Afin d'appréhender et d'optimiser le comportement mécanique de celui-ci, une approche multi-échelle est mise en oeuvre dans ce travail. Celle-ci s'articule autour d'une investigation expérimentale et d'un traitement numérique. L'étude expérimentale s'attache à reproduire, à l'échelle du laboratoire, des surfaces équivalentes à celles issues lors des procédés de mise en oeuvre des outillages. Des techniques de caractérisation spécifiques, à savoir le MEB, l'EBSD, la nanoindentation ou encore l'altimétrie permettent de mettre en évidence un gradient de la stéréologie du matériau en surface et sous-surface. Les hétérogénéités locales induites concernent la morphologie des lattes de martensite, les orientations cristallographiques, l'état d'écrouissage interne mais également le profil de surface. Des essais mécaniques in-situ associés à la technique de corrélation d'images numériques sont réalisés pour des chargements monotones quasi-statiques et cycliques de type traction-traction. Une investigation des champs mécaniques locaux en surface est ainsi effectuée, elle permet d'analyser les schémas de localisations des déformations non linéaires liés aux artéfacts stéréologiques. Le traitement numérique s'intéresse à une modélisation multi-échelle, et plus particulièrement à des calculs par la méthode des éléments finis sur des microstructures virtuelles générées par tesselations de Voronoï. Celles-ci sont effectuées de manière à reproduire les structures martensitiques et considèrent des relations d'orientations spécifiques (de type Kurdjumov-Sachs) à l'issue du traitement thermique entre les lattes de martensite et le grain austénitique parent. Les équations constitutives du modèle de plasticité cristalline (élasto-viscoplastique) de Méric-Cailletaud sont implantées dans le code de calcul par éléments finis Abaqus dans le cadre de l'hypothèse des petites perturbations (HPP) et de la théorie des transformations finies. La formulation du modèle dans le contexte de la théorie des transformations finies est effectuée dans le cadre d'une description spatiale où la notion de dérivée objective est considérée. Celle-ci consiste en celle d'Oldroyd ou de Truesdell de manière à ce qu'une telle formulation soit équivalente à une description lagrangienne. Le traitement numérique a permis de reproduire de manière qualitative les schémas de localisation en surface mise en évidence lors de l'investigation expérimentale. L'influence des divers paramètres stéréologiques, évoqués ci-dessus, sur les champs mécaniques locaux a été analysée. De par cette approche, il a été possible de mettre en évidence certains mécanismes élémentaires, notamment les effets d'interaction et de surface. Enfin, il a été constaté que la prise en compte des rotations des réseaux cristallins par la théorie des transformations finies permet de relâcher certaines zones de localisation des champs mécaniques autour d'artéfacts stéréologiques. / AISI H11 martensitic tool steels are critical mechanical components that behaviour during service is drastically linked to their internal structures and their possible evolution. Their manufacture processes are often at the origin of microstructural changes at the surface, namely the morphology of martensitic laths, the crystallographic orientations, the internal hardening state and the surface profile These aspects can potentially alter the mechanical performance of AISI H11 martensitic steel. In order to get better insight into and optimize its mechanical behaviour, a multi-scale approach involving an experimental investigation and a numerical treatment is taken in this work.The experimental investigation focuses to reproduce, at the laboratory scale, equivalent surfaces to those resulting from tool steels manufacture processes. Specific characterization techniques, namely SEM, EBSD, nanoindentation and altimetry enable to highlight a stereology gradient of the material in surface and sub-surface. The induced local heterogeneities consist in morphology of martensitic laths and crystallographic orientations, internal hardening state and surface profile. In-situ mechanical tests with digital image correlation technique (DIC) are carried out for monotonous quasi-static and tension-tension cyclic loads. An investigation of the local mechanical fields at the surface is thus performed and allows to analyze the localizations schemes of nonlinear strains which are related to stereological artifacts.The numerical treatment is focused on a multi-scale modelling, and more particularly on finite element calculations on virtual microstructures which are generated by Voronoi tesselations. The latters are carried out such that to reproduce martensitic structures and consider a specific orientation relationship between martensitic laths and parent austenitic grains (i.e. Kurdjumov-Sachs) after the heat treatment. The constitutive equations of the (elasto-viscoplastic) crystal plasticity of Méric-Cailletaud are implemented in the finite element code Abaqus in the context of the small strain assumption and the finite strain theory. The formulation of the model in the context of finite strain theory is is given a spatial description where the notion of objective derivative, namely the so called one of Oldroyd or Truesdell, is used in such a way that such formulation is equivalent to a Lagrangian description.The numerical treatment has allowed to qualitatively reproduce the localization patterns at the surface which have been highlighted in the experimental investigation. The influence of the different stereological parameters mentioned above on the local mechanical fields was analyzed. By this approach, it was possible to highlight some elementary mechanisms including interaction and surface effects. Finally, it was found that the inclusion of lattice rotations via the theory of finite strain allows to release certain areas of mechanical fields localization that are related to stereological artifacts.