L'objet de cette thèse est de présenter une étude du comportement thermomécanique des disques de frein automobiles pour la prédiction de leur tenue en fatigue. La stratégie de calcul numérique est repose sur le code de calcul Ansys v.11. Ce dernier qui est basé sur la méthode des éléments finis et qui possède des algorithmes de gestion du contact avec frottement est utilisé pour simuler dans l'application du freinage le comportement du mécanisme malgré son complexité. Dans un premier temps, est présentée une analyse des phénomènes thermiques opérant dans un disque de frein en service (flux de chaleur généré par frottement, gradients thermiques élevés, élévation de température). Cette modélisation est effectuée en tenant compte l'influence d'un certains nombre de paramètre tel que le type de freinage, le mode de refroidissement, les matériaux de conceptions. Ensuite, une étude purement mécanique du contact sec entre le disque et plaquettes est développée avec une bonne prédiction devient un enjeu majeur pour les industriels tout en modélisant le chargement et les conditions aux limites autours du disque. Nous avons utilisé le même code de calcul pour visualiser les déplacements, les déformations globales dans le disque, les contraintes de cisaillement, les contraintes de Von Mises et les outils de contact des plaquettes tout en effectuant une étude paramétrique telle que ( le module de Young des plaquettes, le coefficient de frottement , le type de chargement , la vitesse de rotation du disque ,..) pour voir sa sensibilité sur les résultats de calcul. Ainsi, les analyses faites sur le comportement thermique et mécanique que de ces prototypes montrent que ces types de solutions technologiques représentent de réelles pistes d'amélioration qui répond au besoin de l'ingénieur en charge de la conception des disques de frein.

Class I PI3Ks are a family (α, β, δ and γ) of ubiquitous lipid kinases that can be activated by cell surface receptors to 3-phosphorylate PI(4,5)P2 (phosphatidylinositol(4,5)-bisphosphate) and generate the signalling lipid PI(3,4,5)P3. The PI(3,4,5)P3 signal then activates a diverse collection of effector proteins involved in regulation of cell migration, metabolism and growth. The importance of this network is evidenced by the relatively high frequency with which cancers acquire gain-of-function mutations in this pathway and huge efforts to make PI3K inhibitors to treat cancer. The canonical model describing these events suggests class I PI3Ks are activated at the plasma membrane and generate PI(3,4,5)P3 in the inner leaflet of the plasma membrane where its effectors are activated. The PI(3,4,5)P3 signal can be terminated directly, by the tumour-suppressor and PI(3,4,5)P3-3-phosphatase PTEN, or modified to a distinct PI(3,4)P2 signal, by SHIP-family 5-phosphatases. The PI(3,4)P2 is removed by INPP4-family 4-phosphatases. Published work has shown that PI(3,4,5)P3 signalling can also occur in endosomes and nuclei, however, there is very little data defining the intracellular distribution of endogenous class I PI3Ks that supports these ideas; this is as a result of technical problems such as; their very low abundance, poor antibody-based tools and artefacts generated by overexpression of PI3Ks. Past work has indicated that, in PTEN-null mouse models of prostate tumour progression, either PI3Kβ or PI3Ks α and β, have important roles. Furthermore, the cell types and mechanism involved remained unclear. Recent published work in the host laboratory had indicated that there is an unexpectedly large accumulation of PI(3,4)P2 in PTEN-null cells that might be an important part of its status as a major tumour suppressor. The explanation and prevalence of this observation was unclear but potentially a result of PTEN also acting as a PI(3,4)P2 3-phosphatase in vivo. MEFs were derived from genetically-modified mice expressing endogenous, AviTagged class I PI3K subunits and used in experiments to define the subcellular localisation of class I PI3Ks. We found that following stimulation with PDGF, class IA PI3K subunits were unexpectedly depleted from the adherent basal membrane, in contrast, p85α and p110α, but not p85β and p110β, accumulated transiently in the nucleus. Interestingly, p110β, but none of the other subunits, was constitutively localised in the nucleus. These results support the idea that class I PI3K and PI(3,4,5)P3 signalling occurs in the nucleus. In organoids derived from WT, PI3Kγ-null or PTEN-null mouse prostate, application of PI3K-selective inhibitors revealed that PI3Kα had a dominant role in generating PI(3,4,5)P3 in prostate epithelial cells. The levels of PI(3,4)P2 were also elevated substantially in PTEN-null, compared to WT prostate organoids, use of PI3K-selective inhibitors suggested that it was also generated by PI3Kα. These data were consistent with the idea that PTEN can act as a PI(3,4)P2 3-phosphatase. Surprisingly, raising the pH of the organoids medium dramatically increased accumulation of PI(3,4,5)P3 and PI(3,4)P2, although the cause of this effect was unclear, we hypothesised the pH of the local environment may influence signalling via class I PI3Ks.