Image-based 2D numerical modeling of oxide formation in self-healing CMCs
Résumé
Excellent lifetimes make Self-healing Ceramic-Matrix Composites (CMCs) are promising candidates for jet engine hot parts. These composites are constituted of a 3D arrangement of SiC fiber tows infiltrated by a multilayer matrix. A pyrocarbon interphase acts as a crack deviator, SiC matrix layers bring stiffness, and boron-containing phases are able to produce above 450°C a liquid oxide preventing further oxidation by a diffusion barrier effect. This paper introduces an image-based numerical simulation of the self-healing mechanism under oxygen gas. Existing 0D or 1D models give the time evolution of the oxygen concentration in the weakest fiber and deduce from it a global lifetime through an oxygen-controlled slow crack growth rate law. We propose an approach in which the resolution domain is a 2D FE mesh directly obtained from transverse images of a tow containing the crack. Oxygen diffusion, carbon consumption around the fibers, and conversion of boron carbide into boron oxide are simulated. The model solves mass balances in terms of heights of oxygen (gaseous or dissolved), liquid oxide, pyrocarbon, and boron-containing phase. All the heights are considered perpendicular to the image plane (thin layer approximation), and represent the unit-volume (per square meter) occupied by each phase. Preliminary results on images containing several dozens of fibers and a multilayer matrix are discussed.
Origine | Fichiers produits par l'(les) auteur(s) |
---|