Title: Modeling of microcapsules flowing in a multi-bifurcated networkDate: 2020-04-15 20:36Slug: job_f1f18819cfa3335f365364a35a4ea80eCategory: jobAuthors: Anne-Virginie SalsacEmail: email@example.comJob_Type: ThèseTags: theseTemplate: job_offerJob_Location: CompiègneJob_Duration: 3 ansJob_Website: http://www.utc.fr/~salsacan/ Job_Employer: Université de Technologie de Compiègne - CNRSExpiration_Date: 2020-06-10Attachment: job_f1f18819cfa3335f365364a35a4ea80e_attachment.pdfMicro-capsules, which are fluid droplets enclosed in a thin elastic membrane, are current in nature (red blood cells, phospholipidic vesicles) and in various industrial applications (biotechnology, pharmacology, cosmetics, food industry). They are used to protect and transport active principles, by isolating them from the external suspending fluid. One application with high potential is the use of microcapsules for active substance targeting. But, once injected in the blood flow, the behaviour of the particles is not fully understood. The microcapsules are indeed subjected to large deformations, because of the strong interactions in the flow, but these strongly depend on the vessel network. There is currently a lack of understanding of the distribution of deformable microparticles in a complex bifurcated network.The objective of the project is to study the flow of a dilute suspension of capsules flowing in a capillary network composed of bifurcations in cascade. Their flow and deformation will be modeled numerically with Capsilisk, a numerical code that we have recently developed by coupling the Finite Element solver of Caps 3D with Basilisk (a Finite Volume code) using a Boundary Immersed Method. The goal will first be to optimize the code (ex: parallelization) and set up simulations in bifurcations made of one inlet and several outlets. A database will be generated by changing consistently the geometry of the network and the properties (size and resistance) of the capsules. The last objective of the project will be to build reduced order models using physics-based artificial intelligence techniques in order to predict the capsule distribution in the network in real time.