Preconditioning of multiphysics problems with applications to the biomechanics of the brain

A image


Advanced Numerical Methods for Scientific Computing
Computational Medicine for the Cardiocirculatory System
Kent - Andre Mardal
University of Oslo and Simula Research Laboratory
Thursday 11th July 2019
Aula consiglio VII piano, Edificio 14, Dipartimento di Matematica POLITECNICO DI MILANO
In this talk we will discuss preconditioning algorithms for monolithic schemes of coupled problems involving the coupling of porous and viscous flow as well as fluid-structure interaction and dimension reduction problems. We explore how fractional Laplacian solvers may be utilized to obtain parameter robust schemes. The schemes are discussed in the context of biomechanical modelling of the waste clearance processes in the brain that is believed to fail in various forms of dementia such as Alzheimer's and Parkinson's diseases. Contact:
Kent-Andre Mardal is a Professor at the University of Oslo and an adjuct researcher at Simula Research Laboratory. His research focus is on computational methods for life-science applications, where focus areas are robust multi-physics algorithms, high-performance preconditioning techniques, generic scientific software components, the microcirculation and metabolism of the brain, the cause of dementia such as Alzheimer’s disease and normal pressure hydrocephalus, and multi-dimensional coupling problems. Mardal has made core contributions to the FEniCS software library, the operator preconditioning technique, demonstrated transitional flow at low Reynolds number in blood flow and made crucial observations concerning the g-lymphatic system potentially involved in Alzheimer´s disease development.