StrUcture PREserving Methods on Unstructured grids for continuum Mechanics

forthcoming

SUPREMUM aims to develop and theoretically study innovative structure-preserving (SP) numerical methods to solve challenging non-linear hyperbolic partial differential equations (PDE) modelling continuum mechanics. The ultimate groundbreaking goal consists on devising the first existing numerical methods simultaneously verifying at the discrete level all structural properties of the continuous models: provable thermodynamical compatibility, so that the systems satisfy an extra conservation law for the total energy in order to rigorously prove non-linear stability; conservation of equilibrium solutions, needed to perform very long-time stable simulations; verification of natural involution constraints, of the divergence and curl type; and preservation of the asymptotic limits, arising for some characteristic scales tending to zero. We will devise efficient high order numerical schemes combining different families of numerical methods on unstructured staggered grids configurations, leading to the advance of fundamental mathematical knowledge on numerical analysis. The use of an arbitrary Eulerian-Lagrangian approach will allow the design of an efficient and accurate fluid structure interaction (FSI) solver based on the unified hyperbolic Godunov-Peshkov-Romenski (GPR) model of continuum mechanics able to represent both solids and fluids. Fundamental progress in practical applications of the developed methodologies will initially target the design of a monolithic approach for the simulation of entire cardiovascular systems, coupling 1D blood flow models to the 3D FSI approach. Finally, we will build the bridge between SP methods for hyperbolic PDEs and reduced order modelling (ROM) breaking the frontiers of present research to obtain a novel SP-ROM approach for conservation laws founded on a solid mathematical background. The investigations pursued will impact digital transformation opening new research avenues in numerical analysis, engineering and biomedicine.