Scalable adaptive framework for next generation multiphysics flow simulation
Is Version Of
Efficiently and accurately simulating partial differential equations (PDEs) in and around arbitrarily defined geometries, especially with high levels of adaptivity, has significant implications for different application domains and forms the basis for the next--generation of multiphysics simulation. Moreover, rapid changes in the hardware architecture in the exascale era require novel abstraction methods that ensure code portability across different architecture platforms without compromising overall performance. In this work, we present scalable algorithms capable of creating an analysis-suitable adaptive mesh in the presence of complex objects. Further, we present architecture-agnostic algorithms for efficient and scalable finite element computation. These algorithmic developments unlock the realm of fast, accurate, and efficient solutions for multiphysics PDEs over complex geometries, ushering in a new era of computational possibilities. In this thesis, we showcase this remarkable capability through a series of illustrative examples.