Efficient approximation and preconditioning strategies for models of swirling flow

The Navier-Stokes equations are a fundamental tool for the mathematical modelling of fluid flow. Discretised versions of the equations are routinely solved using computational fluid dynamics software in a variety of important application areas. These range from weather prediction in meteorology to the design of fuel-efficient aeroplanes in aeronautical engineering to human heart modelling in medical physics.

The aim of this project is to develop new computational techniques for efficiently generating numerical solutions of the Navier-Stokes equations expressed in a cylindrical polar coordinate system. In particular, we will design, analyse and test novel approximation strategies that ensure mass conservation at every point in the flow region. The main deliverables will be provably accurate computational tools and open-source software for other researchers to use and develop further.
This project addresses a topic that several international research groups are actively working on. Research funding is needed to enable the principal investigator to make a focused and sustained effort in attacking a topical research problem in an internationally competitive research area.