PERSONAL INFORMATION - PhD project
The subject of the PhD project is:
"Two-level Preconditioned Conjugate Gradient Methods with Applications to Bubbly Flows".
Deflation is a topic in the Numerical Linear Algebra. It is embedded in iterative methods to solve problems in for instance seismic and fluid dynamics more efficiently. More information about deflation can be found here.
Water droplet: an application of a two-phase bubbly flow problem
In the PhD research, the deflation method is examined in more detail. The focus is on both theoretical and numerical aspects.
- Theoretically: it is investigated whether the deflation method can be applied to linear systems with varying and singular coefficient matrices. In addition, we examine the optimal implementation of the deflation method and we compare the method to other well-known two-level preconditioned conjugate gradient methods. We combine the ideas from different fields to improve the deflation method.
- Numerically: the deflation method is applied to moving boundary problems, in particular two-phase bubbly flow problems. We investigate the optimal implementation in 2-D and 3-D simulations. Additionally, some efforts are done to improve the convergence and scalability of the deflation method. Several comparisons are carried out using different two-level preconditioned conjugate gradient methods in order to find the most effective, robust and fastest method for bubbly flow applications.
Some publications with respect to the research can be found here.
|
