Comparison of Two-Level Preconditioners using Deflation Techniques applied to Flow Problems

Master Thesis (2018)
Author(s)

J.Y.Y. Tjan (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Contributor(s)

C. Vuik – Mentor

Gabriela Diaz Cortes – Mentor

Damiano Pasetto – Graduation committee member

H. Schuttelaars – Graduation committee member

Faculty
Electrical Engineering, Mathematics and Computer Science
Copyright
© 2018 Jenny Tjan
More Info
expand_more
Publication Year
2018
Language
English
Copyright
© 2018 Jenny Tjan
Graduation Date
30-08-2018
Awarding Institution
Delft University of Technology
Faculty
Electrical Engineering, Mathematics and Computer Science
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

We investigate the simulation of one-phase and two-phase flow through heterogeneous porous media.The derived matrix, resulting from reservoir simulation of groundwater flow problems, can result in a large and ill-conditioned system, i.e. the matrix has a high condition number, and the modelling takes large computation time. In this thesis report, the Two-Level Preconditioned Conjugate Gradient method with deflation techniques will be considered. Recently, new Two-Level preconditioners are constructed using the AMG approach. In this research we compare this approach with the standard Two-Level preconditioners. We found that the performance of these methods can be improved by using a special starting vector and previous time-step as initial condition. From the results we see that the performance and the memory storage of the methods are similar. However, the cheapest methods per iteration resulted DEF1, DEF2, R-BNN2, and ROM.

Files

Final_Thesis.pdf
(pdf | 2.01 Mb)
License info not available