Back to search

University of Crete (UOC)

Numerical methods for evolutionary differential equations that maintain the positivity of the solution

Abstract

dc:description

We consider the Keller-Segel model of chemotaxis on a bounded domain \Omega\subset\mathbb{R}2. Assuming appropriate conditions in order to have bounded solutions for all times, we discretize the spatial variable with the standard linear finite element method and the temporal variable with implicit Euler and the implicit midpoint rule. We prove that the resulting semi-discrete and fully-discrete scheme via backward Euler can take negative values, therefore our aim is to seek conditions on existing or appropriate modifications of finite element based discretization methods that preserve positivity of the system solution. We recall the results conserving the positivity and the mass conservation of the stabilized schemes which are introduced in R. Strehl et all. (2010). Under suitable smoothness assumptions on the solution, we prove the existence and the uniqueness of the solution of these stabilized schemes and its solution remains positive under a suitable choice of mesh size and time step. Moreover, we prove results concerning the mass conservation and we prove error estimates in L2 and H1-norm in space and L\infty in time. Numerical experiments for various test problems are performed in order to studythe asymptotic behavior of the error in the L2 and H1-norm in space and L\infty in time. Besides the a priori analysis on the stabilized schemes of R. Strehl et all. (2010), we derive residual-based a posteriori error estimates for the fully discrete schemes of implicit Euler and implicit midpoint rule. Since we have a parabolic system of partial differential equations, the analysis includes the appropriate introduction of an elliptic reconstruction operator, similar to C. Makridakis and R. H. Nochetto (2003). The space discretization uses the standard linear finite element spaces and the time discretization is based on the implicit Euler and in implicit midpoint rule. We also derive a posteriori error estimates in L\infty(L2) and L2(H1)-norm. Numerical experiments for various test problems are performed in order to study the asymptotic behavior of the a posteriori error estimators.

Degree

thesis:*
Grantor dc:publisher
University of Crete (UOC)
Year dc:date
2023

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Pervolianakis, Christos
  • Περβολιανάκης, Χρήστος

Subjects

dc:subject × 20

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/53392

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Pervolianakis, Christos; Περβολιανάκης, Χρήστος. Numerical methods for evolutionary differential equations that maintain the positivity of the solution. University of Crete (UOC), 2023. http://hdl.handle.net/10442/hedi/53392