Numerical solution of the Helmholtz Equation using high-order continuous Galerkin methods

We show a continuous Galerkin formulation with high-order polynomials to solve the Helmholtz equation. High-order formulations obtain solutions with less numerical error, and can use curved high-order meshes to approximate the domain. To reduce the computational requirements of the high-order formul...

Descripción completa

Detalles Bibliográficos
Autor: Corella Pérez, Sandra
Tipo de recurso: tesis de maestría
Fecha de publicación:2017
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/106636
Acceso en línea:https://hdl.handle.net/2117/106636
Access Level:acceso abierto
Palabra clave:Difference equations, Partial--Numerical solutions
Numerical modeling
Finite element method
Helmholtz equation
High-order
Static condensation
Equacions diferencials parcials--solucions numèriques
Classificació AMS::65 Numerical analysis::65N Partial differential equations, boundary value problems
Àrees temàtiques de la UPC::Matemàtiques i estadística::Anàlisi numèrica
Descripción
Sumario:We show a continuous Galerkin formulation with high-order polynomials to solve the Helmholtz equation. High-order formulations obtain solutions with less numerical error, and can use curved high-order meshes to approximate the domain. To reduce the computational requirements of the high-order formulation, we apply a static condensation technique. Using this technique, we eliminate a set of unknowns from the global linear system and therefore, we solve a smaller system of equations. Then, we recover the full solution by solving several systems that only involve the unknowns of a single element. In the examples we show that the proposed implementation converges optimally to the analytical solution both for two and three dimensional examples. We also show an application where the mesh is curved in order to better capture the geometry.