Authors

Ngoc Minh Trang Vu, Laurent Lefèvre, Rémy Nouailletas, Sylvain Brémond

Abstract

A method to generate geometric pseudo-spectral spatial discretization schemes for hyperbolic or parabolic partial differential equations is presented. It applies to the spatial discretization of systems of conservation laws with boundary energy flows and/or distributed source terms. The symplecticity of the proposed spatial discretization schemes is defined with respect to the natural power pairing (form) used to define the port-Hamiltonian formulation for the considered systems of balance equations. The method is applied to the resistive diffusion model, a parabolic equation describing the plasma dynamics in tokamaks. A symplectic Galerkin scheme with Bessel conjugated bases is derived from the usual Galerkin method, using the proposed method. Besides the spectral and energetic properties expected from the symplecticity of the method, it is shown that more accurate approximation of eigenfunctions and reduced numerical oscillations result from this choice of conjugated approximation bases. Finally, the obtained numerical results are validated against experimental data from the tokamak Tore Supra facility.

Keywords

Symplectic spatial integration; Pseudo-spectral methods; Balance equations; Port-Hamiltonian systems; Resistive diffusion equation

Citation

BibTeX

@article{Vu_2017,
  title={{Symplectic spatial integration schemes for systems of balance equations}},
  volume={51},
  ISSN={0959-1524},
  DOI={10.1016/j.jprocont.2016.12.005},
  journal={Journal of Process Control},
  publisher={Elsevier BV},
  author={Vu, Ngoc Minh Trang and Lefèvre, Laurent and Nouailletas, Rémy and Brémond, Sylvain},
  year={2017},
  pages={1--17}
}

Download the bib file

References