eprintid: 24230
rev_number: 27
eprint_status: archive
userid: 3678
dir: disk0/00/02/42/30
datestamp: 2018-03-28 09:05:29
lastmod: 2018-04-16 11:01:58
status_changed: 2018-03-28 09:05:29
type: masterThesis
metadata_visibility: show
creators_name: Mehta, Saurabh
title: Matrix Free approach for Raviart-Thomas anisotropic Tensor Product Finite Elements
subjects: ddc-004
subjects: ddc-510
divisions: i-110001
adv_faculty: af-11
cterms_swd: Finite element
cterms_swd: Mixed Finite element
cterms_swd: Raviart-Thomas Finite Element
cterms_swd: high performance computing
cterms_swd: matrixfree
abstract: In this thesis, we develop techniques for fast, efficient cell based operator evaluation of Raviart-Thomas finite elements and thereby extend the existing implementation of matrix free framework in deal.ii. The extensions allow the framework to also support anisotropic vector-valued finite elements in addition to existing isotropic Lagrangian finite elements. Tests on finite element operators in divergence conforming spaces for Mixed diffusion equation and Stokes equations show that the method is sufficiently accurate (relative error 10e-16) and for RT2 is already twice as fast as corresponding sparse matrix based solution.
date: 2018
id_scheme: DOI
id_number: 10.11588/heidok.00024230
ppn_swb: 1654787787
own_urn: urn:nbn:de:bsz:16-heidok-242308
date_accepted: 2018-03-27
advisor: HASH(0x561a62918818)
language: eng
bibsort: MEHTASAURAMATRIXFREE2018
full_text_status: public
place_of_pub: Heidelberg, Germany
thesis_type: Master
citation:   Mehta, Saurabh  (2018) Matrix Free approach for Raviart-Thomas anisotropic Tensor Product Finite Elements.  [Master's thesis]     
document_url: https://archiv.ub.uni-heidelberg.de/volltextserver/24230/8/MehSa_Thesis-1.pdf
document_url: https://archiv.ub.uni-heidelberg.de/volltextserver/24230/14/MehSa_Thesis_old_version.pdf