A direct solver for finite element matrices requiring $O(NlogN)$ memory places

Vejchodský, Tomáš. "A direct solver for finite element matrices requiring $O(N \log N)$ memory places." Applications of Mathematics 2013. Prague: Institute of Mathematics AS CR, 2013. 225-239. <http://eudml.org/doc/287849>.

@inProceedings{Vejchodský2013,
abstract = {We present a method that in certain sense stores the inverse of the stiffness matrix in $O(N\log N)$ memory places, where $N$ is the number of degrees of freedom and hence the matrix size. The setup of this storage format requires $O(N^\{3/2\})$ arithmetic operations. However, once the setup is done, the multiplication of the inverse matrix and a vector can be performed with $O(N\log N)$ operations. This approach applies to the first order finite element discretization of linear elliptic and parabolic problems in triangular domains, but it can be generalized to higher-order elements, variety of problems, and general domains. The method is based on a special hierarchical enumeration of vertices and on a hierarchical elimination of suitable degrees of freedom. Therefore, we call it hierarchical condensation of degrees of freedom.},
author = {Vejchodský, Tomáš},
booktitle = {Applications of Mathematics 2013},
keywords = {sparse direct solver; hierarchical condensation; finite element method; sparse matrices; algorithm},
location = {Prague},
pages = {225-239},
publisher = {Institute of Mathematics AS CR},
title = {A direct solver for finite element matrices requiring $O(N \log N)$ memory places},
url = {http://eudml.org/doc/287849},
year = {2013},
}

TY - CLSWK
AU - Vejchodský, Tomáš
TI - A direct solver for finite element matrices requiring $O(N \log N)$ memory places
T2 - Applications of Mathematics 2013
PY - 2013
CY - Prague
PB - Institute of Mathematics AS CR
SP - 225
EP - 239
AB - We present a method that in certain sense stores the inverse of the stiffness matrix in $O(N\log N)$ memory places, where $N$ is the number of degrees of freedom and hence the matrix size. The setup of this storage format requires $O(N^{3/2})$ arithmetic operations. However, once the setup is done, the multiplication of the inverse matrix and a vector can be performed with $O(N\log N)$ operations. This approach applies to the first order finite element discretization of linear elliptic and parabolic problems in triangular domains, but it can be generalized to higher-order elements, variety of problems, and general domains. The method is based on a special hierarchical enumeration of vertices and on a hierarchical elimination of suitable degrees of freedom. Therefore, we call it hierarchical condensation of degrees of freedom.
KW - sparse direct solver; hierarchical condensation; finite element method; sparse matrices; algorithm
UR - http://eudml.org/doc/287849
ER -