Un algorithme de Gradient Conjugué Projeté préconditionné pour la résolution de problèmes unilatéraux

Nicolas Tardieu; Fabien Youbissi; Eric Chamberland

doi:10.1016/j.crme.2008.10.007

Un algorithme de Gradient Conjugué Projeté préconditionné pour la résolution de problèmes unilatéraux

Translated title of the contribution: A preconditioned Projected Conjugate Gradient for the solution of unilateral problems

Nicolas Tardieu
, Fabien Youbissi
, Eric Chamberland

LaMSID/CNRS (UMR 2932)
Rennes et Université Laval (Canada)

Research output: Contribution to journal › Article › peer-review

Abstract

The Projected Conjugate Gradient method comes from the field of convex optimization, where it is used to solve minimization problems with linear inequality constraints. Thus, it is well suited for unilateral contact. Utilized on the dual form of the contact problem and associated to a Dirichlet preconditioner, it exhibits a performing behaviour, especially as the mesh size increases. To cite this article: N. Tardieu et al., C. R. Mecanique 336 (2008).

Translated title of the contribution	A preconditioned Projected Conjugate Gradient for the solution of unilateral problems
Original language	French
Pages (from-to)	840-845
Number of pages	6
Journal	Comptes Rendus - Mecanique
Volume	336
Issue number	11-12
DOIs	https://doi.org/10.1016/j.crme.2008.10.007
Publication status	Published - 1 Nov 2008
Externally published	Yes

Access to Document

10.1016/j.crme.2008.10.007

Cite this

@article{a2cd152af6204059ae0d1262953e3452,

title = "Un algorithme de Gradient Conjugu{\'e} Projet{\'e} pr{\'e}conditionn{\'e} pour la r{\'e}solution de probl{\`e}mes unilat{\'e}raux",

abstract = "The Projected Conjugate Gradient method comes from the field of convex optimization, where it is used to solve minimization problems with linear inequality constraints. Thus, it is well suited for unilateral contact. Utilized on the dual form of the contact problem and associated to a Dirichlet preconditioner, it exhibits a performing behaviour, especially as the mesh size increases. To cite this article: N. Tardieu et al., C. R. Mecanique 336 (2008).",

keywords = "Conjugate Gradient, Contact, Preconditioner",

author = "Nicolas Tardieu and Fabien Youbissi and Eric Chamberland",

year = "2008",

month = nov,

day = "1",

doi = "10.1016/j.crme.2008.10.007",

language = "Fran{\c c}ais",

volume = "336",

pages = "840--845",

journal = "Comptes Rendus - Mecanique",

issn = "1631-0721",

number = "11-12",

}

TY - JOUR

T1 - Un algorithme de Gradient Conjugué Projeté préconditionné pour la résolution de problèmes unilatéraux

AU - Tardieu, Nicolas

AU - Youbissi, Fabien

AU - Chamberland, Eric

PY - 2008/11/1

Y1 - 2008/11/1

N2 - The Projected Conjugate Gradient method comes from the field of convex optimization, where it is used to solve minimization problems with linear inequality constraints. Thus, it is well suited for unilateral contact. Utilized on the dual form of the contact problem and associated to a Dirichlet preconditioner, it exhibits a performing behaviour, especially as the mesh size increases. To cite this article: N. Tardieu et al., C. R. Mecanique 336 (2008).

AB - The Projected Conjugate Gradient method comes from the field of convex optimization, where it is used to solve minimization problems with linear inequality constraints. Thus, it is well suited for unilateral contact. Utilized on the dual form of the contact problem and associated to a Dirichlet preconditioner, it exhibits a performing behaviour, especially as the mesh size increases. To cite this article: N. Tardieu et al., C. R. Mecanique 336 (2008).

KW - Conjugate Gradient

KW - Contact

KW - Preconditioner

U2 - 10.1016/j.crme.2008.10.007

DO - 10.1016/j.crme.2008.10.007

M3 - Article

AN - SCOPUS:57049157620

SN - 1631-0721

VL - 336

SP - 840

EP - 845

JO - Comptes Rendus - Mecanique

JF - Comptes Rendus - Mecanique

IS - 11-12

ER -

Un algorithme de Gradient Conjugué Projeté préconditionné pour la résolution de problèmes unilatéraux

Abstract

Access to Document

Fingerprint

Cite this