Convex analysis and thermodynamics - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Kinetic and Related Models Année : 2013

Convex analysis and thermodynamics

(1) , (2)
1
2
Nelly Point
  • Fonction : Auteur
  • PersonId : 884962
Silvano Erlicher
  • Fonction : Auteur
  • PersonId : 947903

Résumé

Convex analysis is very useful to prove that a material model fulfills the second law of thermodynamics. Dissipation must remains non-negative and an elegant way to ensure this property is to construct an appropriate pseudo-potential of dissipation. In such a case, the corresponding material is said to be a Standard Generalized Material and the flow rules fulfill a normality rule (i.e. the dissipative thermodynamic forces are assumed to belong to an admissible domain and the flow of the corresponding state variables is orthogonal to the boundary of this domain). The sum of the pseudo-potential with its Legendre-Fenchel conjugate fulfills the Fenchel’s inequality and as the actual value of the dual pair forces-flows minimizes this inequality, this can be used as a convergence criterium for numerical applications. Actually, some very commonly used and effective models do not fit into that family of Standard Generalized Materials. A procedure is here proposed which permits to retrieve the normality assumption and to construct a pair of dual pseudo-potentials also for these non-standard material models. This procedure was first presented by the authors for non-associated plasticity. Now it is extended to a large range of mechanical problems.
Fichier principal
Vignette du fichier
Point2013.pdf (134.84 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01157359 , version 1 (02-04-2020)

Licence

Paternité - CC BY 4.0

Identifiants

Citer

Nelly Point, Silvano Erlicher. Convex analysis and thermodynamics. Kinetic and Related Models , 2013, 6 (4), pp.945-954. ⟨10.3934/krm.2013.6.945⟩. ⟨hal-01157359⟩
136 Consultations
276 Téléchargements

Altmetric

Partager

Gmail Facebook Twitter LinkedIn More