A local thermomechanical model of friction for computing structures under extreme loading conditions - Archive ouverte HAL
Article Dans Une Revue International Journal for Numerical Methods in Engineering Année : 2011

A local thermomechanical model of friction for computing structures under extreme loading conditions

Résumé

A multiscale thermomechanical model of friction is proposed for metallic interfaces submitted to extreme loading conditions with large sliding velocities ( [v] > 100 ms −1) and high contact pressures (P > 1 GPa). This model decomposes the friction problem into a global multidimensional structural problem and a local interface subproblem. At global scale, the structure behavior is governed by an elastoplastic model in large strains, and thermal conduction is neglected. The local model describes the micrometric scale and the underlying thermomechanical mechanisms: mechanical hardening, frictional heating, and plastic work. Using dimensional and asymptotic analysis, the corresponding set of equations at local scale is simplified and reduces to a one-dimensional quasistatic thermoelastoplastic friction model. It is solved locally at each global point by a finite differences subgrid model using a nonlinear time-implicit solver. This solver is coupled to a time-explicit Lagrangian hydrocode used at the global structural scale. The coupling strategy between the solvers is force-based: the frictional stress is evaluated at the local scale by our multiscale friction model and transferred to the global model that can then predict local velocity corrections. This coupling strategy has been successfully tested on real-life cases.

Dates et versions

hal-00655515 , version 1 (29-12-2011)

Identifiants

Citer

Patrick Le Tallec, Franck Dambakazi, Jean-Philippe Perlat. A local thermomechanical model of friction for computing structures under extreme loading conditions. International Journal for Numerical Methods in Engineering, 2011, ⟨10.1002/nme.3252⟩. ⟨hal-00655515⟩
132 Consultations
0 Téléchargements

Altmetric

Partager

More