Identification of insulating materials thermal properties by inverse method using reduced order model
Résumé
The analytical models used by hot-wire type probes are no longer suitable for the characterization of insulating materials. This paper proposes a solution, which is based on a numerical reduced order technique named AROMM (Amalgamated Reduced Order Modal Method) coupled to an inverse procedure. We demonstrate that a single reduced order model provides precise results for insulated materials characterized by different thermal properties, with the advantage of computing 250 times faster than a classical finite element modelization. Such model is then tested through multiple scenarios in order to evaluate the accuracy of the proposed methodology. Results show the importance of the sensitivity of the measurement regarding the sought parameters, which later intervene during the identification process. A statistical study allows us to access a satisfying confidence interval for a common measurement noise. At last, a study on the influence of an eventual thermal contact resistance is conducted.
Origine : Fichiers produits par l'(les) auteur(s)