Towards numerical prediction of cavitation erosion - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Towards numerical prediction of cavitation erosion

Résumé

This paper is intended to provide a potential basis for a numerical prediction of cavitation erosion damage. The proposed method can be divided into two steps. The first step consists in determining the loading conditions due to cavitation bubble collapses. It is shown that individual pits observed on highly polished metallic samples exposed to cavitation for a relatively small time can be considered as the signature of bubble collapse. By combining pitting tests with an inverse finite-element modelling (FEM) of the material response to a representative impact load, loading conditions can be derived for each individual bubble collapse in terms of stress amplitude (in gigapascals) and radial extent (in micrometres). This step requires characterizing as accurately as possible the properties of thematerial exposed to cavitation. This characterization should include the effect of strain rate, which is known to be high in cavitation erosion (typically of the order of several thousands s21). Nanoindentation techniques as well as compressive tests at high strain rate using, for example, a split Hopkinson pressure bar test system may be used. The second step consists in developing an FEM approach to simulate the material response to the repetitive impact loads determined in step 1. This includes a detailed analysis of the hardening process (isotropic versus kinematic) in order to properly account for fatigue as well as the development of a suitable model of material damage and failure to account for mass loss. Although the whole method is not yet fully operational, promising results are presented that show that such a numerical method might be, in the long term, an alternative to correlative techniques used so far for cavitation erosion prediction.

Mots clés

Dates et versions

hal-01134221 , version 1 (23-03-2015)

Licence

Paternité

Identifiants

Citer

Jean-Pierre Franc. Towards numerical prediction of cavitation erosion. Amazing (Cavitation) Bubbles: Great Potentials & Challenges, The Royal Society, Nov 2014, London, United Kingdom. ⟨10.1098/rsfs.2015.0013⟩. ⟨hal-01134221⟩

Collections

UGA CNRS
46 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More