A First Proposal for Fatigue Life Prediction of Carbon Black Filled Natural Rubber at Different Temperatures With an Artificial Neural Network
Résumé
In this paper, the fatigue life prediction of a carbon black filled natural rubber was carried out using an artificial neural network. A wide range of loading ratios were considered under relaxing and non‐relaxing loading conditions in order to fully describe the fatigue response, especially the lifetime reinforcement. As the reinforcement is generally assumed to be due to a highly thermosensitive phenomenon, namely, the strain‐induced crystallisation, tests were performed at three different temperatures. The lifetime criterion was defined in terms of stiffness loss. An original pedagogically based training approach allowed us to satisfactorily predict the fatigue response from limited experimental databases. The method provides Haigh diagrams at temperatures not previously tested and predicts well the reinforcement decrease with the temperature, then saving a considerable number of tests. The results show that the approach finds promising results and opens a new way for reducing fatigue tests of elastomers.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |