Modeling of fatigue crack growth by neural networks


Abstract views: 42 / PDF downloads: 135

Authors

  • Abdelfetah MOUSSOUNI Department of mechanical engineering, University of Tlemcen, Algeria.
  • Nadjia BENACHOUR Department of physics, University of Tlemcen, Algeria
  • Mustapha BENACHOUR Department of mechanical engineering, University of Tlemcen, Algeria.

Keywords:

Crack Growth, Artificial Neural Networks, Fatigue, Multi Layer Perceptron, Aluminum Alloy 2024 T351

Abstract

Fatigue cracks often occur in means of transport such as aircraft, vehicles and ships, as well as in power generation machinery, gas turbines. The crack growth process is complicated for many reasons, including component geometry, manufacturing defects, and applied load. In this paper, a fatigue crack growth model is developed based on the artificial neural networks (ANN) for the V-notch Charpy specimen. The ANN model mainly depends on the cyclic loading conditions and the properties of the materials on input and output the length of the crack. Experimental data on fatigue crack growth of aluminum alloy 2024 T351 for different load ratios obtained from literature were used for this investigation. The predicted crack length is in good agreement with the experimental data.

Downloads

Published

2023-03-18

How to Cite

MOUSSOUNI, A., BENACHOUR, N., & BENACHOUR, M. (2023). Modeling of fatigue crack growth by neural networks. International Conference on Scientific and Academic Research, 1, 215–219. Retrieved from https://as-proceeding.com/index.php/icsar/article/view/298

Issue

Vol. 1 (2023): Proceeding Book of 2nd International Conference on Scientific and Academic Research ICSAR 2023

Section

Conference Papers