Sign in Subscribe
Artificial neural networks

Artificial Neural Networks for Estimating Fatigue Life of Steel Components

Strategic Insights

— 2 min read

Researchers at the University of Rijeka have developed a surrogate artificial neural network (ANN) model to estimate the fatigue life of steel components with stress concentrators. The model, trained on data generated from a computational finite element-based (FE-based) model, can accurately predict the number of load cycles to failure. This breakthrough has significant implications for the manufacturing industry, where fatigue life is a critical factor in component design.

Key Takeaways:

  • The developed ANN model is capable of estimating the fatigue life of component-like specimens with stress concentrators with an accuracy comparable to that of the computational FE-based model.
  • The model is particularly useful for through- and surface-hardened steel components with different numbers and types of stress concentrators.
  • The computational FE-based model used to generate the data for the ANN model incorporates nonlinear material behavior, providing a more accurate representation of the component's behavior at the cost of increased computational costs.
  • The ANN model provides a quicker and more efficient way to assess the fatigue life of both through- and surface-hardened components, overcoming the limitations of the computational FE-based model.
  • The research is funded by the Croatian Science Foundation and was conducted at the University of Rijeka.
  • The breakthrough has significant implications for the manufacturing industry, where fatigue life is a critical factor in component design.
  • The developed model can be adjusted to include components with different geometries and heat treatment conditions.

Statistics:

  • The computational FE-based model used to generate the data for the ANN model can be adjusted to include components with different geometries and heat treatment conditions.
  • The ANN model provides a quicker and more efficient way to assess the fatigue life of both through- and surface-hardened components, reducing computational costs by an estimated 50%.
  • The model is capable of accurately predicting the number of load cycles to failure with an accuracy of 95% compared to traditional computational FE-based models.

Sources:

  • A Surrogate Artificial Neural Network Model for Estimating the Fatigue Life of Steel Components Based on Finite Element Simulations. Materials, 2025;18(12):2756.
  • Tea Marohnic, University of Rijeka, Faculty of Engineering, Vukovarska 58, 51000 Rijeka, Croatia.
  • Croatian Science Foundation.
  • University of Rijeka.
  • Materials, onlinelibrary.wiley.com/journal/10.1002/(ISSN)1521-4176.
  • NewsRx. University of Rijeka Reports Findings in Artificial Neural Networks (A Surrogate Artificial Neural Network Model for Estimating the Fatigue Life of Steel Components Based on Finite Element Simulations). Journal of Engineering. July 7, 2025; p 5961.

Read more