Artificial Neural Network-Based Prediction of Functional Fatigue in Shape Memory Alloys

Strategic Insights -

Researchers at the PSG Institute of Technology and Applied Research have made a groundbreaking discovery in the field of artificial neural networks, successfully developing a model to predict the functional fatigue behavior of shape memory alloys (SMAs). The novel approach utilizes a feed-forward backpropagation artificial neural network (ANN) to capture the complex fatigue response of SMAs under partial thermal cycling and constant stress. This innovative study aims to address the long-standing issue of functional fatigue in SMAs, which exhibit phase transformations during cyclic loading, leading to degradation in recovery strain and thermal hysteresis.

Key Takeaways:

  • The study proposes an artificial neural network (ANN) approach to model the functional fatigue behavior of NiTi SMA under partial thermal cycling.
  • The ANN achieved a prediction accuracy of 94.3%, indicating its reliability in capturing the complex fatigue response of SMAs.
  • The ANN model consisted of two inputs (current and number of cycles) and four outputs (recovery strain, permanent strain, upper cycle temperature, and strain accumulation per cycle).
  • The study utilized a feed-forward backpropagation ANN with a high prediction accuracy, highlighting the potential of this approach for predicting functional fatigue in SMAs.
  • The research was carried out at the PSG Institute of Technology and Applied Research, with the lead author being G. Swaminathan from the Department of Mechanical Engineering.
  • The study includes additional authors, S. H. Adarsh, M. Raju, K. Senthilkumar, and T. Senthil Muthu Kumar, all affiliated with PSG Institute of Technology and Applied Research.
  • The research has significant implications for the development of SMAs in various applications, including aerospace and biomedical engineering.

Statistics:

  • The ANN achieved a prediction accuracy of 94.3%.
  • The ANN model consisted of two inputs (current and number of cycles) and four outputs (recovery strain, permanent strain, upper cycle temperature, and strain accumulation per cycle).
  • The study utilized a feed-forward backpropagation ANN.
  • The research was published in the journal Discover Materials, Volume 5, Issue 1, 2025.
  • The article is available online at https://doi-org.sdpl.idm.oclc.org/10.1007/s43939-025-00396-3.

Sources:

  • NewsRx. Studies from PSG Institute of Technology and Applied Research in the Area of Artificial Neural Networks Described (Artificial neural network-based prediction of functional fatigue behaviour of an NiTi shape memory alloy). Journal of Engineering. October 27, 2025; p 4245.
  • G. Swaminathan et al. Artificial neural network-based prediction of functional fatigue behaviour of an NiTi shape memory alloy. Discover Materials, 2025, 5(1):1-12.