Prediction of the diffusible hydrogen concentration after electrochemical charging utilizing artificial intelligence


Sivesoglu A., Li Y., BAL B.

Materials Research Express, cilt.12, sa.9, 2025 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 12 Sayı: 9
  • Basım Tarihi: 2025
  • Doi Numarası: 10.1088/2053-1591/adf1dd
  • Dergi Adı: Materials Research Express
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Chemical Abstracts Core, Communication Abstracts, Compendex, INSPEC, Metadex, Directory of Open Access Journals, Civil Engineering Abstracts
  • Anahtar Kelimeler: artificial intelligence, FCC metals, hydrogen embrittlement, weight initialization
  • Abdullah Gül Üniversitesi Adresli: Evet

Özet

The concentration of diffusible hydrogen in a material is of high importance as it helps to predict the hydrogen embrittlement effect in the material, and the amount of mechanical properties’ degradation after reaching a critical concentration. Despite that, a simple experimental setup is not available to measure hydrogen concentration at service. In this paper, a multi-layer perceptron (MLP) model is developed using weight initialization, which can estimate the diffusible hydrogen concentration of Face-Centred-Cubic (FCC) metals after electrochemical charging. The input properties of the model include the electrochemical charging parameters of current density, temperature, and charging time as well as the grain size of the specimen. The MLP model with and without the weight initialization was validated and tested with unseen test dataset. The model in both cases showed an excellent predictive performance with a higher accuracy and faster convergence when using weight initialization. A linear correlation of 89% between the experimental and predicted hydrogen concentration was observed. This demonstrates that for the family of FCC metals under electrochemical charging, the estimation of diffusible hydrogen concentration is a feasible path for material safety design analysis.