Machine learning model of RPV alloy embrittlement

Ensemble neural network model for predicting transition temperature shifts of RPV alloys

Note that this model was developed for academic research purposes and should not be used for activities where certified accuracy is required (e.g., regulatory activities) without further discussion. Current work is ongoing to identify potential issues and update the model accordingly. Therefore, the model may evolve in the future and have newer versions that give somewhat different answers. Any questions/feedback should be sent to: Ryan Jacobs ([email protected]) or Dane Morgan ([email protected])

The model can be run on Google Colab:

If you find this model useful, please cite the original reference:

Jacobs, R., Yamamoto, T., Odette, G. R., Morgan, D. "Predictions and Uncertainty Estimates of Reactor Pressure Vessel Steel Embrittlement Using Machine Learning", Materials and Design, 236 (2023) (https://doi.org/10.1016/j.matdes.2023.112491).