Predicting Stress-Strain Curves of Irradiated Eurofer 97 from Indentation Curves using Deep Learning

Pragalbh Vashishtha, Arun Rajkumar, Hariprasad Gopalan

Abstract:

In materials science, understanding the mechanical behavior of irradiated materials is crucial for designing and developing advanced nuclear systems. Eurofer 97, a reduced-activation ferritic/martensitic steel, is a widely used structural material in such a scenario. However, accurately predicting the stress-strain curves of irradiated Eurofer 97 presents a significant challenge due to the complex interaction of various factors.
In this study, we propose a novel approach to predict stress-strain curves of irradiated Eurofer 97 using deep learning techniques. Specifically, we employ an LSTM-based (Long Short-Term Memory) model, a recurrent neural network well-suited for sequence prediction tasks.
The input data for our model consists of indentation curves, commonly used to extract material properties such as hardness and elastic modulus. By training the LSTM model on a dataset comprising indentation curves and corresponding stress-strain curves of Eurofer 97, we aim to capture the underlying relationships between indentation and mechanical behavior. The model learns to recognize patterns in the indentation data and subsequently generates predictions of stress-strain curves for irradiated Eurofer 97.