Prediction of static tensile properties and reverse design of microstructure for dual-phase steel based on deep learning

Key Points

• Tensile properties of dual-phase steel can be accurately predicted using deep learning techniques, with strong implications for material performance.
• The deep learning model demonstrates a correlation coefficient exceeding 0.90, indicating high accuracy in predictions about tensile strength and yield stress.
• Analysis applies a reverse design approach to optimize microstructural characteristics, enabling more efficient material production processes.
• These findings support the potential for using AI in advancing the design and implementation of advanced materials, necessitating further exploration.