AI‑enhanced metallurgical modeling of microstructure and material properties

The X‑Pact® Microstructure Property Model (MPM‑AI) combines metallurgical expertise with data‑driven AI to accurately predict yield strength, tensile strength, grain size, and microstructure - starting from day one, even without extensive historical data. By evaluating upstream conditions and real‑time furnace behavior, MPM‑AI delivers reliable property forecasts for every strip entering the annealing line.

With integrated online calculations and offline simulation, users can analyze process scenarios, adjust parameters, and understand their impact on final material properties. All predictions are securely stored and can be easily visualized or shared through flexible data interfaces.

AI‑enhanced microstructure prediction in continuous annealing

Customer challenges addressed

  • Lack of accurate real‑time prediction of material properties
  • High effort for grade and alloy development, with delays due to lab testing
  • Missing link between process parameters, microstructure, and properties
  • Limited understanding of upstream process influence on final results
  • High downgrade rates, limited ability to simulate process scenarios for optimization
  • Insufficient plant‑specific model calibration due to limited sample data

Key features

  • Reduced downgrades through early deviation detection and improved product quality
  • Immediate high‑accuracy property prediction from day one
  • Online and offline calculation modes for full process transparency, incl. upstream influence
  • Scenario simulation for process optimization and material development; grain size forecasts within defined tolerance ranges
  • Significant reduction of lab testing effort through continuous calculation

Highlights

  • Hybrid Model

    Combination of metallurgical and AI modeling

    MPM‑AI combines a proven metallurgical model with advanced data‑driven algorithms. It ensures accurate material-property predictions from day one, even without historical production data. Thanks to the AI, the system continuously improves its prediction accuracy, ensuring reliability, adaptability, and immediate value for customers.  

  • Monitoring and Predictor Mode

    On-time and predictive calculation

    With its Monitoring and Predictor Modes, MPM‑AI calculates yield strength, tensile strength, and microstructure characteristics on time and even before production begins, respectively. Operators gain early visibility into expected outcomes, enabling faster decisions and more stable production.

  • Simulation Mode

    Powerful offline simulation for optimization

    The integrated Simulation Mode provides a virtual environment to experiment with process parameters without affecting live operations. Users can explore “what-if” scenarios, optimize processes, accelerate material development, and troubleshoot variations - significantly improving product quality and reducing downgrades.