HATFLAT

Holistic Assistance for Cross-Process Analysis and Prediction of Strip and Plate Flatness

In the realm of flat steel production, ensuring the flatness of steel strips and plates holds immense importance as it directly impacts the safety and stability of the manufacturing process and ultimately defines the quality of the end product. While local flatness is effectively controlled during the rolling phase, numerous steel plants encounter notable flatness issues with thin and/or high-strength steel products during subsequent processes and post-production stages.

The HatFlat project aims to investigate the interconnected influences on the development of flatness defects across various stages of the production process. Leveraging the power of digital twins and a unique amalgamation of classical first-principle models and machine learning techniques, this endeavor seeks to gain a comprehensive understanding of the phenomenon while pinpointing the most critical process parameters.

The ultimate goal of the project is to develop an assistance software tool that will empower operators to enhance the flatness of the steel products. By leveraging the insights derived from the holistic understanding attained through the integration of digital twins, classical models, and machine learning methods, operators will be equipped with valuable support in their efforts to optimize flatness during production.