Continuation of the strategic partnership between Dillinger and MECS: AI-based prediction of steel properties

Can the properties of steel be predicted using artificial intelligence (AI)? The answer lies in the microstructure – the microscopic structure of a material. The microstructure stores important information about the manufacturing process and significantly determines the properties of the steel – and who knows this better than our strategic partner and one of Germany’s largest steel producers? With the increasing complexity of modern high-performance steels and ever stricter tolerances, the challenge of precisely analyzing the microstructure is growing. This is where our long-standing strategic partnership with Dillinger, one of Germany’s leading steel producers, comes in.

By using advanced AI technologies and machine learning (ML), we are opening up completely new possibilities in microstructure analysis. With innovative methods, such as correlative microscopy, we can perform automated, reliable and repeatable investigations that enable unprecedented accuracy and depth in quantification. These technologies allow us not only to analyze microstructures, but also to predict material properties based on this data.

In our current research project, we are working on correlating process parameters and microstructural states using ML. The aim is to approximate the phase transformations from the high-temperature austenite phase to the final microstructural states at room temperature. We are investigating different alloys, forming processes and heat treatments in order to develop a model that predicts the resulting microstructures based on defined process parameters. In addition, this model should be able to determine the optimum parameters for generating a desired microstructure – and this should be directly transferable from laboratory conditions to real production processes.

This project is particularly important with regard to the transformation of the German steel industry. The transition to a hydrogen-based production route requires flexible adjustments to existing steel grades and the development of new, sustainable steel grades. Our research is making an important contribution to the future of steel production!

 

Beitrag teilen