Employing machine learning methods in an industry context at Otto Fuchs KG
Michele Matsuo, Mauricio Santaella, Otto Fuchs, Germany
In the digitalization process of the metallurgical industry, the convergence of materials science, simulation, and data-driven plays an important role. As the industry embraces digital transformation, these disciplines enable breakthrough advancements and drive innovation. The objective of this presentation is to showcase work examples of the Advanced Engineering Group of Otto Fuchs KG that highlight the integration of these aspects. The focus will be on discussing three main topics: machine learning (ML) surrogates for finite element (FE) simulation, materials-process-properties design simulation via optimization algorithms, and deployment of machine learning models in the industrial context.
The first topic (surrogates for simulation) demonstrates the utilization of regression models to save computational and time resources in the context of forging FE simulations. By employing these methods, the industry can significantly reduce the computational complexity associated with expensive simulation-based processes. The second topic (materials-process-properties design simulation) showcases the application of optimization algorithms in simulation models to determine
optimal production parameter combination and material constant values. These algorithms are exemplified respective through their use in defects mitigation of ring rolled parts, and in calibration
of the phenomenological microstructure model Johnson-Mehl-Avrami-Kolmogorov. The final part of the discussions (ML-FE models deployment) exemplifies the application of the mentioned works to the industrial production environment. This is achieved through the development of intuitive user interfaces to facilitate the final work usage. In summary, this presentation provides an overview of the approaches and challenges involved on the models‘ development from data acquisition, modeling, and integration of different departments of Otto Fuchs KG. Through these insights, the aim is to highlight the advantages that can be achieved in terms of efficiency, product quality, and technological research as the company progresses in the context of industry 4.0.