Lehrstuhl für Fertigungstechnologie, Universität Erlangen-Nürnberg

Development of a self-learning transformation and dilatometer model for the virtual process design of hot stamping processes

Project Status: Active


The focus of this research project is the development of transformation models, which can be used as “virtual dilatometer” in numerical process simulation. The self-learning model will be trained by extensive experimental data. The project focuses on the isothermal transformation of undercooled austenite of the steel 22MnB5 into ferrite, pearlite and bainite. This is of high relevance in partial hot stamping processes. The required database will be acquired with the help of a deformation dilatometer and accompanying tests. For this purpose, the transformation points will be analyzed in dependency of the thermomechanical load path as well as the resulting phase composition in combination with the mechanical properties. Furthermore, a learning function will be developed, to enables a fast characterization of other materials. By applying an iterative design of experiments the ideal experimental conditions shall be selected to reduce overall test effort.



Research Groups



    • Horn, A.; Hart-Rawung, T.; Buhl, J.; Bambach, M.; Merklein, M.:
      Investigation of the phase transformation in hot stamping processes with regard to the testing facility.
      In: (Edtr.): WGP-Jahreskongress 2020, 2020, accepted

    Letztes Update: 02.01.2020