The structure of a ferromagnetic material can be characterized by its microstructural and magnetic features. When a ferromagnetic material is placed in an external magnetic field, its small magnetic domains start to move. This movement causes a measurable signal called Barkhausen noise (BN). BarFume aims to improve the understanding of how the microstructural features of the sample affect the domain wall movement, i.e. BN, and how this information could be used to improve the accuracy of BN-based quality control methods. In addition, novel machine learning techniques will be developed to automatically classify materials microstructure using BN as input. BarFume is an interdisciplinary project that combines materials characterization to BN measurements, micromagnetic modelling, and advanced statistical analysis. The utilization of the research results is obvious in collaboration with a key Finnish industrial partner.
Goal
- to improve the understanding of how the microstructural features are manifested in the properties of the ensuing Barkhausen noise
- to use this information to improve the accuracy of Barkhausen noise-based quality control methods
- to use new machine learning techniques to construct an automatic classifier of materials microstructures using Barkhausen noise as input
Funding source
Partners
Stresstech Oy, Finland
Technical University of Denmark / DTU Nanolab, Denmark
Italian National Institute of Metrology Research, Italy
