In situ X-ray diffraction during heat treatments to guide ai-modelling of phase transitions in steels
| Reference number | |
| Coordinator | Ferritico AB |
| Funding from Vinnova | SEK 499 000 |
| Project duration | November 2020 - June 2022 |
| Status | Completed |
| Venture | Research infrastructure - utilisation and collaboration |
| Call | Industrial pilot projects for utilisation of large-scale infrastructures for neutron and photon based techniques – 2020 |
| End-of-project report | 2020-03800_Ferritico.pdf (pdf, 211 kB) |
Important results from the project
The project purpose was to measure CCT data in situ during real heat treatments for about 40 steel samples using time-resolved synchrotron x-ray diffraction (SXRD). Six different steel grades from three steel suppliers were investigated. The raw SXRD data was extracted and compiled into a CCT data format, making it possible to i) use it to validate the performance of machine learning models that have been developed to enable CCT data simulation ii) merge the high quality CCT data with the current CCT database and relearn the machine learning models to simulate with improved accuracy.
Expected long term effects
The project did successfully fulfil its targets, i.e. to generate high quality CCT data and to use it ,for the purpose of evaluating the quality of the current CCT simulation models and to improve the models through integration of the high quality SXRD data into the CCT database. In addition, the consortium has written a publication focusing on the CCT modelling and the performance evaluation. The publication is currently under review.
Approach and implementation
Steel samples were heated to austenitization temperatures and then cooled continuously with different rates at different stages of the cooling to investigate e.g. the effect of a rate change at low temperatures for the martensitic transformation. Using the high-frequency detector we could monitor the volume fraction evolution of the ferrite, martensite as well as precipitate phases. Furthermore we could follow the evolution of the carbon diffusion by studying the peak shifts for the fcc and bcc phases.