Digitalization of steel heat treatment process
Reference number | |
Coordinator | Ferritico AB |
Funding from Vinnova | SEK 1 396 638 |
Project duration | November 2020 - November 2022 |
Status | Completed |
Venture | Strategic innovation programme for process industrial IT and automation – PiiA |
Call | PiiA: Digitalization of industrial value chains, spring 2020 |
Important results from the project
The development of steels and production processes is mainly carried out by a time and resource consuming trial-and-error process. The design of new materials, further optimization of current materials and processes, including heat treatment, comes with large investments and often constitute significant challenges. The AI software developed in MaLHT will digitalize and make the steel development process more efficient and enable production of steels with novel properties at lower materials and energy costs
Expected long term effects
Novel steels and effective manufacturing processes can be rapidly developed using the MaLHT platform in a virtual environment, instead of the currently used physical environment. This means significantly enhanced effectiveness in the processes and much less resource (raw materials, energy, etc.) consumption. Furthermore, the AI platform will assist optimization of e.g. strength-to-weight performance and extended life time of steel products. Thus, light-weighting and durable steel products will further significantly reduce the raw materials consumption.
Approach and implementation
The project contained three work packages. In WP1, the project made experimental measurements of steel quenching (CCT) to build databases on which machine learning models could be built. The models were implemented in a web application. In WP2, the coordinator Ferritico arranged interviews and workshops with key stakeholders to define requirements for additional simulation software support being needed in the steel heat treatment. According to input from WP2, the project developed simulation support for Isothermal transformation diagrams in WP3 using the WP1 methodology.