The purpose of unit processes in primary and secondary steelmaking is to refine the steel so that the specified steel composition and temperature are achieved and the steel melt is ready for casting.The control of decarburization, removal of dissolved gases, temperature and inclusions is of particular importance. In this regard, the Finnish steel industry stands already at high level. During the project a novel and highly efficient approach was developed for modelling steelmaking process chain.
During the project a novel and highly efficient approach was developed for modelling steelmaking process chain; Argon-Oxygen Decarburization (AOD), Composition Adjustment by Sealed Argon Bubbling - Oxygen Blowing (CAS-OB), vacuum tank degassing processes (VD). Moreover, a fundamental mathematical model for AOD process, named “AOD Converter Process Simulator”, has been developed and validated. AOD Converter Process Simulator can be used for off-line simulation at Outokumpu Stainless, Tornio. Off-line simulation can be employed for developing new blowing practices and prediction of decarburization behavior with varying steel composition.
- Plane geometry is employed for reactions between steel surface and top-blown gas
- Bubble geometry describes reactions between steel and spherical gas bubbles
- Droplet geometry can be used both for reactions between steel droplets and slag and reactions between slag droplets and steel.
By a smart combination of these, all studied processes could be modeled effectively. Our approach was to first develop modular based models for the small scale phenomena and then combine these to simulate large-scale phenomena. The modules of the unit processes can be combined to form the actual metallurgical processes.
Employing this approach we were able to have very realistic descriptions of the actual processes while achieving a simulation efficiency that is close to real time – an essential feature in on-line simulators. All the models have been validated successfully with experimental data and published in international scientific literature.
The driving force for this research was the need to better understanding the steel making processes in order to improve the processes, material and energy efficiency and to simulate production of new steel grades. There were several approaches already available – such as computational fluid dynamics (CFD) – but most of them were either far too slow or inaccurate for engineering use. Steelmaking industry needs accurate tools that operate close to on-line and to achieve these, a new approach was required. From an industrial point of view, it is extremely expensive to carry out full-scale tests and it would be very useful if processes can be first studied by modelling. Our industrial objective was to have a model that can be used to make the treatment without needing to have sampling during the treatment.
The new methodology opens new possibilities for better understanding the principles of steelmaking and even for developing totally new concepts. With the new tools it is possible to reduce the use of materials, gases and energy and to create a new basis for more sustainable steelmaking.
Models are already successively used to minimize use of energy, raw materials and gases, for example:
- 20% lower nitrogen levels in VD process, Ruukki Metals Oy
- 50% energy saving potential in chemical heating
- 23% shorter AOD processing time, Outokumpu Stainless Oy
- Approximately 30% increase in lifetime of the bottom stirring and increased total lining lifetimes of BOF converters was achieved at Ruukki Metals Oy
Much better understanding of the CAS-OB process, more efficient processing, Ruukki Metals Oy
- Higher strength of steels - less material required
- Corrosion resistance - no painting/zink required
- Better machining and welding properties - less stages in component manufacturing
- Improved inner quality of slabs, the possibility of manufacturing new special steels and new thickness ranges
The process models can be operated via user-interface, which not only makes them more user friendly and accessible, but also enables fast simulation of different practices. In addition, objective is to have these models connected to process control and automation during the next 5 year period.
Aalto University, Ruukki, Outokumpu, University of Oulu, VTT