Development of a Modelling Technique for Non-Equilibrium Metallurgical Processes.

Scandinavian Journal of Metallurgy,  28 (1999),  285
Language: English


Development of a Modelling Technique for Non-Equilibrium Metallurgical Processes

| Modigell M. | Traebert A. |
Institut für Verfahrenstechnik, RWTH Aachen, Germany

| Monheim P. |
Mannesmann Demag Metallurgie, Duisburg, Germany

| Hack K. |
GTT-Technologies, Herzogenrath, Germany

Keywords: | chemapp | process simulation | steelmaking | promosys | slag |


High-temperature processes with fluid phases involved usually exhibit a behaviour close to thermochemical equilibrium. This can be attributed to high reaction rates and high mass respectively heat transfer rates in turbulent flow. Therefore a simulation of the process assuming thermochemical equilibrium is often appropriate. However, a more detailed analysis of fluid/fluid or fluid/solid technical processes shows deviations from equilibrium due to considerably limited heat and mass transfer, caused e.g. by incomplete mixing. A modeling technique is developed to enable the calculation of complex non-equilibrium phenomena. With this technique, a process is simulated using an appropriate arrangement of simple, local equilibrium reactors coupled by defined heat and mass exchange.

As an example, the simulation of a LD converter process was undertaken with focus on non-equilibrium phenomena. The LD process was chosen as it is widely investigated, which proved helpful for an evaluation of the applied technique. In modeling great emphasis was placed on a state of the art representation of the thermochemistry of the material systems involved. A newly developed modular process balancing system, ProMoSys®, was successfully applied incorporating the thermochemical applications library ChemApp, which is well known for its abilities to treat the chemistry of metallurgical processes. The validation of the simulation model shows good agreement between reported and calculated values for species concentration development over time. Real-process phenomena can be predicted.