The submitted article deals with the use of physical and numerical modelling to study the process of the steel flow in an asymmetric five-strand tundish that continuously casts steel. For the purposes of physical modelling, a 1:4-scale plexiglass model was used as the operating tundish, and for numerical modelling, the geometry of the operating tundish was created on a 1:1 scale. A model liquid (water) was used in the physical modelling of the melt flow process, while liquid steel was used as the standard flowing medium in the numerical modelling. We assessed the relevant operating parameters influencing the characteristics of the flow of the bath in the tundish—the shape of the turbulence inhibitor, the position of the ladle shroud in relation to the turbulence inhibitor and the distance between the ladle shroud orifice and the bottom of the turbulence inhibitor. The preliminary results show that optimal steel flow characteristic results are achieved by using the TI3-C configuration. The results from both modelling methods achieved the same characteristics, therefore verifying the results of each other and demonstrating that when taken together, the results of physical and numerical modelling can be considered sufficiently informative.

When describing the behaviour and modelling of real systems, which are characterized by considerable complexity, great difficulty, and often the impossibility of their formal mathematical description, and whose operational monitoring and measurement are difficult, conventional analytical–statistical models run into the limits of their use. The application of these models leads to necessary simplifications, which cause insufficient adequacy of the resulting mathematical description. In such cases, it is appropriate for modelling to use the methods brought by a new scientific discipline—artificial intelligence. Artificial intelligence provides very promising tools for describing and controlling complex systems. The method of neural networks was chosen for the analysis of the lifetime of the teeming ladle. Artificial neural networks are mathematical models that approximate non-linear functions of an arbitrary waveform. The advantage of neural networks is their ability to generalize the dependencies between individual quantities by learning the presented patterns. This property of a neural network is referred to as generalization. Their use is suitable for processing complex problems where the dependencies between individual quantities are not exactly known.

The continuous steel casting process must simultaneously meet the requirements for production performance, quality and safety against breakouts. Knowing the thickness of the solidified shell, particularly at the exit of the mould, is useful for the casting process control and breakout prevention. Shell thickness is difficult to measure during casting; in practice, it is predicted by indirect methods and models. But after undesired rupture of the shell and leakage of the liquid steel, it is possible to measure the shell thickness directly. This article is focused on the problem of the growth and measurement of the solid shell obtained after the breakout of a round block with a diameter of 410 mm. An original methodology was developed in which a surface mesh of points was created from the individual scanned parts of the block using a 3D laser scanner. Research has shown differences of up to 6 mm between the maximum and minimum shell thickness at the mould exit. A regression function of the average shell thickness on time was found. The results of the real shell growth were further used for the verification of the original numerical model of cooling and solidification of the round block.