Search

The similarity between steel ladles and hot-water models regarding natural convection phenomena has been systematically analysed through examination of the numerical solutions of turbulent Navier-Stokes partial differential equations governing the phenomena in question. The numerical solutions have been obtained by using a computational fluid dynamics (CFD) simulation method. Key similarity criteria for non-isothermal physical modelling of steel ladles with hot-water models have been derived as Fr m =Fr p and (βΔT) m =(βΔT) p where the subscripts m and p stand for the water model and the prototype steel ladle, respectively. Accordingly, appropriate conditions fulfilling the above criteria, such as model size, water temperature, time scale factor and the scale factor of boundary heat loss fluxes, have been proposed and discussed. As a result, water models with geometry scales between 1/5 and 1/3 and using hot-water of temperature higher than 45°C are appropriate for simulating natural convection phenomena in steel ladles.

A 1/4-scale hot-water model of industrial 107-tonne steel ladles was established in the laboratory. With this physical model, thermal stratification phenomena due to natural convection in steel ladles during the holding period before casting were investigated. By controlling the cooling intensity of the water model to correspond to the heat loss rate of steel ladles, which is governed by dimensionless numbers Fr and βΔT, temperature distributions in the water model can simulate those in the steel ladles. Consequently, the temperature profile in the hot-water bath in the model can be used to deduce the thermal stratification phenomena in liquid steel bath in the ladles. In addition, mathematical simulations on fluid flow and heat transfer both in the water model and in the prototype steel ladle were performed using a computational fluid dynamics (CFD) numerical method. The CFD model was validated against temperatures measured in the water model. Comparisons between mathematically simulated temperature profiles in the prototype steel ladle and those physically simulated by scaling-up the measured temperature profiles in the water model showed a good agreement. Therefore, it can be concluded that, as long as accurate heat loss information is known, it is feasible to use a 1/4-scale water model to non-isothermally simulate fluid flow and heat transfer in steel ladles during the holding period before casting.

Based on a prototype installation, the mixing time and the fluid flow pattern of CAS (Composition Adjustment by Sealed argon bubbling) were studied by a water model analog simulation. The decisive ...