Abstract
In this paper, we extend the pragmatic model for alumina feeding, presented in Johansen et al. Light metals 2022, to treat the associated evolution of bath temperature. The underlying idea of the framework is to reduce its computational overhead to allow for very fast simulation of the dissolved alumina, particle alumina, and bath temperature evolution within aluminium reduction cells. Two main advances presented in this work: (1) a generic method to read velocity data from any CFD simulation to the pragmatic model and (2) inclusion of governing equation to simulate the bath temperature evolution. The physics associated with the various heat loss modes from the bath as well as the temperature dependence of the alumina solubility in the bath is represented in the model. The demonstration cases discussed in this work show that the proposed framework runs much faster than real time (up to 75 times). The average bath and metal temperature in the demonstration case is observed to depend on the feeding pattern used to add alumina into the bath. The proposed framework can be used to observe the spatial and temporal changes of dissolved and particle alumina as well as the bath temperature within a cell.