The need for increasingly accurate data regarding physical slag propertiesPhysical slag properties is clear and will be extremely valuable in making the simulations quantitative and more realistic. Some of these properties are difficult to determine, which is inherent to the high temperatures in pyrometallurgyPyrometallurgy. At the Sustainable Materials Science research group of Ghent UniversityUniversities, we focus on determining electrical conductivitiesElectrical conductivity, viscositiesViscosity, and surface tensions of slagsSlag via a combined experimental modelingModelling method. These values are used to set up and optimizeOptimize structure–property relationships, which are essential for obtaining accurate processes and predictive digital twinsDigital twin for industry. Besides the experimental work at UGent, molecular dynamicsMolecular dynamics (MD) simulations are used to investigate oxidic slagOxidic slags physical properties. With MD simulations, one can follow the time evolution of a system (consisting of atoms or molecules) by integrating Newton's equation of motion. The input for an MD simulation mainly consists of a force field as well as initial positions, velocities, mass, and charge of each atom. The determination of the transport properties in existing literature of molecular dynamicsMolecular dynamics simulations in oxidic systems is typically based on incorrect assumptions. For example, to determine the ionic part of the electrical conductivityElectrical conductivity, the frequently-used Einstein–Stokes equation fails to take into account all ionic types and neglects correlated motion. It is better to use the more general Einstein relationship and most optimal to also make a critical comparison with experimental values, both from literature and from our own lab.