Behind every accurate model and innovative solution is a team turning complex material behaviour into practical engineering insights. In this series of articles, we explore the important role Ex Mente’s Material Science team plays in helping us to Advance Through Insight. In this second article we discuss how at Ex Mente we do not necessarily see slag as a necessary evil byproduct of smelting. We know that slag is a critical control lever for your entire operation. It governs heat transfer, protects your refractories, and ultimately dictates the purity of your metal. But to control the slag, you must first understand the invisible architecture holding it together: its structure. This is where our materials science team steps in. We provide foundational insights that support purpose-built computational models by our Multiphysics team.

Why does structure matter? The physical properties of molten slag, viscosity, diffusivity, electrical and thermal conductivity, are not arbitrary numbers. They are the direct result of the slag’s internal “web”, specifically the degree of polymerisation of the silicate network.

Imagine this network as a series of interconnected chains.

1. High Polymerisation: In a highly polymerised network, silicate tetrahedra form long, interconnected chains through bridging oxygens. This restricts movement, resulting in higher viscosity (resistance to flow) and lower electrical conductivity (hindered ion movement).
2. Low Polymerisation: When we introduce network modifiers like Calcium Oxide (CaO) or Magnesium Oxide (MgO), we break these chains by increasing the fraction of free oxygens, loosening the network. The result? A fluid slag with lower viscosity and higher electrical conductivity.

By understanding these microscopic interactions, we can estimate how your slag will behave under different temperatures and compositions, moving beyond guesswork to accurate control.

To deliver these insights, we use dedicated tools, developed in-house, that allow us to quantify slag structure and the effects it has on its material properties. For developing these tools, we use methods like molecular dynamics that allows us to simulate atomic movements, giving us a near fundamental understanding of how ions interact within slag. We also offer an open-source Python tool called auxi-mpp, which provides a collection of material physical property models sourced from published literature.

The materials science team supports the Multiphysics modelers so they can support you. This ensures that the solution you receive for slag foaming, energy optimisation, or emissions control, is built on a foundation of rigorous, validated material science.