Computational modelling tools have been developed in the last 20 years with the final aim of investigating, designing and optimizing thermal plasma processes for material treatment, such as nanoparticle synthesis and micrometric powders spheroidization in RF and DC plasma systems, plasma spray through DC non-transferred arc torches, plasma cutting and welding in DC, AC and pulsed transferred arcs, plasma waste treatment in AC and DC arc furnaces.
Modelling tools developed at UNIBO are based on a commercial CFD code in which several physical modules have been implemented, which are able to take into account most relevant physical phenomena occurring in thermal plasma systems for material treatment:
- Plasma thermo-fluid dynamics considering fully 3D realistic geometries of the plasma system and turbulence through Reynolds-Averaged Navier-Stokes and Large Eddy Simulations models, relying on accurate computation of plasma thermodynamic and transport properties for custom working gas mixtures, under steady-state or transient conditions;
- Electro-magnetic fields coupled with plasma dynamics;
- Plasma radiation and re-absorption;
- Micrometric solid/liquid/gas precursors thermal histories, trajectories and evaporation, considering their interaction with plasma discharge;
- Transport, diffusion and demixing of gas mixtures and vapour phases;
- Multi-phase systems (solid-liquid-plasma) with moving boundaries;
- Nanoparticle nucleation, growth, transport and deposition on the plasma system walls using different numerical approaches for solving the aerosol general dynamic equation that allows the prediction of nanoparticle size distribution with no assumptions on its shape;
- Chemical and thermal non-equilibrium effects and near-electrode phenomena.
These models have been validated in the past for most of the plasma processes investigated, and especially for plasma systems for nanoparticle synthesis under the framework of the FP7 project SIMBA.
The Laboratory for Industrial Applications of Thermal Plasmas is equipped a cluster of 80 processors for parallel computing to perform the modelling activities.
