Plasma modelling with Plasimo – design and applications

Original title : The Art of Modelling Plasmas and Gas Discharges. With the advent of cheap, yet powerful computers, self-consistent modelling is becoming a powerful tool for better understanding the behaviour of plasmas and gas discharges. Even timedependent modelling of non-equilibrium discharges in three dimensions starts becoming viable. In order to create a computer model for a discharge, one must take into account virtually the entire landscape of modern physics, like (non-)equilibrium plasma chemistry, the barycentric flow field, the gas temperature, plasma-wall interaction, radiation transport and the electro-magnetic field. For modelling individual aspects of plasma behaviour, a variety of commercial plasma simulation programs are available, like FLUENT, Opera-2d, CHEMKIN or ChemApp. While such commercial programs may be superior in their respective domains of application, this is not necessarily the case for self-consistent plasma models in which such programs are embedded: the lack of control overclosed-source – hence unmodifiable – external programs may result in missing model features, substantial communication overhead or convergence problems. With these observations in mind, the plasma modelling toolkit Plasimo is being developed at Eindhoven University of Technology. Plasimo is an open framework that has been designed with extendibility in mind. It is meta-configurable in the sense that models can be composed (at runtime) by defining the plasma geometry and selecting appropriate chemical, electro-magnetic, transport and radiation sub-models. Plasimo supports kinetic, fluid and hybrid transport calculations. A modular and orthogonal design ensures that new modules can be added without changing the core infrastructure, and used irrespective of the context for which they were initially developed. In the first part of this contribution we will discuss Plasimo's current feature set and shed some light on the principles that underlie its design. The application of Plasimo to various low- and highpressure plasma sources will be demonstrated in the second part of this contribution. Amongst others, we will discuss metal-vapour plasmas, plasma light sources and a microwave-driven deposition plasma. Special attention will be paid to high-pressure metal-halide plasma lamps that feature convection-driven axial inhomogeneities that cannot be understood on the basis of local thermal equilibrium. A noteworthy feature of the lamp model is the fully spatially and spectrally resolved treatment of radiation transport. The radiative source terms are fed back into the enthalpy equation, making the model fully self-consistent. In the last part of this contribution we will reflect on the present status of plasma modelling more generally. We will put ourselves in the position of a "benevolent dictator", not bridled by scarcity of manpower, funding or other practicalities. We will try to imagine what plasma modelling roadmap he would come up with. From such Utopian ideal, we will extract the plans that we have for the future of Plasimo. Since Plasimo (including the source code) is available for peer researchers in plasma science and other fields of physics, we wholeheartedly invite our colleagues to participate in its further development.