In a tokamak, the hot plasma conditions required for fusion are maintained using magnetic fields. To a good approximation, the plasma particles (electrons and deuterium/tritium ions) are constrained to follow the magnetic field lines, which wind around toroidal flux surfaces. Unfortunately the confinement is not perfect. Collisions between particles can cause them to diffuse across the flux surfaces and eventually be lost from the plasma: this is called classical transport in general. In a toroidal plasma there is a class of particles that are trapped in regions of low magnetic field. Their orbits are larger than the Larmor orbit and give rise to correspondingly larger transport: neoclassical transport.
It has long been known that in tokamaks the transport is typically an order of magnitude larger than the predictions of the neoclassical theory. This “anomalous” transport is thought to be a consequence of turbulence. In a plasma, this turbulence can be driven by a variety of small-scale instabilities. The likely candidates are instabilities related to drift waves, sound waves and micro-tearing modes. Our work in this area explores aspects of these instabilities. It is important because the amount of transport determines the size that a tokamak reactor needs to be: the more transport, the larger it needs to be.
The main emphasis of our research is theoretical and computational modelling of both the linear and non-linear instability characteristics. Our computational work employs the GS2 code, which simulates the plasma turbulence in a narrow (flux) tube of plasma that is aligned with the magnetic field lines. A particularly interesting feature of tokamak transport is that it can exhibit bifurcations: a small change in a plasma parameter (such as heating power) can lead to a sudden, dramatic improvement in confinement in a narrow region of plasma. This is called a transport barrier and is thought to be a consequence of strong flows that can be generated, which then tear the turbulence apart. Our research explores how such a flow affects the stability of the linear modes that are candidates for driving plasma turbulence (mentioned above). Here, the formalism for how to treat such a situation remains incomplete, so our work involves analytic theory to improve our understanding of the basic mechanisms as well as more detailed numerical modelling with the GS2 code. In parallel with this activity, we perform computationally intensive non-linear calculations of the turbulence in the presence of flow. These calculations make use of the HECToR parallel computer facility.