Turbulent and collisional transport in magnetic fusion devices

SNIC 2018/3-136


SNAC Medium

Principal Investigator:

Istvan Pusztai


Chalmers tekniska högskola

Start Date:


End Date:


Primary Classification:

10303: Fusion, Plasma and Space Physics



In order for fusion to live up to its promise - a clean large scale energy source with abundant resources - several scientific and technological challenges need to be met. Most of these are directly linked to the transport of heat, particles and momentum between the hot core and the edge of the fusion plasma. The performance of a fusion device, and ultimately its viability as a reactor, largely depends on the transport through a region that makes up only a small fraction of the fusion plasma: the transport barrier. In these thin formations located at the plasma edge the otherwise dominant turbulent transport is strongly reduced due to sheared flows that reduce the spatial correlation between turbulent fluctuations. Accordingly, transport due to Coulomb collisions between particles becomes more relevant in this region. Indeed, recent studies performed on a moderate size fusion device, ASDEX-Upgrade, established that ion heat transport can be fully accounted for assuming that the transport is purely collisional. Last year we have studied the collisional transport across the edge transport barrier of the largest currently operating tokamak, JET, and found that, surprisingly, the transport in these transport barriers must be dominantly turbulent. This not only shows that turbulence can survive the sheared flows in the transport barrier, but also indicates that the size of the tokamak affects the turbulent nature of the transport barrier; a question of major importance when the magnetic confinement fusion community prepares for the operation of an even larger device, ITER, currently being built in France in an international collaboration, noting that our estimates for its performance is, to large degree, based on extrapolations from current devices. The size of the device can be characterized by how small a typical ion Larmor radius is compared to the characteristic linear dimension of device; a quantity denoted by rho*. Within a work package of EUROfusion on modeling for medium size devices (MST1), we plan to study the scaling of transport inside the transport barrier in dedicated experiments scaling rho*, with particular emphasis on its collisional/turbulent nature. For the study we use the GATOOLS transport package of General Atomics, including the collisional transport code NEO and the turbulent transport code CGYRO. The latter is a new version of the widely used state-of-the-art tokamak turbulence code GYRO, optimized for collisional environments, as the plasma edge. In sharp transport barriers, where the ion orbit width is comparable to the length scale of the profile variations, coupling of the perturbations across the magnetic surfaces, referred to as "radially global effects", may become important. This aspect is not captured by conventional collisional solvers, instead it is addressed by computationally extremely expensive full-f codes that are not suitable for exploratory studies or performing parameter scans. Our group is currently the main developer of the collisional transport solver PERFECT, an additional tool to be used in the study, that is unique in that it accounts for global effects, while remaining computationally cheap compared to full-f codes.