CCFE’s spherical tokamak MAST is the first device to observe finger-like plasma structures emanating from methods applied to mitigate harmful instabilities. The images could help find a solution to one of the biggest plasma physics problems standing in the way of the development of fusion power.


High-speed video image of the MAST plasma obtained at the start of an ELM. The mark indicates the X-point. (Image: CCFE)

So called ‘Edge Localised Mode’ or ELM instabilities are powerful plasma events which have the potential to damage components in future machines like ITER. Not surprisingly then, understanding ELMs and developing techniques to deal with them is a high priority in fusion research. The goal of plasma physicists is to stop ELMs from happening at all, but this is easier said than done. For example, reducing the plasma’s pressure can suppress ELMs – but this also impairs its ability to confine energy, one of the key factors for achieving fusion conditions. Another way of tackling the problem is not to eradicate the instabilities but to control them at a manageable level, limiting the amount of harm they can do. This approach, known as ‘ELM mitigation’, is the subject of an intensive campaign of studies worldwide.

Magnetic perturbations

MAST, like many other fusion devices, is using a technique called resonant magnetic perturbation (RMP) to mitigate ELMs. The idea is to apply small magnetic fields around the device to punch holes in the plasma edge and reduce the pressure – and thus the confinement – in a measured way. This technique has been remarkably successful in mitigating ELMs on several tokamaks.

One surprising feature of the RMP method is that as the pressure at the edge of the plasma drops, the number of ELMs increases. With our understanding of ELMs, lower pressure should lead to fewer ELMs, because there is less motivation for the plasma to release energy and particles. Something else must be happening. The clue is in lobe structures that have recently been spotted in images of plasmas inside MAST during RMP experiments.

The lobes are caused by the resonant magnetic perturbation, which throws particles off course as they move around the magnetic field lines in the plasma, effectively changing their route and eventual destination. Some particles end up outside the magnetic field lines, forming small offshoots near the base of the plasma, at the ‘X-point’. These lobes change the shape of this area of the plasma, which in turn has the effect of lowering the pressure threshold at which ELMs are triggered – causing more frequent but less powerful ELMs. So by shaking up the plasma and deforming a specific part of it, researchers should be able to produce a stream of small, manageable ELMs that will not damage the tokamak – an effective way of keeping them in check.

ELM Fingers

False colour images of the X-point region of the plasma captured by the camera during an H-mode. Top: without RMPs; Bottom: with RMPs, showing the finger-like lobe structures emerging from the edge of the plasma. (Image: CCFE)

First images of plasma lobes

First predicted by US researcher Todd Evans in 2004, the lobes – known as homoclinic tangles – were only seen for the first time during experiments at MAST in December 2011, thanks to the machine’s exceptional high-speed cameras. CCFE scientist Andrew Kirk, who leads ELM studies on MAST, says: “This could be a very important discovery for tackling the ELM problem, which is one of the biggest concerns for physicists at ITER. The aim for ITER is still to remove ELMs completely, but it is useful to have back-up strategies which mitigate them instead. The lobes we have identified at MAST point towards a promising way of doing this.”

The lobes are useful for another reason; they are a good indicator of how well the resonant magnetic perturbation is working. The changes in ELM size and frequency tell physicists that puncturing the edge of the plasma with extra magnetic coils is having an effect, but the lobes show how far this effect is penetrating into the plasma. “The length of the lobes is determined by the amount of the magnetic perturbation the plasma is seeing,” explains Andrew Kirk. “So the longer the ‘fingers’, the deeper the penetration. If the fingers are too long, we can see that it has gone too far in and will start to disturb the core, which is what we want to avoid.”

New codes

The next phase of the research will involve developing codes to map how the lobes are formed around the plasma. “We already have codes that can determine the location of the fingers but we cannot, at present, predict their length due to uncertainties in how the plasma reacts to the applied perturbations. Our measurements will allow us to validate which models correctly take this plasma response into account,” says Andrew Kirk. “Our next job will then be to write 3D codes which can calculate the plasma stability in the presence of these lobe structures. This will mean we can produce accurate predictions for ITER and help them tame the ELM.”
Nick Holloway, CCFE