News & events
Making the invisible…visible | 10/03/2015
One of the biggest challenges facing fusion physicists is controlling the plasma inside a tokamak reactor.
Plasma – a gas of the fuels that are heated to start the fusion process – is difficult to keep stable, and seeks to escape the magnetic field that confines it within the machine. This results in 'instabilities' which make the plasma wobble and fluctuate, taking energy away from it and affecting the tokamak's performance.
Decades of research on tokamak experiments worldwide has led to a deep understanding of a myriad of different plasma instabilities with exotic names (from Edge Localised Modes to Tearing Modes, Kink instabilities and Sawteeth). Just as importantly, researchers are developing methods to stop them occurring, reduce their effect or stabilise them altogether.
Amongst all these challenges has been the fact that most of these instabilities, certainly those deep inside the plasma, are invisible to high-speed camera videos – until now, that is. University of York PhD student David Ryan is currently working at Culham and he applied cutting-edge video magnification techniques to footage of plasmas in the MAST tokamak to see what would emerge.
Specifically, he applied so-called Eulerian Video Magnification to the videos – a technique developed at the Computer Science and Artificial Intelligence Laboratory at MIT which detects very subtle (normally invisible) changes in intensity and magnifies them in a new video. In this way, the MIT researchers observed the blood pumping through a person's face and a baby's heartbeat.
Applying this technique to MAST video data, David magnified the growth of a particularly important instability, the Neo-classical Tearing Mode (NTM). NTMs can grow rapidly as the plasma gets hotter and in some cases can result in abrupt terminations of the plasma, known as 'disruptions' – making them important to stabilise or at least mitigate. The resulting video shows a large, saturated NTM propagating around the plasma edge; remarkable activity when the original video footage showed nothing visible.
Warning: the video features strobe-like images.
Further use of this technique has enabled new instabilities and filament-like structures to emerge from high-speed camera footage – posing new and interesting questions into what is happening in the hot tokamak plasma.
David was surprised and delighted with the power of this new technique: “One of the most exciting aspects of scientific research is that you never know what you might find. I was sceptical that a magnetic perturbation such as an NTM would even be detectable with a visible light camera, but to get such a detailed picture of its evolution and spatial structure was really quite amazing.”
Looking to the future, David is already excited about future applications. “The MAST high-speed cameras produce an immense trove of useful data, but we lack sufficient video processing algorithms to exploit it," he said. "With some modifications, Eulerian Video Magnification could be used to systematically enhance and characterise a wide range of plasma phenomena hidden in the footage, making the high-speed camera a far more powerful tool for fusion researchers.”