The Beam Emission Spectroscopy (BES) experimental technique has been used in fusion plasma experiments for decades. Nearly all major fusion experiments are equipped with BES diagnostics, and a variety of applications have been developed. The basic idea is that an atomic beam can penetrate the magnetic field of fusion devices and the beam fluorescence is measured. The light emission originates from excitation by plasma particle collisions; hence the light intensity gives information primarily on local plasma density.
Principle of the beam emission spectroscopy diagnostic
Light rays in and port structure of EAST alkali beam diagnostics
Transport of energy and particles in magnetic fusion devices is mostly driven by small-scale turbulence. Fusion plasma is a self-regulating system, where parameter profiles drive the small turbulence and turbulence sets the parameter profiles. Additionally turbulence itself is regulated by self-generated flows. BES diagnostics are capable of measuring all 3 elements: the plasma density profile, the ion scale (cm) turbulence structures, and through their movement the flows. The temporal resolution in these experiments is typically 1 microsecond, the spatial resolution is cm.
Our group designs, manufactures, installs and operates BES diagnostics in several major fusion devices around the globe. Heating beam BES diagnostics were installed at MAST, KSTAR and EAST tokamaks, while alkali beams are operated in collaborations at JET, KSTAR, EAST, Compass and ASDEX Upgrade tokamaks. Wendelstein 7-X stellarator is also being equipped with an alkali beam BES diagnostic. This diversity gives a unique opportunity for multi machine comparison of several important physics phenomena.
BES diagnostics built and partly operated by the members of the BES research group around the world.
Multi-device measurements of type I ELM triggering mechanism
The type I ELMy H-mode is the baseline operating scenario for ITER. However, the energy predicted to be released by a type I ELM in ITER is unacceptably high. The challenge is to produce a non-linear model that can predict when an ELM will be triggered and predict the ELM energy loss. The present understanding is that type I ELMs result from the peeling-ballooning instability. In this theory the edge pressure gradient grows in the inter-ELM period until the peeling-ballooning stability boundary is crossed at which point the ELM is triggered. However, it has been observed on several devices that the experimental profiles can exist near to this stability point for a substantial fraction of the inter-ELM period raising the question: what ultimately triggers the ELM? Normally there is a large velocity shear at the edge of the H-mode plasma but the non-linear theory (explosive growth stage) of the ELM predicts that the filaments associated with ELMs have to push out through this edge of the plasma and that these filaments can only grow through a region of small velocity shear. Hence, in order for the ELM to erupt the edge shear velocity has to be reduced. Experimental observations of a suggests that a counter rotating mode localised at the top of the pedestal that could be a mechanism that reduces the flow shear in this region.
Several devices have reported a precursor oscillation that grows up in the 50 – 150 microseconds before an ELM is triggered. On MAST, a 2 dimensional 4x8 pixel turbulence imaging beam emission spectroscopy (BES) diagnostic has been used to characterise the edge plasma region with a 2cm spatial and 500 kHz temporal resolution. The excellent signal to noise (up to 300) and very high signal to background (up to 50) has enabled a detailed study of the density variation pattern before and during ELMs.
During the ELM crash the BES system observes filament structures, similar to those observed previously on MAST using visible imaging that erupt from the edge of the plasma and rotate in the co-current (ion diamagnetic) direction. What is new is the observation of the growth of a mode which is localised near to the top of the pedestal that grows up in ~ 100 microsecond before the ELM crash. The mode is characterised with a frequency of ~ 20 kHz, a 6-10 cm poloidal wavelength (corresponding to a toroidal mode number of n~40) and a radial extent of ~ 2cm. The mode is observed to propagate in the counter current (electron diamagnetic direction). At a time when the plasma edge profiles are at the peeling ballooning stability limit, the mode appears to modify the flow in the pedestal region allowing the filaments associated with the peeling ballooning mode to grow and erupt from the plasma edge.
In this project the measurements from MAST, KSTAR and from 2016 Compass are compared and checked against modelling results.
MAST –U BES upgrade
A turbulence imaging 2D BES system has been installed on MAST at 2010. This diagnostic was operating very successfully in M8 (2011) and in M9 (2013) measurement campaigns. A major results was these years that the core turbulence measurements of various plasma scenarios were compared model calculations. Also ELM precursor waves were characterized the first time in the history of the MAST operation. The experiment was shut down for a major upgrade in 2013 and is expected to be operational in 2017. The BES diagnostic is also being upgraded during the shutdown. The optics was redesigned and being manufactured. A larger detector will provide more detailed information and also new mechanics is being manufactured.
Engineering drawings of BES observation system designed and built for the MAST Upgrade spherical tokamak in Culham, Oxfordshire, UK.
Multi machine characterization of ELM free H-mode phenomena
The H-mode (High confinement mode) is a plasma scenario which is now considered as the operation state of a future fusion reactor. The plasma confinement advances considerably with the formation of a plasma edge transport barrier that characterizes H-mode, along with the steepening of the plasma edge profiles (pedestal formation), and the increase of the plasma core temperature and density, which is essential for a fusion reactor with positive energy balance. The edge localized modes (ELM) those appear in H-mode, are followed by large, periodic and abrupt release of particles and energy. These bursts reach the wall that has a finite mechanical and thermal load capacity. A future reactor would not be economically feasible without the suppression or the control of the ELMS according to our current knowledge, due to the need of the frequent wall replacement.
The ELM control and suppression is in the focus of fusion research, and several technology are already available for the reduction of the wall heat load. Along with these, it would be obviously beneficial if we would find a plasma scenario that has the advances of the H-mode, but lacing the difficulties caused by the ELMs. These plasma scenarios are called ELM free H-modes, and are in the spotlight recently due to the aforementioned conditions.
Another interesting point about the ELM free H-modes is their typical appearance, when the plasma goes from L-mode (low confinement mode) to H-mode, which is called the L-H transition. The exact cause of the L-H transition is not known so far, and thus it is in the focus of the fusion research, however several theories were born for the explanation. The ELM free H-mode regimes can be important in L-H transition physics considering their appearance in the transition state.
The BES team participated the operation as well as the development of the Lithium beam emission spectroscopy diagnostic at the Joint European Torus (JET) and at the ASDEX Upgrade (AUG) tokamak. We made several developments on both diagnostics between 2011 and 2016 those were focused mainly to JET. As a result of the developments we are capable of doing plasma edge electron density profile measurements with ten-thousandths second time resolution at both devices that makes us capable observing numerous fast phenomena.
The M-mode phenomena at JET and the I-phase phenomena at AUG are ELM free H-mode plasma operation regimes. We showed with our diagnostic, that the two phenomena modulates the plasma edge density and the transport between the plasma and the wall with about 1kHz, as well as that the dynamics of the plasma density profile is very similar on both machines. We would like to extend our study on more devices, and make a more comprehensive analysis on data recorded during similar plasma states.
A lens of the Li-BES system optics at JET. We collect the light from the beam on the detector with that.
KSTAR SOL measurements
A 2D BES system was installed on KSTAR in 2012. The diagnostic has been commissioned in 2012 and the first low noise level, low background level measurements were done in 2013. After installing a 64 pixel ultra-fast camera unit based on Avalanche Photodiode array detectors in 2014, the plasma measurements became possible outside the confined region, in the so called scrape-off layer region. The fluctuations are measured in a 4cm by 16cm region, of which a 4cm by 4cm region is observing the scrape-off layer. In the scrape-off layer, filamentary structures can be found, also known as blobs, which are propagating radially outwards and they play an important role in the overall plasma transport. The BES measurement on KSTAR is possible to distinguish these structures from the bulk plasma and their dynamics can be also analyzed in detail. The following figure depicts an average blob of several hundred blob events just outside the plasma edge.
Wendelstein 7-X alkali beam diagnostic
An alkali beam diagnostic is being designed and manufactured for the Wendelstein 7-x stellarator in the framework of the Eurofusion WPS1 project in collaboration between Wigner RCP Budapest and IPP Greifswald. The ion gun will be installed in 2016 as the observation system is planned to be installed in 2017.
The primary objective of the diagnostic is to measure density profiles, island structures, turbulence and turbulence flow properties in the SOL as well as the edge region of Wendelstein 7-X. The main target is to measure upstream densities at high spatial (1cm) and temporal (1µs) resolutions.
The engineering drawings of the ion gun built for Wendelstein 7X stellarator.
Beam emission spectroscopy diagnostics on the EAST tokamak in Hefei, China
The BES group collaborates with the EAST tokamak team in Hefei, China. This fusion device is a large superconducting tokamak experiment studying the confinement of fusion plasmas in long discharges, up to 100 second. We built and jointly operate two diagnostic systems for the tokamak: a Lithium-beam injector with an optical observation system and a BES observation system (built by Adimtech Kft.). The Li-beam injector is similar to the one installed on the KSTAR tokamak and aims at measuring plasma turbulence and edge density profile via the detection of the beam light emission.
The BES observation system was built by Adimtech Kft, the spin-off company of the Hungarian Academy of Sciences. It collects the light emitted by a heating Deuterium beam via a 8x16 pixel Avalanche Photodiode matrix with the aim of measuring plasma turbulence properties and fast instabilities.
Modeling of the beam light emission for both systems is done by the Budapest University of Technology and Economics.
Both diagnostic systems are preforming their first measurements with some density profiles and turbulence properties already obtained.
Photograph of the BES observation system in the laboratory before installation in the tokamak. The long tube is installed into the cryostat port, it leads light collected by the front lens in the bottom right corner. The black frame contains the detectors and optical filters. a vacuum window.