Probing Plasma Instabilities with high performance simluations

Lead Research Organisation: University of York
Department Name: Physics


The fusion performance of tokamaks is strongly influenced by the effectiveness of the background magnetic field in holding particles and energy inside the central plasma volume and away from the tokamak's walls. Fluctuations in the electromagnetic potentials, resulting from a multitude of small scale plasma instabilities, can limit the confinement that can be achieved in these tokamak plasmas. The study of these instabilities and the associated turbulent fluctuations is therefore vital in order to better understand the physics processes leading to enhanced losses and hence fusion performance degradation. The small scale and turbulent nature of the fluctuations can pose a challenge to experimental measurements of instability properties and data interpretation. Both theoretical and numerical studies of instabilities are therefore extremely useful approaches which can aid interpretation of the experimental data as well as allowing more flexibility in probing the underlying physics in detail.
In this project the researcher will develop experience with using leading edge gyrokinetic simulations to investigate the properties of instabilities in experimentally relevant scenarios. As a part of this work, a novel approach to capturing important effects arising from the slow spatial variation of the background equilibrium will be exploited to study the impact of equilibrium properties on the instability structure and the knock on effect for diagnostic measurements and interpretation. A key motivation of this project, alongside improving our fundamental understanding of plasma instabilities and associated turbulence, will be to explore the consequences of instability character for experimental diagnostics such as the Doppler Back Scattering (DBS) and Beam Emission Spectroscopy (BES) system. This will also aid efforts to validate the gyrokinetic tools through comparison of measured and predicted fluctuation signals and the researcher will be encouraged to exploit their gyrokinetic simulation skills to provide interpretive support in the diagnosis of experimental data. This project aligns closely with the goals of a wider collaborative project involving the universities of York and Oxford and the Culham Centre for Fusion Energy (CCFE), providing the researcher with the opportunity to work closely with experimentally focussed PhD students as well as with experimental and theoretical experts within the international fusion community on these areas.

Planned Impact

Identifying a solution to the energy problem is crucial to the UK economy and quality of life. In the near term a range of renewable options must be developed, eg wind and solar, but it is unlikely that these will provide the base-load supply required. Nuclear is an option for a carbon-free base-load and, in particular, fusion energy is safe and relatively clean. If it can be achieved, fusion would bring the largest economic benefits to those countries that lead the way to build the first fusion power plants, but ultimately most people in the world will benefit from fusion in some way.
ITER, the largest international science project on Earth, will operate from 2020 to answer the final physics questions and most technology questions required to construct the first demonstration magnetic fusion energy (MFE) power plant, DEMO. We will train the ITER generation of UK fusion scientists who will have the expertise to win time on this key facility against international competition. This is crucial to build experience that will feed into the design of DEMO, ensuring the UK remains at the forefront. EU design studies for DEMO are already under way, with manufacture of prototype components likely to follow soon. There are a number of beneficiaries from this training: (1) it will benefit Culham Centre for Fusion Energy (CCFE), providing well-trained new staff to replace those retiring, keeping the UK at the forefront of fusion energy research, competitive for ITER time and leading elements of DEMO design/prototype development; (2) it will provide expertise for the growing UK industry involvement in fusion, helping to win contracts for ITER and DEMO prototype components; (3) it will ensure the UK has a cadre of fusion experts to advise Government on future directions. We expect to train 60 students in MFE, approximately balanced across plasmas, materials (relevant for IFE also, see below) and related fusion technologies.
For inertial fusion energy (IFE), NIF in the US is the most advanced device in the world, and some expected it would achieve fusion conditions, i.e. ignition. In its 2012 ignition experiments, this did not happen, but the reason why is still uncertain. The immediate need is to understand this, which requires experts to win time on international facilities (including NIF), understand why ignition did not occur and so develop a roadmap to IFE based on the new knowledge. This will benefit the UK Government by providing experts to advise on an appropriate strategy, able to compare the relative merits of IFE and MFE because of our training across both areas. If IFE proves viable, then it will need to integrate fusion technologies in a similar way to ITER and DEMO, bringing benefits to industry. We expect to train 15 students in high energy density physics (HEDP), spanning IFE and lab astrophysics; the MFE materials students' expertise is also relevant for IFE reactor design.
Expertise in HEDP is required by AWE for its science-based approach to underpinning the UK's nuclear deterrent, and is a key element of the UK's strategy to comply with the Test Ban Treaty. The new Orion laser facility at AWE can replicate the conditions in a nuclear warhead, enabling advanced computer codes to be tested. Our students will have the expertise to work with Orion, which requires skilled scientists as it establishes its programme. Also the materials and computational scientists amongst the ~60 MFE students will be of value to AWE.
We will train students in the cooler exhaust plasma of a tokamak. Similar plasma conditions are used in manufacturing industries (coatings, computer chips, etc) so we will develop a skill base that will benefit a number of such companies. Materials research for fusion is also relevant for fission. The popularity of fusion amongst students is a good way to bring outstanding students into the field, providing expertise that benefits the growing nuclear industries and supporting the Government's nuclear policy.


10 25 50
Description The key findings of this award are still emerging and evolving. However, so far, the key findings are as follows: (1) Computer simulations of fusion plasmas are known to require a lot of computing resources due in part to the large number of particles (~10^19 per metre cubed). One approach to making this easier is to simulate a small slice of plasma (less plasma = fewer particles), known as the "local approximation". This only works if the plasma conditions are approximately constant across the plasma slice. This is true in the plasma core where conditions are relatively constant. However, there is often concern that this assumption is questionable at the plasma edge where the temperature and density change drastically over short distances. Simulation results before and during this award appear to indicate that there may be an additional concern in edge simulations that significant changes in the magnetic field alignment over short distances can also invalidate the local approximation and thus necessitate more computationally intensive simulations of a larger section of the plasma. However, mathematical work carried out as part of this award has found that the invalidating effect of rapid magnetic field alignment change is cancelled out by the opposite effect on other parts of the local approximation. Therefore, simulations are under-way to test the temperature and density concerns. If successful, this will: (a) provide confidence in a method that can be applied to the analysis of other tokamak plasmas; and (b) provide insight into the behaviour of one particular type of plasma oscillation under certain conditions, which is of interest for various reasons mainly related to machine efficiency and protection. (2) Another set of simulations are ongoing to investigate the spatial layout of where another type of plasma oscillation occurs. This is to support experimental work to find evidence of such oscillations in a physical device. Preliminary results indicate that useful information will be found but further analysis is required to make the results more accurate and interpret the implications thereof. (3) To support the two items discussed above, numerous bug-fixes and new features have been added to the simulation code (gs2) as part of this award. This constitutes a useful contribution to the improvement of an important, world-class simulation code.
Exploitation Route For item (1), await publication of the findings and: (a) take these into account in future simulations; and (b) use the physics results to improve other models that use this data as input. For item (2), await publication of the findings and use that information in experimental searches for the plasma phenomenon of interest. For item (3), some updates are available for anyone to use right now and more changes will become available imminently (currently undergoing review before being published via the source code version control system).
Sectors Energy

Title Mode Tracking 
Description When using the GS2 code to simulate plasma vibrations, by default it only reports the most unstable mode of vibration. However, it is sometimes necessary to track a particular mode of vibration through parameter space to areas where it is not the most unstable. The has been done previously using an alternative solver that can report many modes of vibration. However, this is computationally expensive and thus necessitates low resolution in parameter space. Furthermore, it is difficult to reliably identify the mode of vibration that one is trying to track. I developed a method where small steps through parameter space are used such that the mode of vibration is similar to that at the previous point so the previous solution can be used as the starting point for the next search. The disadvantage of this is that the different points in parameter space are no longer independent so cannot be solved in parallel. However, this is outweighed by the advantages: namely that the solution is reliably found at each point and in a very short time so the overall runtime is comparable to the previous method. NB: Not published yet but will be in due course at which point it will be available to all. In the meantime, the method is available to others who have seen my talks on this subject. 
Type Of Material Improvements to research infrastructure 
Year Produced 2016 
Provided To Others? Yes  
Impact Enabled investigation of the effect of plasma shaping on turbulent instabilities (also to be published in due course). 
Description GQL Project - Leeds 
Organisation University of Leeds
Country United Kingdom 
Sector Academic/University 
PI Contribution Colaboratory project with University of Leeds. I helped them to test their generalised quasilinear approximation. It has previously been tested in hydrodynamic geometries. This project tested it for the first time in a fusion relevant plasma turbulence regime - the Hasegawa-Wakatani (HW) equations. I developed a model in their code Dedalus and optimised its performance. Work is on-going with a publication likely following completion of the project. I also showed them our finite difference code BOUT++ and developed a standalone highly-optimised spectral HW code.
Collaborator Contribution Instruction on how to operate their spectral code Dedalus, how their approximation works, how spectral methods in general work and the bigger picture, namely using direct statistical simulation instead of direct numerical simulation as a quicker route to calculating the statistics of a system.
Impact Publication likely in due course. Multi-disciplinary: Applied Mathematics and Plasma Physics
Start Year 2018
Title GS2 
Description GS2 - A physics package for simulating plasma turbulence. The code was already in existence but I have contributed to its development as part of this award. 
Type Of Technology Software 
Year Produced 2019 
Open Source License? Yes  
Impact Improved the code to output geometry information in < 1 second instead of > 5 minutes. Added a framework to automate testing of the code as changes are made. Helped transition to a better source code version control system. Wrote various post-processing scripts that are expected to contribute to the basis of the official analysis program in due course. Improved the memory demands of loading data files by a factor of 4 which enabled working with large data files with limited RAM (and also generated a speed-up in runtime). 
Description School Visit (York) 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact ~20-30 pupils across a wide age range (primary and secondary) attended an event at their school where we had various interactive physics demonstrations about astrophysics, robotics/programming, spectroscopy, etc. Staff and students at the school all seemed to enjoy the activity. The school reported increased interest in related subject areas and remained in contact with the university for various follow-up activities.
Year(s) Of Engagement Activity 2017