Fusion plasma theory and modelling
The Plasma Theory and Modelling group focuses on understanding these fundamental properties of plasmas through a variety of different academic perspectives.
About
Big machines, hot plasma, and cool technology
- Plasma physics lies at the centre of many exciting technologies that are undergoing rapid development in the scientific and engineering communities. In particular, the drive behind magnetically-confined nuclear fusion is gaining significant momentum with the recently emerging desires of governments and national populations to get away from fossil-fuel based power. Construction has recently started in France on the first fusion reactor--dubbed ITER--that will consistently produce more energy than it consumes. The technical and scientific challenges inherent to the construction of ITER are some of the greatest that the scientific and engineering communities have ever had to face. Consequently, a race is currently underway to ascertain as much understanding as possible about the fundamental workings of magnetically confined plasmas, so as to best ensure the success of ITER when construction is completed in 2024.
The Plasma Theory and Modelling group focuses on understanding these fundamental properties of plasmas through a variety of different academic perspectives. With a particular focus on magnetically-confined plasma physics, the group has an eclectic range of expertise including modelling and computer simulation of plasma turbulence and equilibria, fluid dynamics, mathematical theory of dynamical systems, plasma diagnostic design and using advanced statistical techniques in analysing experimental data. The group also actively fosters international collaboration in plasma research and is currently engaged in research with scientists in the UK, Germany and the United States.
Beyond nuclear fusion, the Plasma Theory and Modelling group has members whose interests also include the dynamics, modelling and computer simulation associated with weather, fluid flow and space plasma phenomena, to name just a few.
If you have any interest in plasma theory or modelling please feel free to contact any of the researchers in the group; we would be very happy to talk with you!
Advanced Monte Carlo Modelling for Network Tomography
The study of toroidally confined plasmas requires the tackling of very complex inverse problems, like inferring the internal behaviour of the plasma from limited data provided by sensors. A cross disciplinary research project was undertaken by Matthew Hole and honours student Ashley Barnes, where the modelling and inference skills of the group were applied to computer network tomography.
Network tomography, much like the Positron Emission Tomography (PET) used in medical imaging, is the inferring of internal structures, behaviours and properties of a network by analysing traffic passing between nodes in a network. Where PET analyses the behaviour of positrons passing through tissue to image the internal structure, network tomography can describe the inference of traffic flow, node status and structure of a network. Like investigating the behaviour of plasma currents from limited sensor data, problems in this field require the inference of network properties from the traffic data between a limited number of ‘monitor’ nodes that one has access to.
This project focuses on optimising the placement of these monitor nodes so as to maximise the chance that certain network behaviours can be inferred. So far, we have used a novel stochastic network model to demonstrate that certain monitor placement algorithms perform better than others in dynamic, load balancing networks.
Potential Student Projects
There are several additional ideas that would work well as honours or masters projects, including:
- Breaking networks down into topological constituents and determining whether certain topological properties lend themselves to better inference than others.
- Implementing the Open Shortest Path First (OSPF) network protocol within the model in order to introduce OSPF poisoning attacks. These attacks could then be another network behaviour to detect and locate with the Monte Carlo inference.
If you’re interested in learning more, please contact Ashley ashley.barnes@anu.edu.au or Matthew matthew.hole@anu.edu.au
Presentations
An overview of the research activity within the group is available in the following seminars.
-
Document
-
Document
