G8 Multilateral Research Funding Nu-FuSE

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Physics and Astronomy

Abstract

The primary focus of fusion energy research over the past decades has
been on magnetic confinement devices called tokamaks. The UK's JET
facility is currently the largest tokamak in the world but will be surpassed
by the international ITER device, now under construction in France.
Providing computational resources in support of ITER is a dedicated High
Performance Computer for Fusion (HPC-FF) at Jülich, Germany, which will
soon be expanded at Rokkasho, Japan in the International Fusion Energy
Research Center within the ``broader approach" framework between
Japan and the EU. Operation is planned to start in 2012. Simulation is
required in three principal areas: plasma physics, the powerful controlling
plasma-solid interaction/interface, and materials science. We have
constructed a consortium involving six countries (France, Germany, Japan,
Russia, UK and USA) with expertise in all three of these applications
domains as well as the underpinning computational science techniques.
We propose to use these skills to undertake an integrated research
programme focussed on investigating the scaling of key codes which have
relevance for providing experimentally validated predictive capabilities for
magnetic fusion systems.
On the path towards an economical magnetic confinement fusion reactor
integrated numerical models can speed up technological but also physical
progress, even mitigating possible bottlenecks.
Our proposal concentrates on codes for three scientific areas, the plasma
itself, the materials from which a reactor will be built, and the physics of the
plasma edge.
Previous research on materials and plasmas has been conducted
independently, and a key aspect of the proposal is to ensure that scientists
working in these areas are well versed in all the issues affecting putative
devices. We will train a cohort of young scientists who are genuinely
expert in "Fusion Energy", as opposed to the current division of expertise
between plasma physicists, reactor engineers andmaterials scientists.
This group will comprise both the researchers paid for by the project, and
the students funded by the constituent universities to work alongside them.
It will also bring together the international group of senior scientists (PIs)
from different fields united in the goal of supporting a practical fusion
energy device. The collaborative training courses will ensure that
expertise in one area is matched by an understanding of the other.
Particularly in materials science, researchers are using codes developed
to treat a wide range of materials. The issues relevant for fusion are more
specific, and there is plenty of
scope for both algorithmic and parallelisation developments to lead to
significant speed-ups.
By concentrating on community codes, we will ensure that the exascale
developments of the project are of benefit to a wide range of external
users, in addition to the scientists working on the project itself.

Planned Impact

Current worldwide energy consumption has risen 20-fold during the 20th century and the growth rate shows no
sign of saturation. Of the current 15 TW load 80-90% is derived from fossil fuels, but with peak oil imminent and
coal supplies limited, a sea change in energy production is vital. Renewables will make a contribution to future
energy production, but they are generally unrelated to seasonal and geographical demands - while nuclear fission
raises significant environmental and political worries. Nuclear fusion, on the other hand, promises a low pollution
route to generate a large fraction of the world's energy needs sustainably. However, the scientific and engineering
challenges in designing such a reactor are formidable and commercial power plants are not expected before 2050.
Real progress therefore needs to be made now if fusion is to be relevant as coal and oil decline.


Fusion power is famously always 30 years in the future. To some extent this is due to the stop-start nature of funding over the
last few decades, but also due to the formidable difficulties encountered both in physics and engineering. Many of these were
"unanticipated" in the sense that they are irrelevant to smaller devices. Materials is the classic case, for fusion devices to date
the radiation dose is negligible compared with the lifetime-limiting doses in fission reactors. Even at ITER, the doses are not extreme, and since
continuous operation for many years is not envisaged, checking and repair is possible. But a commercial fusion tokamak reactor
would produce far more, and more energetic radiation, in the form of very high neutron dose. The timescales for R&D, licensing
and formal testing mean that it is perfectly possible that a wholly new class of materials could take 20-30 years to be usable: so this
process needs to start now, which means proceeding by simulation and extrapolation of experimental results without the necessary experimental verification.


It is not yet known which area of physics (or interaction between areas) will prove the rate limiter in progress towards a commercial reactor. Consequently it is important to make progress on all fronts to identify possible bottlenecks in as-yet unbuilt devices. Any advances in understanding could speed implementation of the fusion programme by several years. This is the major impact of the project.

Publications


10 25 50
publication icon
Ackland G (2011) The MOLDY short-range molecular dynamics package in Computer Physics Communications


publication icon
Ackland G (2012) Temperature dependence in interatomic potentials and an improved potential for Ti in Journal of Physics: Conference Series


publication icon
Ackland G (2009) Twenty five years of Finnis-Sinclair potentials in Philosophical Magazine

publication icon
Ackland G (2010) Materials science. Controlling radiation damage. in Science (New York, N.Y.)

publication icon
Ackland GJ (2015) MATERIALS SCIENCE. Bearing down on hydrogen. in Science (New York, N.Y.)

publication icon
Ackland GJ (2012) Mechanical properties: Overcoming old barriers. in Nature materials



 
Description Nu-FuSE is an international project (funded through the G8 Research Councils Initiative on Multilateral Research Funding) looking to significantly improve computational modelling capabilities to the level required by the new generation of fusion reactors.

The focus is on three specific scientific areas: fusion plasma; the materials from which fusion reactors are built; and the physics of the plasma edge. This involves computing at the "exascale" level across a range of simulation codes, collaborating together to work towards full integrated fusion tokamak modelling.

The largest supercomputers today can perform over a Petaflop of calculations per second on real scientific applications. However Exascale systems are planned for 2016-2018, performing an Exaflop is a million million million calculations per second, a thousand times faster. To exploit these systems effectively for fusion modelling creates significant challenges around scaling, resiliency, result validation and programmability


The project focussed on meeting these challenges by improving the performance and scaling of community modelling codes to enable simulations orders of magnitude larger than are currently undertaken.

This upscaling allowed important new science discovery.
We discovered that the stability of plasmas is better at large sizes than one might have guessed from extrapolation of small-scale calculation. We also uncovered the mechanism for radiation resistance in so-called ODS steels.

In addition the project has developed a community of fusion scientists with the requisite skills to model fusion devices as a whole, rather than focusing on one part as is currently common practice.

Nu-FuSE is an international project sponsored by the G8 group of leading industrial nations. It is led by Professor Graeme Ackland at The University of Edinburgh, and includes research teams in Caderache (France, and also the location of the next generation of Fusion reactors, ITER), Edinburgh (UK), Princeton (USA), Garching and Jülich (Germany), Keldysh Institute of Applied Mathematics (Russia), and Tsukuba (Japan).
Exploitation Route Improvement in fusion devices such as ITER. Real-time modelling of plasmas to detect potential instabilities.
Sectors Energy
URL http://www.nu-fuse.com/
 
Description Code implemented wordwide in prediction of fusion plasma behaviour.
First Year Of Impact 2013
Sector Energy
Impact Types Societal,Economic
 
Description PPPL 
Organisation Helmholtz Association of German Research Centres
Department Jülich Research Centre
Country Germany, Federal Republic of 
Sector Public 
PI Contribution Led G8 collaboration
Collaborator Contribution Collaborative project + computer time in NuFuSE
Impact Multidisciplinary work on supercomputing for fusion devices Shared expertise in coding and computing Access to various pre-exascale machines for code development
Start Year 2011
 
Description PPPL 
Organisation International Development Research Centre
Country Canada 
Sector Public 
PI Contribution Led G8 collaboration
Collaborator Contribution Collaborative project + computer time in NuFuSE
Impact Multidisciplinary work on supercomputing for fusion devices Shared expertise in coding and computing Access to various pre-exascale machines for code development
Start Year 2011
 
Description PPPL 
Organisation Keldysh Institute of Applied Mathematics
Country Russian Federation 
Sector Academic/University 
PI Contribution Led G8 collaboration
Collaborator Contribution Collaborative project + computer time in NuFuSE
Impact Multidisciplinary work on supercomputing for fusion devices Shared expertise in coding and computing Access to various pre-exascale machines for code development
Start Year 2011
 
Description PPPL 
Organisation Princeton University
Country United States of America 
Sector Academic/University 
PI Contribution Led G8 collaboration
Collaborator Contribution Collaborative project + computer time in NuFuSE
Impact Multidisciplinary work on supercomputing for fusion devices Shared expertise in coding and computing Access to various pre-exascale machines for code development
Start Year 2011
 
Description PPPL 
Organisation University of Tsukuba
Country Japan 
Sector Academic/University 
PI Contribution Led G8 collaboration
Collaborator Contribution Collaborative project + computer time in NuFuSE
Impact Multidisciplinary work on supercomputing for fusion devices Shared expertise in coding and computing Access to various pre-exascale machines for code development
Start Year 2011
 
Title MOLDY 
Description MOLDY is a parallelised OpenMP short-ranged molecular dynamics program, first written at Harwell Laboratory in the 1980s. The program is rewritten in a modular fashion to allow for easy user modification, in particular the implementation of new interatomic potentials. Using Link Cells and Neighbour Lists, the code fully exploits the short range of the potentials, and the slow diffusion expected for solid systems. The code allows for a wide variety of boundary conditions, including constant pressure, temperature and strain rate. It also incorporates molecular statics via the conjugate gradients minimisation of the enthalpy. The code will enable simulation of millions of atoms using short range potentials. Currently modules for Embedded Atom, Finnis-Sinclair, Lennard Jones and Morse potentials exist. In addition, the "magnetic" potential formalism of Ackland and Wallenius is available for separate compilation. Alloys containing a number of elements can be simulated, subject only to the available potentials. 
Type Of Technology Software 
Year Produced 2009 
Impact Molecular dynamics calculations by multiple groups worlwide 
URL http://code.google.com/p/moldy/