Integrated Tokamak Modelling Task Force (ITM-TF)

The path that leads toward using fusion as a sustainable source of energy for the future goes via the exploitation of the next generation of fusion devices, and in particular, ITER, which is designed to demonstrate the scientifi c and technical feasibility of fusion. In order to predict ITER performance in terms of energy production and to guarantee optimal operation, high level physics modelling and simulations are required. Modelling of a tokamak experiment is necessary to fully comprehend the experimental results, furthermore, benchmarking models against each other (for verifi cation) and the experiment (for validation) increases their prediction capability in view of the realisation of a fusion reactor.

Pär Strand is an Associate Professor at Chalmers University of Technology, Gothenburg, Sweden and has been acting as the Leader of the ITM Task Force since 2006, after holding a deputy position since the start of the project. He is also the Coordinator of the EUFORIA – “EU Fusion for ITER Applications” project which is funded under the EU FP7 Capacities programme. Dr. Strand is also the ITM-TF representative on the HPC-FF Board and Chairs the ITER Integrated Modelling Expert Group advising ITER Fusion Science andTechnology on modelling issues. Pär Strand’s vita can be found here:

Tokamak physics covers a wide range of areas such as plasma turbulence, magneto- hydrodynamics (MHD), radiofrequency (RF) wave propagation, plasma- wave interaction and plasma-wall interaction. Moreover, the operation of a high performance tokamak requires a number of technological elements including super-conducting magnets, RF wave generators and antennas, fast ion sources and accelerators, plasma diagnostics and real-time control schemes. In addition to the tight coupling of the physical phenomena among each other, technology and physics are also strongly linked: RF wave launching conditions depend on the antenna characteristics, plasma equilibrium is controlled by external magnetic field coils and the interpretation of measurements requires the knowledge of the diagnostic hardware. Although individual models exist for most of these elements and components, a realistic modelling of a tokamak experiment ought to cover all of these physical and technological aspects, whilst taking into account the complexity of their interactions. This requirement results in the emergence of the Integrated Modelling concept.

A wide choice of models, using different levels of sophistication in the physics description, can be applied to describe tokamak plasma phenomena prior to the transcription of the mathematical model into a numerical code. Besides, different numerical methods can be used to solve the same type of mathematical model resulting in an even wider choice of numerical codes for the fusion plasma community. Some of these codes have been extensively verified and validated and integrated into simulators for specific purposes. However, they are often only suited to the geometry settings and experimental data from a specific tokamak and seldom adopt generic descriptions for subsystems (e.g. heating, diagnostics) which match any given experimental device. Moreover, each simulator uses its own format for general input/output and interfaces to its internal modules. This makes the exchange of data and physics modules (code portions containing a physical model) between codes tedious, thus slowing down the benchmarking efforts and the development of physics modules that may be shared. In view of ITER exploitation and the development of future fusion reactors, the challenge is to provide an integrated simulator that meets the reliability standards required by the operation of a nuclear device, and thus enables the transparent use of the available numerical models.

Lars-Göran Eriksson has worked at the JET Joint Undertaking and at CEA Cadarache. He was appointed ITM deputy task force leader in October 2006. He is now at the European Commission in Brussels and is the Directorate General for Research Unit J4 – Fusion Associations Agreement. His duties still include the deputy leadership of the ITM task force. Lars-Göran Erkisson’s vita can be found here:

The European Integrated Modelling approach

The choice of Integrated Modelling made by the ITM-TF is unique and original. It entails the development of a comprehensive and completely generic tokamak simulator that includes both the physics and the machine and can be used for any fusion device. The simulation platform will be fully modular, fl exible and independent of a programming language.

The choice of modularity implies that each module contains a single physical model and that the communication between the modules is standardised. A set of common rules clearly specifi es the format of the data that is to be consistently exchanged between modules (data-structure). The complexity of coupling the codes is therefore transferred to the defi nition of a datastructure which should be generic (able to describe and exchange information concerning both physical quantities and technical objects, not assuming the origin of those) and extendable to allow for the integration of new physics, as well as more elaborate machine geometries and experimental data in the future.

The Consistent Physical Object (CPO) concept has been developed to this purpose: CPOs are the only blocks of consistent data exchanged between physics modules during an integrated simulation. The defi nition of CPOs for the different classes of physics codes derives naturally from the input/output logics of a physics problem; CPOs are identical for all modules addressing the same physics problem and produced as a whole by a single module, thus ensuring the consistency of the data. A CPO might be, for instance, the plasma equilibrium data that is made up of plasma profi les, plasma boundary, a fl ux surface coordinate system and other data.

Thanks to this conceptual paradigm in Integrated Modelling, the physics and programming aspects are separated so that an external or wrapping software can deal with the exchange of data, the link to the experimental database and access to the simulation platform.

Code developers wishing to integrate their code into the ITM simulation platform must provide it in a modular form with the CPOs as input and output arguments. Libraries written in different programming languages and developed by ITM, automatically handle the exchange of CPOs between modules. An open source software (KEPLER) featuring a user-friendly interface allows the user to choose, add and link together the physics modules and launch the desired integrated simulation (a simulation workfl ow). A central computer platform was made available in 2008 on which the ITM suite of tools is tested and released – the “Gateway” hosted at ENEA/CRESCO. It allows ITM-TF users to run simulations locally or within the framework of GRID (a grid of geographically distributed supercomputers) or High Performance Computer environments. Additionally, the High Performance Computer for Fusion (HPC-FF) at FZ Jülich was made available in autumn 2009.

Rui Coelho is an assistant researcher at the Instituto Superior Técnico of Lisbon (IST) and has been assigned to the Theory and Modelling group of IPFN since 2005. This group investigates non-linear MHD plasma instabilities and real-time digital signal processing on plasma diagnostics. He is the contact person for the EFDA ITM-TF and MHD Topical Group at the IST Euratom Association. He joined the ITM-TF Leadership as deputy in mid 2008 and is charged with the coordination of the Experimentalists and Diagnosticians Resource Group (EDRG). Rui Coelho’s vita can be found here:

ITM-TF work programme and achievements

The Integrated Tokamak Modelling Task Force was created in 2003 with the task of coordinating the development of a coherent set of European simulation tools to be benchmarked on existing tokamak experiments, with the ultimate aim of providing a validated simulation package for ITER exploitation. ITM-TF is divided into four Integrated Modelling Projects (IMPs) that focus on the following areas of physics: plasma equilibrium and MHD, transport code and whole discharge evolution, transport and micro-instabilities and fi nally, heating, current drive (H&CD) and fast particles. Their work programmes refl ect some of the needs expressed by ITER, most of which require an Integrated Modelling platform:

  • IMP12: Mitigation and avoidance of disruptions (in coordination with EFDA Topical Group MHD and PWI TF), ELMs and other instabilities;
  • IMP3: Pellet injection, ELM triggering;
  • IMP4: Effects of edge turbulence on core energy; momentum, fuel and impurity confi nement; consistent characterisation of turbulent transport over the full wave spectrum;
  • IMP5: Coupling between sources and energetic particle effects; coupling of energetic particle instabilities to MHD.

The “Infrastructure and Software Integration Project” (ISIP) is in charge of the development of the code platform and the implementation of the ITM data-structure and tools. Two additional subgroups coordinated by the TF leaders guarantee the link with the experimentalists and the provision of the experimental databases: The “Atomic, Molecular, Nuclear Surface” Data (AMNS) and the “Experimentalists and Diagnosticians Resource Group” (EDRG). In 2007, the “ITER Scenario Modelling Working Group” (ISM) was established as part of ITM-TF with the aim of assisting the ITER International Organisation in systematic predictive modelling of all reference scenarios, using the major existing integrated modelling tools, whilst the ITM code platform was in development. ISM produc very valuable work in terms of bothcode cross-comparisons and validation and scientifi c publications, addressing a wide range of physics issues for ITER. ISM activity will continue in 2010 in collaboration with ITPA (International Tokamak Physics Activity) and the international fusion community, aiming to include ITM codes and modules as they become available. ITM-TF is working in close collaboration with the EUFORIA – “EU Fusion for ITER application” project towards creating transparent access to capacity and capability computing.

A central ITM-TF project is the development of the European Transport Solver (ETS). The project is an exceptionalone in that ITM-TF usually only coordinates code development. It is motivated by the fact that none of theexisting transport codes meet all of the ITM requirements, namely modularity, fl exibility and standardised interfaces. In terms of the physics, the ETS is designed to solve the standard set of one-dimensional time-dependent equations which describe the evolution of the core plasma, including several ion species (impurities). The solver itself is designed with a modular approach enabling the separationof the physics from the numerics, thereby facilitating the testing/usage of the numerical schemes that best suit a particular scenario.

During the fi rst phase of ITM-TF, surveys and cross-verifi cation of the available models and numerical codes were performed within the individual IMPs and the data-structure was extensively discussed. Equilibrium, linear MHD stability, core transport and RF wave propagation, as well as the poloidal fi eld systems and some diagnostics were the fi rst topics addressed. Data-structures have been fi nalised for these and are being expanded to address, among others, non-linear MHD, edge physics and turbulence as well as neutral beam propagation. Alongside the development of the physics concepts, ITMTF has produced tools to manipulate the data-structure and use it in fully fl exible and modular simulation workfl ows.

Gloria Falchetto holds a permanent position at the CEA Research Institute on Magnetic Fusion (IRFM, Cadarache, France) where she is part of the “Transport, Turbulence and MHD group” which is working on turbulence simulations and modelling. She was appointed deputy leader of ITM Task Force in mid 2009, after being deputy leader of IMP4.

Gloria Falchetto’s vita can be found here:

The ITM-TF has now achieved the development of an initial release version of a fully modular and versatile simulator containing all essential functionalities. The simulator is now ready to be used for the fi rst physics applications. In 2010, the validation of tools can start: the database is being fi lled with machine descriptions (JET, Tore Supra, MAST, FTU, FAST and AUG) as well as experimental data (some discharges of Tore Supra and JET) and modules from the different IMPs are available for integration into the transport solver. The next few years will see the validation of the simulator for a complete discharge based on existing experimental data with the available modules, the integration of more quantitative physics models (“ab-initio”) and the integration of the entire modelling of the device. Gloria Falchetto, IRFM, Rui Coelho, IST, Christian Konz, IPP ITM-TF: Gateway: