Heterogeneous architectures are increasingly common in modern High-Performance Computing (HPC) systems. Achieving high-performance on such heterogeneous systems requires new approaches to application development that are able to achieve the three Ps: Performance, Portability, and Productivity.In this paper, we provide an overview of the state-of-the-art for developing high-performance, portable and productive multi-physics applications with particular focus on the simulation of a plasma fusion reactor. Simulating such a complex system relies on both fluid- and particle-based simulations, and coupling interfaces between these two domains. We also review the current state-of-the-art in reasoning about the performance, portability and productivity of HPC applications.

Accurate computation of power deposition is expected to be critical for the successful design of plasma-facing components (PFCs) in reactor-scale magnetic confinement devices. The work provides analysis and model computations to help explain the high levels of accuracy obtainable using relatively coarse meshing of PFCs to treat cases with sharp-edged shadowing of one PFC by another. It emerges that even small misalignments of the surface grid with respect to the shadow edge or “terminator” may greatly improve the accuracy to which the total power to the tile is calculated, hence a strategy for error control emerges which emphasizes meshing with different constraint rather than different size of triangulation.

This work presents an overview of the integrated strategy developed, as part of the DEMO Key Design Integration Issue 1 (KDII1), to protect the EU-DEMO first wall (FW) from planned and unplanned plasma transients by employing discrete limiters. The present Breeding Blanket (BB) FW design, which aims at minimizing the loss of neutrons while travelling to the breeding zone, is able to withstand steady state heat fluxes up to ≈1-1.5 MW/m² [1], which is not sufficient to guarantee its integrity for most plasma-FW direct contact. This is different from ITER, which has a FW designed for peak heat loads up to 4.6 MW/m2 [2], and it does not have the DEMO BB breeding related requirement. A series of documents was compiled in the DEMO Pre-Conceptual Design Phase, in support of the KDII1. The work presented here was presented at the 2020 DEMO Gate 1 (G1) review, and collects also the comments of the panel and the relative additional studies triggered by them. The design process, presented in this paper was adopted to systematically evaluate the impact of design changes, or new physics inputs, on the FW protection strategy and integration issues. It includes compiling the list of transients, and performing the relative plasma simulations, the design of discrete limiters and the evaluation of their capability to reduce the heat flux density on the FW, and finally a preliminary analysis of the heat loads effects on the Plasma Facing Components (PFC). All these aspects, together with preliminary limiter design, where considered since the beginning, in an integrated way.

The charged particle heat load expected for the DEMO Single Module Segment first wall (FW) during current off-normal plasma scenarios indicates that protection is needed for avoiding/reducing damage to the breeding blanket FW due to the deposition of a huge amount of energy in a small timescale [1]. Within the EUROfusion framework of heat load analysis and design of DEMO wall and FW protections during plasma transients (identified as “Key Design Integration Issue 1” [2]), extensive reworking has led to FW and limiter designs that keep the flat-top maximum heat load on both the FW and limiter plasma-facing surfaces within engineering limits. The limiter strategy appears promising for both normal and off-normal plasma events, therefore the study will be focussed on a FW equipped with limiters. As a continuation of the work started in [3], which has highlighted the weakest point of the older FW design and led to the new FW layout, the impact of misaligned segments and limiters on the charged particle heat flux pattern is investigated for the “limited” FW (i.e. FW protected by limiters). The study is carried out by 3D field line tracing codes SMARDDA/PFCflux [4], [5] and covers normal operation scenarios (ramp-up and steady-state) with the aim of producing heat flux penalty factor distribution to identify the worst case scenarios. As far as the normal transient events are concerned, the results in [3] are updated. In addition, during steady-state operation, deformation of in-vessel components due to mechanical loads such as ferromagnetic forces acting on EUROfer and different thermal expansion of adjacent segments, leads to the exposure of edges that are shadowed in the FW undeformed configuration. As a novel approach, flexible geometrical transformations simulating this kind of normal operation misalignment are implemented for studying the impact on the charged particle heat load of the induced differential deformations.

Within the framework of EUROfusion DEMO First Wall and limiter design activities, the protection of the First Wall against power deposition peaks is being considered. During steady-state operation, the radiative power from the plasma could be considered more uniformly spread on plasma-facing components than the charged particle power deposition. The presence of openings (i.e. gaps between segments and ports) and the introduction of limiters breaks the continuity of the wall and opens the possibility of localized high heat flux values on toroidal-facing gaps due to charged particles striking the wall. The heat flux pattern is likely to be altered by misalignments between components due to manufacturing, assembly or non-uniform/non-symmetrical operational conditions in the segments. In this paper, the 3D field line tracing code SMARDDA is used for studying the impact of misalignment on the heat load distribution for a periodically segmented DEMO First Wall, specifically the Single Module Segment Concept. The present work does not address any study on limiter misalignments, which will be treated separately. The work covers normal operation conditions (ramp-up and steady-state), considering both the cases of bare First Wall (without limiters, as reference) and First Wall protected by limiters. The main aim of the work is understanding how the power deposition on the First Wall (due to charged particles) is affected by the presence of a radial or vertical misalignment between segments and starting a workflow to be applied later to all the possible plasma scenarios and First Wall layout. Heat flux penalty factor maps have been created to identify the worst cases among the ones analyzed. The related heat flux maps are relevant for thermal assessments of a simplified DEMO model to be performed later on, particularly in terms of maximum temperature values for the current DEMO design.

The dynamics at the edge of fusion confinement devices is driven by interchange instabilities and involves the motion of plasma across two regions-the "core region" and the scrape-off layer (SOL)-distinguished by whether field lines are, respectively, closed or connected to the wall. Motivated by this phenomenon, we present an extensive linear stability analysis of a two-layer plasma model encompassing the coupled interactions between the region with closed field lines and the SOL. We focus on the effect of varying the particle diffusivity and ion viscosity, revealing the significant variation in the spatial structure of the critical modes. In addition, we have investigated the dependence of the stability threshold on the ratio of the width of the region with closed field lines to that of the SOL; this dependence is strong when the ratio is sufficiently small, but becomes insignificant once the ratio is of order unity.

The anticipated heat flux limit of the European DEMO first wall is similar to 1 - 2 MW/m(2). During transient and off normal events, the heat load deposited on the wall would be much larger than the steady state heat load and exceed the first wall limit, therefore the breeding blanket first wall needs to be protected. This involves dedicated discrete limiters in certain regions of the machine that would take the brunt of the heat load as well as adequate shaping of the first wall. The current concept envisages limiters at a few (3-4) equatorial ports to cope with the ramp-up of the plasma; upper limiters (in similar to 8 upper ports) are considered for upward vertical displacement events. Two design options have been considered for these limiters: a modular design where the limiter plasma facing components are attached to individual plates that are assembled together so that transient electro-magnetic loads can be reduced, and in case of damage the plates can be replaced/repaired individually; and a divertor-like design where the plasma facing components are attached to a single Eurofer cassette. Other limiters considered include inner wall limiters in case of plasma contraction and lower limiters may be needed for downward vertical displacement events. The thermal hydraulic finite element analysis results show that the integrity of the cooling pipes can be maintained during the anticipated transient events. The limiters are considered to be sacrificial and designed to be replaceable independently from the breeding blanket system. The design has to allow that installation, removal or replacement of the limiters can be performed remotely. Strategy to tackle outstanding issues and required R&D is also discussed.

We present an extensive linear stability analysis of a two-dimensional fluid model used to study the plasma dynamics in the scrape-off layer of tokamaks. The model equations are based on the Braginskii fluid equations under the assumptions of drift ordering and electrostatic plasma. The model also employs the commonly used slab geometry approximation, whereby the magnetic field is assumed constant and straight, with the effects of curvature reintroduced as effective gravitational terms. We study the linear instability in the system by solving a boundary value problem, thereby extending previous studies, which focused on a local analysis. Furthermore, we demonstrate that the governing plasma equations for the scrape-off layer can be viewed as describing a thermal convection problem with additional effects. The new features include a non-uniform basic state gradient, linear damping terms, and additional advective terms. We characterize the conditions at the onset of instability and perform an extensive parameter scan to describe how the stability threshold varies as a function of plasma parameters.

The design of plasma facing components (PFCs) requires knowledge of the charged particle heat load in the scrape-off layer (SOL). Ray-tracing codes like PFCFlux can model this heat load assuming that particles follow the magnetic field lines. Calculations on limiter equilibria underestimate the heat load significantly. In fact, not all the power circulating in the SOL is reported on the wall, with 80% of the total power circulating in the SOL missing in the worst cases. This paper explains why some power is missing in this case, and presents different ways to rescale the heat load to recover all the power coming from the SOL. The maximum heat load on the limiter for a given magnetic configuration can change from 1 MW/m(2) without rescaling to values from 3.5 MW/m(2) to 21.7 MW/m(2) depending on the rescaling method.

Within the DEMO first wall 3D shape design activity, studying the effect of misalignment started in 2017. Such assessments have been conducted in the past for ITER and heat flux penalty factor maps have been created [1], this route could be a feasible approach in the case of DEMO as well. This paper details the methodology that allows the effects of misalignments to be assessed for DEMO. The test cases focus on steady-state plasma operation (start of flat top). The aim is to understand the effect of basic misaligned cases, for example radial protrusion/recession or poloidal rotation of a single module. To do so, particle tracing software codes such as SMARDDA and PFCflux have been used to create heat flux maps that reach the first wall surfaces. These heat flux maps, combined with the specified radiative heat load, are used as input for simplified FE models of the blanket modules. As a result, not only the effect on heat flux, but also on the temperature (and later stress) distributions can be estimated. The paper describes how the obtained results can be implemented in ANSYS in the identified critical cases from the test matrix that has been studied. The results obtained from the nominal heat flux map are compared to the misaligned cases. The mitigating effect of the 3D nature of the heat conduction on the peak temperature is discussed. This work paves the way to assessing more realistic combined misaligned cases (such as misalignment from different thermal expansion, or due to electromagnetic loads etc. of neighbouring blankets) in the future.

Built around the SMARDDA modules for magnetic field-line tracing [IEEE Tr. Plasma Sc. 42 (2014) 1932], the SMITER code package (SMARDDA for ITER) is a new graphical user interface (GUI) framework for power deposition mapping on tokamak plasma-facing components (PFC) in the full 3-D CAD geometry of the machine, taking as input a user-defined specification for parallel heat flux in the scrape-off layer (SOL) and a description of the equilibrium magnetic flux. The software package provides CAD model import and integration with the ITER Integrated Modelling and Analysis Suite (IMAS), parametric CAD components catalogue and modelling, CAD de-featuring for PFC surface extraction, meshing, visualization (using an integrated ParaView module), Python scripting and batch processing, storage in hierarchical data files, with several simulation cases in one study running in parallel and using message passing interface (MPI) for code speed-up. An integrated ParaView module can combine CAD geometry, magnetic field equilibrium, meshes and results for detailed setup analysis and a module is under development for full finite element computation of surface temperatures resulting from the power deposition patterns on 3-D PFCs. The code package has been developed for ITER, but can be deployed for similar modelling of any tokamak. This paper presents and discusses key features of this field-line tracing environment, demonstrates benchmarking against existing field-line tracing code and provides specific examples of power deposition mapping in ITER for different plasma configurations.

The neutral beam power into JET can be increased by increasing the neutraliser gas flow (and gas pressure), hence ensuring a higher neutralisation efficiency. This has the potential to increase the loading on the ion source backplate due to backstreaming electrons. Measurements of the total backplate power loading due to backstreaming electrons and also the arc discharge are presented for the JET EP2 neutral beam injectors as the neutraliser gas flow rate is varied. The measurements are carried out for normal gas delivery operation and for the gas delivery method (grid gas delivery) that is required for the production of tritium beams at JET. Modelling of the backstreaming electron power load and the power distribution is also carried out and the former is compared to the experimental measurements. The backplate loading is more sensitive to the gas flow rate in the case of grid gas delivery. The measurements and calculations can be used to understand the thermo-mechanical performance of the system to balance the benefit of increased neutral beam power against potential increased fatigue and reduced lifetime of the backplate.

The problem of parameter fitting for nonlinear oscillator models to noisy time series is addressed using a combination of Ensemble Kalman Filter and optimisation techniques. Encouraging preliminary results for acceptable sampling rates and noise levels are presented. Application to the understanding and control of tokamak nuclear reactor operation is discussed.

The present DEMO breeding blanked design heat load capability is limited to approximate to 1 MW/m(2) for steady state plasma loading, due to the specific requirements on high neutron irradiation capable materials, and high coolant temperature for efficient energy conversion. While this limit is achievable in nominal conditions in the present DEMO blanket concept designs, the greatest challenges arise from the occurrence of plasma transients. The results of simulations of a number of plasma transients are presented in this paper. 3D field-line tracing codes have been used to analyses the maximum heat flux and energy density for a specific first wall shape design, and optimize it. A scoping study has been performed with the thermal analysis code RACLETTE, using a broad range of transient input heat fluxes, on a series of high heat flux (HF) components concepts with tungsten armor, Eurofer steel or copper alloy as heat sink materials, and helium or water as coolant. The results permit the identification of the operational space of the peak HF density that can be tolerated by different plasma facing components, for the different transient models.

The SMITER Graphical User Interface (GUI) is a simulation framework integrated into SALOME, an open-source platform allowing the combination of numerical solvers from different physical domains. The SMITER GUI provides a user-friendly interface to the SMITER (SMARDDA for ITER) suite which contains modules that allow import of CAD objects, meshing, python scripting, fast and accurate shadowing calculations, and prediction of the power deposition by the plasma on first wall and divertor plasma-facing components (PFCs), together with data analysis and visualization. The SMITER GUI has been further developed and extended with additional features and additions. One significant improvement is the integration of the ITER Integrated Modelling and Analysis Suite (IMAS) with SMITER. IMAS is a framework that is being actively developed and used by the ITER organization and EUROfusion community.The main goal of the integration of IMAS with SMITER was to set the foundations for interoperability between SMITER and the IMAS databases or Interface Data Structures (IDSs) that provide a uniform means of data archival and retrieval within the IMAS framework. This allows SMITER to use the IMAS IDSs to store the meshed geometry needed as input by the SMITER GUI to run the case computations. Moreover, the mesh and output results can be saved directly to IDSs if needed. To provide these new capabilities additional interface SMITER GUI utilities have been developed which provide tools for data conversion and transformation between the IDSs and the mesh files.

In ASDEX Upgrade, the propagation of cold pulses induced by type-I edge localized modes (ELMs) is studied using electron cyclotron emission measurements, in a dataset of plasmas with moderate triangularity. It is found that the edge safety factor or the plasma current are the main determining parameters for the inward penetration of the Te perturbations. With increasing plasma current the ELM penetration is more shallow in spite of the stronger ELMs. Estimates of the heat pulse diffusivity show that the corresponding transport is too large to be representative of the inter-ELM phase. Ergodization of the plasma edge during ELMs is a possible explanation for the observed properties of the cold pulse propagation, which is qualitatively consistent with non-linear magneto-hydro-dynamic simulations.

The work extends the linear fields' solution of compressible nonlinear magnetohydrodynamics (MHD) to the case where the magnetic field depends on superlinear powers of position vector, usually, but not always, expressed in Cartesian components. Implications of the resulting Lie-Taylor series expansion for physical applicability of the Dolzhansky-Kirchhoff (D-K) equations are found to be positive. It is demonstrated how resistivity may be included in the D-K model. Arguments are put forward that the D-K equations may be regarded as illustrating properties of nonlinear MHD in the same sense that the Lorenz equations inform about the onset of convective turbulence. It is suggested that the Lie-Taylor series approach may lead to valuable insights into other fluid models.

In order to have a better closure for magnetohydrodynamic (MHD) equations, a common approach is to obtain the ion fluid pressure tensor by directly computing the moments of an ion distribution function, obtained by a particle-in-cell solver of the Vlasov or Boltzmann equation. This is the so-called hybrid approach. Long MHD simulations are required for problems such as investigating the properties of the sawtooth cycle. In such long hybrid simulations, collisions are required to relax the distribution function after violent MHD events, and to obtain the selfconsistent neoclassical transport. In this paper, we present a new approach to ion self-collisions, based on temperatureand velocity-shifted Maxwellian distributions. It is shown that the approach emulates the effect of the background reaction, without the need to explicitly implement it. Arbitrariness in the choice of the closest Maxwellian is removed. The model compares very well with binary collision Monte-Carlo simulations. The practical implementation as a Fokker–Planck module in a hybrid kinetic/MHD simulation code is discussed. This requires an additional manipulation in order to conserve energy and momentum.

FISPACT-II is a code system and library database for modelling activation-transmutation processes, depletion-burn-up, time dependent inventory and radiation damage source terms caused by nuclear reactions and decays. The FISPACT-II code, written in object-style Fortran, follows the evolution of material irradiated by neutrons, alphas, gammas, protons, or deuterons, and provides a wide range of derived radiological output quantities to satisfy most needs for nuclear applications. It can be used with any END-compliant group library data for nuclear reactions, particle-induced and spontaneous fission yields, and radioactive decay (including but not limited to TENDL-2015, ENDF/B-VII.1, JEFF 3.2, JENDL-4.0u, CENDL-3.1 processed into fine-group-structure files, GEFY-5.2 and UKDD-16), as well as resolved and unresolved resonance range probability tables for self-shielding corrections and updated radiological hazard indices. The code has many novel features including: extension of the energy range up to 1 GeV; additional neutron physics including self-shielding effects, temperature dependence, thin and thick target yields; pathway analysis; and sensitivity and uncertainty quantification and propagation using full covariance data. The latest ENDF libraries such as TENDL encompass thousands of target isotopes. Nuclear data libraries for FISPACT-II are prepared from these using processing codes PREPRO, NJOY and CALENDF. These data include resonance parameters, cross sections with covariances, probability tables in the resonance ranges, PKA spectra, kerma, dpa, gas and radionuclide production and energy-dependent fission yields, supplemented with all 27 decay types. All such data for the five most important incident particles are provided in evaluated data tables. The FISPACT-II simulation software is described in detail in this paper, together with the nuclear data libraries. The FISPACT-II system also includes several utility programs for code-use optimisation, visualisation and production of secondary radiological quantities. Included in the paper are summaries of results from the suite of verification and validation reports available with the code.