DECI 11th Call

Find below the results of DECI-11 (Distributed European Computing Initiative) call.

Projects from the following research areas:

Astro Sciences (6) Bio Sciences (14) Earth Sciences (3) Astro Sciences (5)
Engineering (1) Materials Science (20) Plasma & Particle Physics (3)

Descriptions of projects follow.

Astro Sciences (6)


Resource Awarded: 7 992 000 core hours on EPCC – Archer and EPCC – Blue Joule


Project Title: Explosive Energy Releases in the Solar Corona
Project Leader: Klaus Galsgaard, University of Copenhagen, Denmark
Resource Awarded: 7 560 000 core hours on RZG – Hydra

Prof. Fernando Moreno-Insertis – Instituto de Astrofisica de Canarias, Tenerife, Spain
Daniel Nobrega Siverio – Instituto de Astrofisica de Canarias, Tenerife, Spain
Åke Nordlund – University of Copenhagen, Denmark
The Sun’s magnetic field is generated in the solar convection zone, and from there it spreads into the outer atmospheric region, the solar corona. The convective motions below the solar surface (the photosphere) inject stress into the magnetic field that propagates into the corona, where the magnetic field determines the dynamic. There for, in the corona, the magnetic field continuously tries to evolve towards new force free states. Occasionally, new stable magnetic field configurations can only be reached through the formation of strong, localised electric current concentrations. Such ‘electric current sheets’ are locations in space where fast conversion of magnetic energy into thermal and kinetic energy can take place. The formation of electric current sheets therefore signals the onset of explosive energy release processes, and a detailed understanding of the conditions for forming current sheets is needed to understand where, when and why such explosive events form. Current sheets occur in many different situations, and observationally they give rise to specific localised signatures such as significant plasma heating, systematic high velocity plasma flows, and accelerations of a population of particles to relativistic energies.

In this project we will investigate two special cases of magnetic energy release in the coronal magnetic field. The first case is the formation of jets in ‘open’ magnetic fields in the Sun’s polar regions, where emerging small bipolar field structures interact with the unipolar background field, producing both fast plasma jets shooting far up into the corona, and creating associated hot loop systems. Two very different models for these events exist; one where magnetic flux emergence from the convection zone into the corona drives the process, but also a scenario where the horizontal advection of a previously emerged bipolar region relative the background field drives the interaction. At present only relative simple numerical models of the emergence scenario have been investigated. We intend to expand on the simple models by creating a family of different models, and use the model data to derive (synthetic) observational parameters; this will make direct comparisons between models and observations possible for the first time.

The second part of the project relates to ‘solar flares’, where the intent is to use more realistic models for the magnetic field configurations, based on direct field extrapolations from real observational data. To stress these models in a realistic way we will extract photospheric motion patterns directly from observations. This approach will extend the current, more static and idealised models, providing much more detailed information on the dynamical evolution of the magnetic field, and from this the reason for the fields local collapse into strong electric current concentrations. This will help understanding the collapse process, in relation to both the initial field structure and the imposed driving of the system. The ultimate goal is improving our possibilities to predict locations of solar flares in the complex solar magnetic field; this is in turn an important step in predicting the onset of space weather storms.


Project Title: Dynamical analysis of packed multiplanetary systems with help of MPI/HDF5 numerical framework and task management system Mechanic
Project Leader: Mariusz Słonina, Nicolaus Copernicus University, Centre for Astronomy, Toruń, Poland
Resource Awarded: 8 400 000 core hours on CYFRONET – Zeus BigMem

Prof. Krzysztof Goździewski – Nicolaus Copernicus University, Centre for Astronomy, Toruń, Poland
Dr. Cezary Migaszewski – Nicolaus Copernicus University, Centre for Astronomy, Toruń, Poland
The Kepler photometric mission discovered several low-mass packed multiplanetary systems. Many of them, such as Kepler-11 or Kepler-33 host more than four planets with semi-major axis < 0.5 AU. The dynamical study reveals a complicated dynamical structure with marginal long-term stability, which is most likely maintained by the multiple mean motion resonances between planets. The packed and mostly chaotic architecture of these systems raises the question on mechanisms leading to such configurations. We propose massive numerical study on Kepler systems, that includes statistical search for long-term stable orbital configurations. It will cover the volume of detected systems and help in more precise estimation of the orbital elements, which is critical in terms of determining the long-term stability. To achieve this, we use genetic algorithms (penalized with the fast chaos indicator MEGNO) combined with mapping of the phase-space. This numerical technique is built on top of our new PC/cluster MPI task management system Mechanic and has been utilized and validated in our study of the Kepler-11 system.


Project Title: Streaming instability in global protoplanetary disks
Project Leader: Prof. Michal Hanasz, Nicolaus Copernicus University, Centre for Astronomy, Toruń, Poland
Resource Awarded: 12 400 000 core hours on ICHEC – Fionn-thin and SURFSARA – Cartesius

Kacper Kowalik – Nicolaus Copernicus University, Centre for Astronomy, Toruń, Poland
Dominik Wóltański – Nicolaus Copernicus University, Centre for Astronomy, Toruń, Poland
We live in the times when we discover a new planet orbiting a distant star almost every two days. However, after the centuries of theoretical research we still struggle to explain their origin. Of course we have some clues, some shards of knowledge that, thanks to hard work of many scientists, shed some light on the mysterious ways of how the planets are born. One possible scenario concerns protoplanetary disks so rich in the building material that gravity itself is able to fragment the disk into large blobs of dust and gas that later form giant, Jupiter-like planets. The other theory assumes that planets are built by slow, but steady growth of small dust grains that collide and stick together forming larger and larger bodies up to the point when gravity again starts to dominate further evolution. Although that process can feasibly explain growth up to cm-sized pebbles, the progress is greatly hindered for larger than 10cm particles as they are more likely to fragment during collision than grow. However, supercomputer simulations performed by A. Johansen and A. Youdin found a process – so called Streaming Instability – that can locally enhance concentration of small pebbles to the levels where gravity can squish them into km sized rocks, which we call planetesimals. In order to achieve very high resolution simulations of protoplanetary disks are usually limited to the local approximation of a small box that corotates with the gas on a chosen orbit – so called `shearing box`. This approach neglects various physical conditions like global evolution of disk’s properties. Our project is aimed to investigate both streaming instability and selfgravity in a more complex setup of radially extended disk. The previously performed computer simulations were focused on streaming instability in quasi-global disk, but completely neglected the effects of self-gravity of the dust. Though the analysis of the results clearly shows that concentrations of dust that formed in our simulations were massive enough, so that gravity would start to play a major role in their evolution. With the help of Tier-1 computational resources, we will be able to increase the resolution of our simulations and simultaneously include the effect of self-gravity of the fluid. That will allow us to verify the validity of the proposed theory in a more complex and realistic computational setups.


Project Title: Planck LFI data analysis
Project Leader: Hannu Kurki-Suonio, University of Helsinki, Finland
Resource Awarded: 4 000 000 core hours on CSC – Sisu

Andres Curto – Institute of Physics of Cantabria, Santander, Spain
Simona Donzelli – IASF Milano, Italy
Elina Keihänen – University of Helsinki, Finland
Reijo Keskitalo – University of California, Berkeley, USA
Theodore Kisner – Lawrence Berkeley National Laboratory, USA
Luis Mendes – European Space Agency, SCIOPS, Madrid, Spain
Marina Migliaccio – University of Cambridge, UK
Diego Molinari – IASF Bologna, Italy
Paolo Natoli – University of Ferrara, Italy
Martin Reinecke – Max Planck Institute for Astrophysics, Garching, Germany
Matti Savelainen – University of Helsinki, Finland
Anna-Stiina Suur-Uski – University of Helsinki, Finland
Maurizio Tomasi – IASF Milano, Italy
Jussi Väliviita – University of Helsinki, Finland
Planck is a European Space Agency satellite mission, whose task is to map the structure of the cosmic microwave background (CMB) in unprecedented detail, surpassing the accuracy of previous missions, like the NASA WMAP. The cosmic microwave background is radiation from the Big Bang, and it shows us the structure of the early universe.
Planck constrains cosmological models and examines the birth of large-scale structure in the universe. It is thought that this structure originates from quantum fluctuations in the very early universe during a period of accelerated expansion called inflation, but Planck results are needed for a better understanding of this. The main scientific results from Planck are cosmological, but as a by-product, Planck will also yield all-sky maps of all the major sources of microwave to far-infrared emission, opening a broad expanse of other astrophysical topics to scrutiny. Planck was launched in May 2009. The first cosmological results were published in March 2013; there will be a second release in 2014; and a final third release with the final results from the mission in 2015.
Planck carries two instruments, the Low-Frequency Instrument (LFI) and the High-Frequency Instrument (HFI), utilizing different technologies. It is important to map the sky at many frequencies to be able to separate the cosmic microwave background from the astrophysical foreground radiation. Two Planck data processing centres (DPCs) have been set up, one for each instrument, but the most resource-intensive tasks need to be done on supercomputers.
Because of the weakness of the signal and the high accuracy desired, Planck data analysis is a complicated task, requiring sophisticated statistical methods to separate out the signal from instrument noise and systematic effects.
Simulation work will dominate the computational load of Planck data analysis. Analysis pipelines will indeed be predominantly run on simulated rather than real data since Planck analysis codes either require simulations for self-calibration and validation of results, or depend critically on simulations for the results themselves.
The objective of this project is to carry out resource-intensive tasks that are needed as part of Planck LFI data analysis. These tasks include:
timeline-to-map Monte Carlo simulation of Planck data: thousands of realizations of instrument noise and cosmic microwave background signal; also simulations of astrophysical foreground radiation signal; the analysis of this data in parallel to the analysis of the real data from the sky
cosmological parameter estimation for a number of cosmological models, in particular those related to multi-field inflation.


Project Title: The WHALE project: Wdm HAlos with Lagrangian phase-space Elements.
Project Leader: Dr. Raul E. Angulo, CEFCA, Cosmology, Teruel, Spain
Resource Awarded: 500 000 core hours on RZG – Hydra

In this project we aim to perform high-resolution numerical simulations of a single WDM halo. For this, we plan to use novel and promising numerical methods. In particular, our simulations will feature a recently developed method to compute gravitational forces, which suppresses discreteness noise to a minimum. This characteristic is essential to perform cosmological simulations of structure formation in WDM cosmologies. Extending our approach, we will be able to directly trace the evolution of Lagrangian phase-space volume elements. Our program tests the fundamentals of cosmological simulations, as well as enhances our understanding of halo formation and the implications of a Warm DM particle. Our estimates place the requirements of this project in 2048 CPUs and 1.6 million CPU hours. These expensive computer resources are needed for our novel approaches to succeed and to enable a notable progress in the field of cosmology.

Bio Sciences (14)


Project Title: Molecular mechanism of ATP Binding by Synapsins
Project Leader: Dr Luca Maragliano, Italian Institute of Technology (IIT), , Genova, Italy
Resource Awarded: 1 700 000 core hours on EPCC – Archer

The synapsins are a family of neuronal proteins present in invertebrate and vertebrate organisms. They have been characterized as having a key function in synapse formation and plasticity, in particular as modulators of neurotransmitter release at the pre-synaptic terminal. The available crystal structure of Synapsin I (Syn I) bears strong similarities with a class of ATP-binding enzymes, and several experimental results also point to a role of ATP in Syn function. Presently however the molecular basis of this mechanism are still unknown. In the present project, we propose to characterize in full atomic detail the process of ATP binding by Syn I. Specifically, we will study the ATP dependent protein conformational transition by means of Free Energy calculations in the wild-type molecule as well as in mutants that have shown different ATP affinity.


Resource Awarded: 2 000 000 core hours on CSC – Sisu and CSC – Sisu_XC40


Project Title: Transport Mechanisms in the ClC Channels and Exchangers
Project Leader: Prof. Simon Berneche, University of Basel, BIOZENTRUM, Basel, Switzerland
Resource Awarded: 3 350 000 core hours on PDC – Lindgren

Membrane channel proteins are sophisticated molecular machines which facilitate the transportation of ions, protons, and small molecules across biological membranes. Among them, chloride channels (ClC) are a specific family which transport chloride ions via electrostatic diffusion dependent on both the ionic gradient and transmembrane voltage. Many studies have shown that several ClC proteins are not only simple channels, but also behave as chloride ion/proton exchangers, in which the flux of chloride ions is coupled to that of protons in the opposite direction. However, the proton conduction pathway and the coupled transport mechanism of chloride ions and protons are still poorly understood. The present proposal is to investigate the chloride ion/proton transport mechanism of ClC-ec1. It was shown that the protonation states of E148 and/or E203 play an crucial role in both chloride ion binding and proton transport mechanisms. Here we propose to use classical molecular dynamics (MD) simulation to characterize the protein structure under different protonation states of E148, E203 and other acidic amino acids in the vicinity of the permeation pathways. We will particularly look at conformational changes of the pore that could affect ion binding affinity and eventually favor ion transport. Furthermore, the combined effect of ion binding and protonation of key residues can potentially regulate the formation of water chains likely to act as proton wire. The aim is to identify which conditions favor the formation of stable proton wire. Based on these results, potential of mean force calculations will be performed on our local facilities to characterize the permeation of ions and protons using classical MD and QM/MM simulations.


Project Title: Computer modeling in biomedical engineering
Project Leader: Prof. Nenad Filipovic, University of Kragujevac, Faculty of Engineering, Kragujevac, Serbia
Resource Awarded: 189 000 core hours on ICHEC – Fionn-hybrrid

MSc Tijana Djukic – University of Kragujevac, Faculty of Engineering, Kragujevac, Serbia
This project deals with computer modelling techniques that are applied in biomedical engineering. There are three major fields of interest in this project. The first one is related to the process of atherosclerosis and blood flow simulations, while the other two parts deal with computer aided detection of breast cancer and prediction of cancer behavior.

The first application uses computational fluid dynamics approach to model the blood flow through the arterial tree using geometrical models created from clinical data for specific patients. This application is designed to provide patient-specific computational model of the cardiovascular system and to assist medical staff by providing the possibility to analyze current state of patient’s arteries, possible outcomes of several treatments and select the most appropriate treatment for the specific patient. The second application deals with computer modelling of tumor progression and behavior of tumor cells and tissues under the influence of diverse treatment strategies. Computer simulation tools also uncover a new avenue to optimize and control tumor growth and may have broad implications for the treatment of cancer.

The third application is designed as a tool that can help to improve the detection of abnormalities in breast cancer screening. Breast cancer screening with mammography has been shown to be effective for preventing death caused by breast cancer. This type of software can help the radiologists in the diagnosis process and decrease the number of errors, by indicating suspicious signs, or by classifying lesions in benign or malignant category.


Project Title: Conformational transitions in the ATP-lid of HSP90: Induced fit or conformational selection?
Project Leader: Dr. Yvonne Westermaier, Universidad de Barcelona, Spain
Resource Awarded: 1 500 000 core hours on BSC – MinoTauro

The present study on the ATP-lid of HSP90, a pharmaceutically relevant oncology target is aimed at shedding light on the differences in conformational plasticity in the presence or absence of known fragments and small molecules. Unbiased, atomistic simulations in the upper microsecond range in explicit solvent will be used to generate a large number of conformational ensembles which will serve as a basis to address the conformational preferences in terms of the induced fit or the conformational selection concept. We expect to get useful atomistic insights for designing ligands which are able to “freeze” HSP90 in a particular conformation and therefore have a higher binding affinity and residence time. The project addresses the two extremes which underlie protein-ligand interactions, namely induced fit (1) and conformational selection (2). While in the former, binding is obtained by a specific structural change, the later selects the adequate protein conformation from the unbound ensemble. The two concepts are actively discussed in the scientific community. Interestingly, enzyme selectivity is often achieved via large conformational changes, such as the ones happening in DNA polymerases after nucleotide binding from an open to a closed state, by the kinetic partitioning of the enzyme-bound nucleotide species (3). Similarly, the nucleotide-dependent conformational change governs specificity and analogue discrimination by HIV reverse transcriptase (4). Furthermore, a periplasmic binding protein (PBP) called Lysine-, Arginine-, Ornithine-binding protein, which has more than 100 crystal structures with a huge variety of substrates, can undergo large-scale hinge bending motion from open to closed upon ligand binding (5). In that study, a protein-ligand encounter complex is formed by rigid body rotations. This process is dominated by conformational selection. The protein-ligand interactions then induce conformational changes to reach the fully bound state (induced fit). This duality has been confirmed by NMR for another PBP (6). In our study, we will explore which concept applies better to our test system or whether the two not mutually exclusive concepts are predominant in different situations, starting from different apo and holo Heat Shock Protein 90 (HSP90) crystal structures. A conformational selection or encoding model seems probable, given previous work on the apo, ADP- and ATP-bound N-terminal (NT) HSP90α revealing that functional motions induced by the ligands are already present in the apo MD (7) and the knowledge from crystal structures on the conformational changes in the lid region. Despite the fact that these conformational differences in the HSP90 binding pocket are well established (8), it is still a matter of debate whether these are ligand-induced or encoded in the protein. Targeting the ATPase activity of HSPs is an attractive drug design strategy, given that several of its client proteins influence all the hallmarks of cancer (9). An overwhelming range of molecules directed at HSP90 are currently in development (10-15). We are interested in studying the NT ATP-binding domain of HSP90α because known inhibitors do most often target this domain. Notably, the shape and size of the pocket are governed by a conformational change taking place predominantly in the “ATP-lid” (16), which needs to be considered for drug design purposes (17, 18). Known inhibitors target one of three different conformations of this lid, i.e. either the closed, helical or open one. The helical conformation is of particular interest, as it enables the opening of a hydrophobic subpocket which is exploited by several drug candidates reviewed by Roughley et al. (15). However, this conformation seems to be energetically penalized relative to the open and closed conformations (17). Fluorescence resonance transfer experiments revealed that the transitions between the conformational states and the nucleotide binding/unbinding are mainly thermally driven and that nucleotides can bind to the NT open and closed state without strictly forcing the protein into a specific conformation (19). It remains to be seen how this relates to the binding site opening in the NT monomer, which we study in silico. To understand the nature of this conformational transition, we performed so far relatively long unbiased (over 1 μs) Molecular Dynamics (MDs) on 3 representative apo crystal structures of the NT, which led to a partial conformational interconversion. For obtaining the complete transition from intermediate conformations and for extracting kinetic information of this conformational change using a Markov State Model as described in (20), we would need to run the already existing MDs probably at least until reaching 100 μs. Convergence will be checked periodically while the simulations will be running. These apo simulations will provide an idea on variety of the conformational ensembles for the lid conformations depending on the starting conformation. We expect the large-scale motions in the apo HSP90 to be non-random, similarly to findings for adenylate kinase (21), where such motions preferentially follow the pathways of catalytically competent states, as a hierarchy of substrates (22) in accordance with the underlying energy landscape seems to govern dynamics and binding of enzymes in general. Furthermore, we would like to start the same type of MDs starting from holo crystallographic structures in the helical conformation to be able to compare the stability of the helix in the presence of ligands. Several publicly available crystal structures of diverse ligands and fragments bound to the helical conformation will be investigated. The moieties which bind in proximity to the helix include resorcinols, furane derivatives, pyrazoles or phtalazines. In a first phase, we intend to simulate a HSP90 with a resorcinol-based ligand that we will also test experimentally using kinetic capillary electrophoresis. Subsequently, at least 3 physicochemically distinct fragments will also be simulated bound to the helical conformation of HSP90. Depending on the outcome, holo MDs could then also be extended to both the closed and open staring conformations. In this respect, previous MD work on the NT in presence and absence of the natural substrate showed that functional motions induced by the ligands are already present in the apo form (7), but it is still a matter of debate whether these are ligand-induced or encoded in the protein. Conveniently, a method for quantitatively comparing induced fit (a) and conformational selection (b) was published (23). (a) is represented by the local atom RMSDs (root-mean square deviations) between all unbound and the bound reference state and (b) by the average backbone atom RMSDs. Although we do not intend to simulate binding/unbinding, which would require biased MDs to be carried out to speed up the calculations, we expect to get useful atomistic insights which we will use to design ligands which are able to “freeze” HSP90 in a particular conformation and therefore have a higher binding affinity and residence time. Our work on the conformational plasticity of the NT HSP90 helix will help elucidating the conformational plasticity of this important pharmacological target, providing important hints for inhibitor design.


Project Title: Atomistic simulation of internal protein dynamics and protein-protein interactions under cell-like conditions
Project Leader: Prof. Dr. Matteo Dal Peraro, University of Basel, Basel, Switzerland
Resource Awarded: 8 750 000 core hours on CSC – Sisu

The cytosol is the main compartment inside living cells. It is a very crowded medium where solutes reach concentrations as high as 300 – 400 g/L forcing molecules into unspecific inter actions and repulsions. These perturbations alter macromolecules in many ways, affecting their diffusion, structure, internal dynamics, activity and interaction capabilities. Such effects ultimately affect the equilibria and cellular processes in which these macromolecules are involved, hence the importance of studying these phenomena. This is especially important for proteins, which provide the main machinery that keeps cellular biology running. Molecular dynamics (MD) simulations stand as a very useful tool to model crowded cellular environments, having several advantages over experimental approaches and allowing to probe the effects and responses of individual components at atomic resolution.13 The main drawback of using MD to explore crowding effects is that usually very large systems must be simulated to achieve meaningful results, leading to high computational loads for conventional resources. But with the availability of supercomputing architectures such as those provided by this call, these kinds of studies are becoming feasible. We would now like to capitalize on our and others’ previous works by simulating big systems with crowders of different sizes and varied chemical nature as those found in the cytosol, focusing on their effect on ubiquitin’s internal dynamics and on its capacity to establish interactions with other proteins as observed in physiologically relevant complexes. We will also explore the aggregation potential of the molecules and the stability of these aggregates.


Project Title: Network analysis of protein dynamics
Project Leader: Elena Papaleo, Danish Cancer Society Research Center, Copenhagen, Denmark
Resource Awarded: 1 375 000 core hours on PSNC – Chimera

James Fraser – University of California, San Francisco, USA
Kresten Lindorff-Larsen – University of Copenhagen, Denmark


Project Title: Deciphering the intricacy of electron tunnelling in biological systems
Project Leader: Dr. Jan Rezac, Institute of Organic Chemistry and Biochemistry, Prague, Czech Republic
Resource Awarded: 8 671 622 core hours on PSNC – Cane

Prof. Vincente Moliner Ibanez – Universitat Jaume I, Dep. de Química Física i Analítica, Valencia, Spain
Dr. Aurélien de la Lande – Université Paris-Sud 11 – CNRS, Laboratoire de Chimie Physique, Orsay, France
Electron transfer (ET) is one of the most common chemical reactions encountered in biochemistry, for instance along the respiratory chain, in the photosynthetic system or in numerous enzymes that need inputs of electrons to fulfill their catalytic role. It is fascinating to realize that ETs often occur over large distances (1-2 nm), basically governed by quantum tunnelling. The question when and how the long range electron transfer between two active sites of a protein takes place remains the last missing piece needed for description and understanding of the reaction mechanism in these enzymes. In this project, we focus on non-coupled dicopper monooxygenases and, in particular, on peptidylglycine aplha-Hydroxylating Monooxygenase (PHM). PHM has the remarkable ability to catalyze the hydroxylation of a substrate C-H bond through the activation of a dioxygen molecule with the help of two remote copper active sites that are separated by a solvent cleft. The central objective of the present project is to decipher the interplay between the structure and the dynamics of the PHM enzyme and of the long range electron transfer. We will address this topic by high level quantum chemistry tools. We will use density functional theory (DFT), which is needed to cope with the complexity of the electronic structures of copper centers, within a hybrid DFT/MM simulation protocol in order to accopunt for the dynamics of the protein. To obtain accurate results, high performance computing resources are necessary. Apart from contributing to the knowledge of the mechanism of this particular enzyme (which is still subject of an intense debate), our calculations will also answer more general questions on the role of tunnelling and protein dynamics in enzyme catalysis.


Resource Awarded: 6 187 500 core hours on EPCC – Blue Joule


Project Title: Modeling of action of molecular switches in G-protein-coupled receptors
Project Leader: PhD DSc Slawomir Filipek, Warsaw University, Faculty of Chemistry, Warsaw, Poland
Resource Awarded: 4 000 000 core hours on IPB – PARADOX

G protein coupled receptors (GPCRs), also called 7TM receptors, form a huge superfamily of membrane proteins that, upon activation by extracellular agonists, pass the signal to the cell interior. Biochemical and crystallographic methods together with molecular dynamics simulations and other theoretical techniques provided models of receptor activation based on the action of so-called “molecular switches” buried in the receptor structure. We plan to investigate three types of GPCRs 5-HT1B, SMO1 and FPR1 by molecular dynamics simulations to find out action of molecular switches.


Project Title: Calculation of protein−peptide interactions
Resource Awarded: 1 500 000 core hours on CSCS – Rosa and NIIF – NIIFI SC

Peptides act as partners of proteins in various biological and therapeutic processes. Thus, precise description and prediction of protein−peptide interactions is essential. Application of computational calculations can become a real alternative of experiments in this field. The project aims at overcoming present limitations of calculations of protein−peptide interactions at two important points. The first goal is the development and validation of a precise method for determination of hydration structure of protein−peptide complex interfaces. The second goal is the development and testing of binding thermodynamics calculators of protein−peptide complexes using correctly hydrated structures.


Project Title: Simulating the effect of fiducial markers on high-intensity focussed ultrasound treatments of the prostate
Project Leader: Dr. Jiri Jaros, Brno University of Technology, Department of Computer Systems, Brno, Czech Republic
Resource Awarded: 1 569 000 core hours on WCSS – Supernova

Dr. Dean Barratt – University College London, Department of Medical Physics and Bioengineering , London, UK
Dr. Ben Cox – University College London, Department of Medical Physics and Bioengineering , London, UK
Vojtech Nikl – Brno University of Technology, Department of Computer Systems, Brno, Czech Republic
Dr. Bradley Treeby – University College London, Department of Medical Physics and Bioengineering , London, UK
Ondrej Vysocky – Brno University of Technology, Department of Computer Systems, Brno, Czech Republic
Drahoslav Zan – Brno University of Technology, Department of Computer Systems, Brno, Czech Republic
Prostate cancer is one of the most common cancers for men in Europe and a leading cause of cancer- related death. For patients with early-stage disease, the cancer is often treated via the insertion of small permanent radioactive seeds (a procedure known as brachytherapy), or using external beam radiation therapy. The latter procedure usually also involves implanting a small number of gold markers into the prostate to verify the position of the gland between treatments. Unfortunately, for 1 in 4 patients undergoing these treatments, their cancer will recur. However, for some of these patients, further treatment using a different therapy technique is still possible (this is called salvage therapy). In this regard, high-intensity focused ultrasound (HIFU) has shown a lot of promise. HIFU works by sending a focused beam of ultrasound into the prostate using a transrectal ultrasound probe. At the focus of the beam, the acoustic energy causes sufficient heating via acoustic absorption to kill cells in a localised region (and thus treat the tumour), while the surrounding tissue is left unharmed. The use of HIFU as a salvage therapy for prostate cancer has previously been reported in a number of clinical studies. However, the impact of implanted markers and brachytherapy seeds on the efficacy of the treatment is still largely unknown. The objective of this project is to use numerical simulations to investigate and quantify the effect of implanted markers on prostate HIFU treatments. Specifically, a numerical model of wave propagation and heat transfer will be used to systematically quantify the effects of individual implanted markers, including the shape of the ultrasound beam, the predicted region of treated tissue, and the treatment time. These will be the first large-scale simulations of this kind, as previous acoustic models have been unable to deal with the computational complexity of simulating these effects. The model (called k-Wave) is based on an efficient k-space pseudospectral discretization of the nonlinear continuum equations. Spatial gradients are computed using the Fourier collocation spectral method (which is based on the FFT), while the existence of an exact solution to the linearized wave equation is exploited to improve the accuracy of computing temporal gradients. Overall, this allows for a significant reduction in both the number of time steps and the number of grid points per wavelength required for accurate simulations. The code is parallelised using MPI, which partitions the 3D domain across multiple nodes and uses a parallel 3D FFT routine. Each time step requires fourteen global 3D FFTs, and a large simulation may have thousands of time steps, consuming up to 100,000 core hours.


Project Title: Examining the structure of the intrinsically disordered syndecan’s ectodomain by molecular mechanics simulations
Project Leader: Hector Martinez-Seara Monne, Tampere University of Technology, Finland
Resource Awarded: 5 000 000 core hours on EPCC – Archer

Reinis Danne – Tampere University of Technology, Finland
Tomasz Rog – Tampere University of Technology, Finland
Syndecans constitute a major family of heparan sulfate transmembrane proteoglycans which are crucial in cell-cell and cell-matrix interactions. Four syndecan forms (or paralogs) exist in vertebrates, whose expression is tissue specific and developmentally regulated. Most of the cells express at least one syndecan form. Their pathogenesis includes wound healing disabilities, obesity, cancer proliferation, and promotion of infectious diseases. Syndecans are composed of a signal peptide and three core units: cytoplasmic, transmembrane and ectoplasmic domains. Syndecans in their mature form are always covalently attached to glucosaminoglycans in which several ligand binging sites are located. Although the molecular structures of cytoplasmic and transmembrane domains are known, the structure of the ectodomain remains elusive.

In this project we want to prove that the ectodomain is intrinsically disordered driven by the presence of the highly charged and long glucosaminoglycans unbranched polymer (20-60 nm). In order to achieve this goal we will simulate by means of molecular dynamics two form of the syndecan family, syndecan-2 and syndecan-4, in their native environment, i.e. immersed in a lipid membrane. In each simulation the ectodomain will have three glycosaminoglycans attached, which is the maximum occupancy for the chosen syndecans. These simulations are the first ever to include simultaneously three of the most biologically relevant molecular classes: lipid (membrane), protein and sugar polymers (glycosaminoglycans). These atomistic simulations spanning into the microsecond time scale and covering over two million atoms will provide detailed information of syndecan structure, which is crucial to find molecular target for drug development. This is of key importance because syndecan plays a major role in the development of several pathologies. Additionally, this work is essential for any future attempt to tackle syndecan-related biologically processes like the signal transduction which is known to involve oligomerization.


Resource Awarded: 10 000 000 core hours on CSCS – Rosa

Earth Sciences (3)


Project Title: Addressing the nanoscale complexity of clay-water interfaces in MD simulations: Effects of the substrate compositional disorder on the properties of surface species
Project Leader: Dr Andrey Kalinichev, Ecole des Mines de Nantes, SUBATECH UMR6457, Nantes, France
Resource Awarded: 6 300 000 core hours on VSB-TUO – Anselm

Dr. Narasimhan Loganathan – Ecole des Mines de Nantes, SUBATECH UMR6457, Nantes, France
Brice Firmin Ngouana wakou – Ecole des Mines de Nantes, SUBATECH UMR6457, Nantes, France
Adjunct, Postdoctoral researcher Marek Szczerba – Institute of Geological Sciences, Krakow Division, Krakow, Poland
The primary research focus of our project is quantitative molecular-scale investigation of the effects of the adsorption and transport of cationic radionuclides in clay and cement materials by means of classical molecular dynamics (MD) simulations. This choice is based on the understanding that the behavior of hydrated cations in clays has already been studied by several research groups in significant detail. However, these previous simulations typically covered just a 18 Å x 20 Å surface area which had substantially limited the accurate representation of the entire complexity and diversity of surface adsorption sites, including substitution disorder and local inhomogeneity of surface charge distributions. We have now developed a new set of larger-scale (~55 Å x 65 Å) muscovite and montmorillonite models, which were systematically constructed to introduce as much structural disorder in the distributions of the Al/Si and Mg/Al substitutions as possible. This brings the modeling of the clay surface complexity to the qualitatively new level. In addition, we have developed completely new realistic models of so-called interstratified illite/smectite – one of the important components of the COx clay of the French nuclear waste repository. The ionic adsorption on this complex clay material cannot directly predicted as simply a linear combination of adsorption on illitic and smectitic components, and has not been modeled before. CLAYFF force field allows us to systematically study all these effects of structural disorder and inhomogeneity on the local structural and dynamic properties of the interfacial and interlayer ions and H2O molecules and to quantify these effects for a number of monovalent and divalent cations relevant to radioactive waste disposal and storage.


Project Title: Ensemble numerical weather prediction of heavy precipitation events: Tests of the Meso-NH system in the framework of the `Distributed Research Infrastructure for Hydro-Meteorology` (DRIHM) EU project.
Project Leader: Dr. Evelyne Richard, CNRS & Université Paul Sabatier, Laboratoire d’Aérologie, Toulouse, France
Resource Awarded: 3 500 000 core hours on CYFRONET – Zeus BigMem

Dr. Juan ESCOBAR MUNOZ – CNRS & Université Paul Sabatier, Laboratoire d’Aérologie, Toulouse, France
The Mediterranean region is susceptible to heavy precipitation and severe flooding. These events can be very devastating in the densely-populated coastal regions and need to be better forecasted. The current proposal aims at running an ensemble prediction system (EPS) based upon the Meso-NH model. The Meso-NH EPS system specifically addresses the uncertainties associated with the model representation of physical processes such as turbulence and cloud microphysics. This work is carried out in the framework of the EU DRIHM project which aims at developing a prototype e-Science environment to provide an easier access to hydro-meteorological data and models, and facilitate the collaboration between meteorologists, hydrologists.


Project Title: Super Computing of Extreme Natural Events
Project Leader: Dr Elisabetta Fiori, CIMA Research Foundation , Savona, Italy
Resource Awarded: 2 400 349 core hours on EPCC – Blue Joule

The World Conference on Disaster Reduction (Hyogo, 2005) defined among its thematic priorities the worldwide improvement of cooperation in HydroMeteorology Research (HMR) activities for the prevention and mitigation of risk associated with the occurrence of severe hydrometeorological events. This statement was confirmed at the joint press conference of the Center for Research on Epidemiology of Disasters (CRED) with the United Nations International Strategy for Disaster Reduction (UNISDR) Secretariat, held on January 2009, where it was noted that flood and storm events are among the natural disasters with major impact on human life. Damage caused by extreme precipitation events strongly burdens the budgets of industry, national governments, and international organizations. People affected by severe precipitation events often face economic ruin. This led European and worldwide national authorities to answer to the fundamental question of civil protection organizations on dealing with flash-flood and debris flow: Do we have enough tools to predict when and where heavy rain will impact on small and medium size catchments in complex orography areas? This challenging task requires focusing on the understanding and prediction of severe meteorological phenomena through the use of Numerical Weather Prediction (NWP) models. As known from the literature, numerical models include uncertainties deriving from insufficient and/or incorrect initial conditions, omission or inadequate representation of sub-grid scale processes, errors related to data assimilation techniques, or even poor knowledge of process interaction and parameterization.

Engineering (5)


Project Title: Future-proof High Performance Numerical Simulation for CFD with FEASTFLOW (2)
Project Leader: Prof. Stefan Turek, TU Dortmund University of Technology, Institute for Applied Mathematics, Dortmund, Germany
Resource Awarded: 1 500 000 core hours on CINECA – PLX

This DECI11 project is a follow-up to the recently completed DECI8 project with acronym FFF. Here we again aim at the development of efficient, reliable and future-proof numerical schemes for the parallel solution of partial differential equations (PDEs) arising in industrial and scientific applications. Here, we are especially interested in technical flows which can be found in a wide variety of (multi-) physics problems. In our approach, both numerical and hardware efficiency are addressed simultaneously: On the one hand, the transition of today’s computational hardware towards parallel (and heterogeneous) architectures is in progress and therefore all levels of parallelism (vectorisation, parallelism on the core level in multi- and many-core CPUs and accelerator devices like GPUs and finally on the node level within distributed memory clusters) have to be exploited. Algorithms and whole solvers have to be tailored with respect to the target hardware in order to achieve a significant amount of the parallel peak performance. On the other hand, the sole concentration on hardware efficiency does not carry out the whole job (and in some cases may be counter-productive): Numerical efficiency plays a crucial role and itself includes multiple levels that can be optimised. Starting with the overall numerical and algorithmic approach required for the solution of a given domain specific problem (i.e. discretisation of the governing equations in time and space), stabilisation, linearisation of non-linear problems and finally the solution of the linear problems and smoothing therein, all these aspects together with the aforementioned levels of parallelism bear a large amount of interdependencies.


Project Title: Fully Localised Edge States in Boundary Layers
Project Leader: Dr. Philipp Schlatter, KTH, Department of Mechanics, Stockholm, Sweden
Resource Awarded: 12 500 000 core hours on EPCC – Archer

Dr. Yohann Duguet – CNRS, LIMSI, Orsay, France
Bruno Eckhardt – Philipps-Universität Marburg, Germany
Prof. Dan Henningson – KTH, Department of Mechanics, Stockholm, Sweden
Understanding the appearance of turbulent spots in boundary layer flows is of crucial importance in order to maintain the viscous drag at low levels, e.g. for flow control. In this proposal we focus on a boundary layer flow held parallel by homogeneous wall suction. The aim of this European collaboration is to compute the coherent structures acting as precursors for turbulent spots, the recently discovered “edge states”. Edge states correspond to nonlinear equilibria of the Navier-Stokes equations lying at the dynamical border between laminar and turbulent flow in phase space. Investigating their spatial structure, their temporal dynamics and their stability characteristics as the Reynolds number varies will lead to an accurate description of the general conditions under which a boundary layer flow can transition to turbulence. The focus here, contrarily to former edge states studies, lies in the use of an extended numerical domain in order to catch the full spatial localisation of the corresponding coherent structures. The use of a well-tested and highly parallel spectral code, combined with an iterative bisection strategy, makes this search highly feasible. However, the localisation also requires large computational domains which justifies the need for PRACE Tier-1 resources.


Resource Awarded: 2 519 041 core hours on PDC – Lindgren and PDC – beskow


Project Title: Hybrid RANS/LES model for discontinuous Galerkin finite element numerical code
Project Leader: Dr Antonella Abba, Politecnico di Milano, Dipartimento di Ingegneria Aerospaziale, Milano, Italy
Resource Awarded: 2 000 000 core hours on NIIF – NIIFI SC


Project Title: Using LES for studying the Precessing Vortex Core in gas turbine combustors
Project Leader: Dr. Ferry Tap, Dacolt International BV, , Maastricht, The Netherlands
Resource Awarded: 3 500 000 core hours on SURFSARA – Cartesius

Prof.Dr. B.J. Boersma – Delft University of Technology, The Netherlands
Dr. Giel Ramaekers – Dacolt International BV, Maastricht, The Netherlands
Modern gas turbines for power generation or aircraft propulsion are constraint by legislation to produce less NOx emissions. Such low NOx combustion systems have a very narrow operating window where the flame is burning in a stable way. Flame stability is achieved by introducing a strong swirling motion to the combustion air entering the combustion chamber. This swirling motion leads to strong recirculating flow patterns inside the combustion chamber, including a flow structure called the Precessing Vortex Core (PVC). This PVC interacts with the flame; such interactions in gas turbine combustors are a trending topic in current gas turbine research. It is believed that the PVC is a key phenomenon to understand in order to design stable low NOx combustion systems.

Dacolt has used Computational Fluid Dynamics (CFD) tools to study a model gas turbine combustor for which detailed experimental data is available. Both cold flow and reacting (hot) flow simulations have been performed. It was found that, at minimum, the unsteady Reynolds-Averaged Navier-Stokes (RANS have to be solved using the Reynolds Stress Model (RSM) for the turbulent viscosity in order to capture the PVC. A next level of refinement was achieved by using Detached Eddy Simulation (DES), allowing to capture more of the turbulent velocity fluctuations. Still, significant deviation from experimental results in terms of velocity, mixing and species profiles is observed. DES has shown to predict better results than URANS RSM.

A next level of refinement is to use Large Eddy Simulation (LES) techniques. It is estimated that such advanced turbulence modelling should be able to fully describe the flow dynamics and hence PVC behavior. However, LES simulations are at least an order of magnitude more expensive in terms of computational effort as compared to URANS or DES. Such computational resources are far beyond the capacity of typical resources of engineering companies like Dacolt. A typical LES simulation on the burner studied to date is estimated to require approximately 50.000 CPU-core hours.

The aim of this project is to gain further insight in the PVC characteristics and its interaction with the flame. From such in-depth understanding an engineering methodology can be derived to do practical combustor simulations. It is proposed to proceed in three steps, involving in total an estimated 20 simulations:
1) Use LES to study the PVC flow structure in great detail (9 simulations): assess the effect of subgrid turbulence models, grid discretization and other numerical parameters
2) Characterize PVC / flame interactions (6 simulations): perform cold and hot flow simulations for three reference flames (stable and unstable) for which experimental data is available and study the interactions between the flame and PVC.
3) Distill an engineering CFD methodology (8 simulations): investigate how the high-fidelity LES set-up can be simplified to be affordable for engineering purposes, without losing key information. This work would be carried out by Dacolt engineers who are already familiar with the burner CFD setup Also it is planned to work with Delft University of Technology, by defining an MSc thesis assignment on especially the fundamental first step.


Informatics (1)


Project Title: Towards Web-Scale Semantic Parsing
Project Leader: Prof. Dr. Oepen Stephan, Department of Informatics, University of Oslo
Resource Awarded: 5 000 000 core hours on UIO – Abel

Materials Science (20)


Project Title: Ab-initio investigations of electronic states and transport phenomena in carbon-based nanostructures with spatial inhomogeneities
Project Leader: Tomasz Slusarski, Adam Mickiewicz University, Faculty of Physics, Poznań, Poland
Resource Awarded: 750 000 core hours on FZJ – JuRoPA

Tomasz Kostyrko – Adam Mickiewicz University, Faculty of Physics, Poznań, Poland
The subject of the studies will be carbon nanostructures and simple carbon-based molecules (like hydrocarbons). The materials of interest includes, among others, graphene, carbon nanotubes, fullerenes, and carbon-based chains (polyynes, cumulenes and alkanes). Our studies will be performed with the help of the advanced ab initio code, which was specifically designed to deal with the systems including thousands of atoms in the unit cell and for this reason it is very effective tool for analysis of the nanostructures. The main objective of the project will be a systematic study of possibilities of controlling the electronic states and the transport properties of such nanostructures by external electrostatic potential fields or defects introduced purposefully into the nanostructures. We will also study the influence of intrinsic defects that can appear in the carbon-based nanostructures e.g. during their synthesis (vacancies, substitutional impurities, Stone-Wales defects etc.). We plan to describe and systematize the effects of the interplay of various symmetry limitations on the electronic and transport properties of the studied materials. We will try to relate the transport characteristics to the actual defect configurations in the junctions involving the nanostructures and by these means we hope to obtain the indications for the experimental methods of detecting and identifying various defects in the materials of interests.


Project Title: Molecular-level understanding of CeO2 as a catalyst for partial alkyne hydrogenation
Project Leader: Dr Ganduglia-Pirovano Maria, Centro Superior de Investigaciones Científicas (CSIC), Spain
Resource Awarded: 100 000 core hours on FZJ – JuRoPA

Dr. Javier Carrasco – Consejo Superior de Investigaciones Científicas, Instituto de Catálisis y Petroleoquímica, Madrid, Spain
Dr. David López Durán – Consejo Superior de Investigaciones Científicas, Instituto de Catálisis y Petroleoquímica, Madrid, Spain
Recently, a high catalytic performance of pure ceria for the partial hydrogenation of alkynes to olefines has been reported [1]. This type of reaction, widely exploited in steam crackers for purification of olefin streams as well as in the manufacture of fine chemicals, is conventionally carried out over palladium based catalysts. With a propene (ethene) selectivity of 91% (81%) at a high degree of alkyne conversion and stable behaviour, pure ceria is one of the most efficient catalysts ever reported for these industrially relevant reactions. The project aims to elucidate the structure and functioning of pure ceria as catalyst for partial alkyne hydrogenation. The fundamental understanding of the partial and total acetylene hydrogenation reaction mechanisms on CeO2(111) as a prototype reaction, coupled to the detailed analysis of the configuration and properties of the active sites, is expected to help identify the key properties of a successful new catalytic systems for partial alkyne hydrogenation. The expected intense interplay between theory and our experimental collaborators at ETH-Zurich will allow a greater opportunity to explore implications of the results for technological relevant applications. The calculations employed the DFT+U approach as implemented in the VASP code. Transition structures along reaction pathways are searched using the nudged elastic band method. The use of PRACE Tier-1 infrastructure or equivalent is mandatory to perform the calculations proposed here due to the high computational requirements of NEB calculations, and the large number of (relatively large size) systems to be consider. [1] G. Vilé, B. Bridier, J. Wichert, and J. Pérez-Ramírez, Angew. Chem. Int. Ed. 51, 8620 (2012).


Resource Awarded: 4 207 500 core hours on EPCC – Blue Joule


Project Title: Electric and electromagnetic field effects on nanomaterial-protein systems
Project Leader: Niall English, University College Dublin, School of Chemical and Bioprocess Engineering, Dublin, Ireland
Resource Awarded: 4 492 800 core hours on CINECA – PLX

It is proposed to study both thermal and non-thermal effects of static electric and electromagnetic (e/m) fields on nanoscale systems of interest to biology and nanomedicine, using non-equilibrium molecular dynamics simulation approaches developed previously by the PI. This will extend our understanding of important potential applications for nanoparticle-protein interactions and binding, e.g. mechanisms for thermal ablation mediated by e/m heating of nanotubes, and of molecular and ionic transport through nanochannels in bio-membranes. Such computational projects could be used to identify preliminary strategies to, for instance, promote drug delivery, inhibit cancer cells via thermal ablation, or influence nanoparticle-protein interactions. In particular, molecular simulation will offer a unique insight at the atomistic level of how static electric and e/m radiation affect nanobiological structures and systems. In addition, simulation constitutes a very cost-effective way of assessing project feasibility, without the need to develop expensive prototype equipment. This simulation of moderately large systems for, crucially, long timescales will necessitate the use of GPU computing for maximum impact.


Resource Awarded: 9 744 000 core hours on EPCC – Archer


Project Title: Trapping of He and H on dislocations and grain boundaries in W and W-Ta alloys: an atomistic study
Project Leader: Dr. Dmitry Terentyev, SCK-CEN, Institute of Nuclear Materials Science, Mol, Belgium
Resource Awarded: 425 000 core hours on EPCC – Archer

Experiments involving direct contact of plasma with material at the sub-threshold implantation conditions (i.e. plasma energy is too low to generate Frenkel pairs) do reveal that essential amount of plasma components is stored in subsurface grain. Conventionally, the trapping is associated with the presence of natural defects such as vacancies. However, recent experimental investigations of trapping and release of D in pure Tungsten (W) and tungsten-tantalum (W-Ta) alloys show that there is a considerable amount of trapped D in the bulk up to several microns depth, where no vacancies can be expected since their thermal concentration at the implantation temperature ( The investigation of atomic-scale mechanisms of diffusion and trapping of Hydrogen and Helium is very important as they are the main plasma components in Nuclear Fusion Devices, in Tungsten(W) and WTa alloys which are the main candidates for first wall and divertor materials in DEMO. Wa-Ta random alloy is believed to be more attractive solution than pure polycrystalline W as Ta is suggested to improve the mechanical strength and ductility of W. However, the impact of Ta on H/He retention is not well known up to now. A limited number of experimental data shows that, in general, Ta suppresses the blistering in high temperature/high flux exposure. Whereas, in the low temperature low flux range, addition of Ta leads to a higher retention. Hence, there is a need to understand physical processes laying behind these interplay of Ta effects.
Here, we characterize dislocations and grain boundaries as possible traps for plasma components. We apply density function theory (DFT) calculations to characterize the interaction of He and H with a screw dislocation and high angle tilt grain boundaries in W and W-Ta dilute alloys. The obtained data can be used in coarse grain models (such as Rate Theory or Object Kinetic Monte Carlo) to account for accommodation and detrapping of D and He from the considered lattice defects. Thus, the present project will directly contribute to the part of experimental programme focusing on the assessment of retention and plasma-induced surface modification of W and W-Ta alloys, carried within the EFDA framework.


Project Title: Graphene-Ferroelectric Interfaces: a first-principle investigation
Project Leader: Dr. Zeila Zanolli, Forschungszentrum Julich, Institute of Advanced Simulation, Julich, Germany
Resource Awarded: 3 600 000 core hours on PDC – Lindgren

Graphene is a perfect infinite single layer of sp2-bonded carbon atoms densely packed into a benzene-ring structure. The confinement of electrons in two dimensions and the peculiar symmetry of the carbon network give graphene exceptional electronic properties that make it a promising material for carbon-based nanoelectronics and spintronics. In particular, the performances of such devices rely on the exceptional intrinsic carrier mobility of graphene. However, extrinsic scattering sources due to standard SiO_2 substrates limit the mobility. Hence, the quest for alternative substrates is mandatory in order to increase the mobility beyond the extrinsic limits. Among the possible candidates, ferroelectric (FE) substrates are the most promising due to their ultrahigh dielectric constant and hysteretic dielectric response to an electric field. The first would results in reduced scattering at the graphene-substrate interface and could explain the observed ultrahigh mobility in non-suspended graphene. The latter could be responsible for the nonvolatile memory achieved in top-gated graphene-ferroelectric FET devices. In addition, polarization domains can be written in FE materials, resulting in a substrate with tunable periodic potential that would allow the engineering of the electronic properties of graphene without etching. Last but not least, magnetism could be induced in graphene by a magnetic FE substrate. The objective of the present proposal is the theoretical investigation of graphene-FE systems in order to understand the physical phenomena occurring at the interface between the two materials and exploit this knowledge to propose new electronic and spintronic devices. Due to the nanometric size of the systems under investigation quantum-mechanical simulations with atomistic resolution, as first principles Density Functional Theory (DFT) techniques, is the method of choice to achieve an accurate description of their properties.


Project Title: Graphene precursor phase and graphene-SiC(000-1) interface by first principles thermodynamics and a massively parallel structure search
Project Leader: Dr. Volker Blum, Max Planck Society, Fritz-Haber-Institute, Berlin, Germany
Resource Awarded: 14 700 000 core hours on RZG – Hydra and VSB-TUO – Anselm

Graphene films grown by Si sublimation on SiC are among the most promising materials combinations for future graphene applications based on existing semiconductor technologies. To refine the growth quality and the resulting electronic properties (insulating, metallic, electronically doped etc.) of these films, it is essential to understand the exact atomic structure of the interface. On the so-called `Si face` of SiC crystals, the relevant interfaces are, in fact, largely understood, but the same is not true for other SiC surfaces. In particular, the `C face` (SiC(000-1) in crystallographic notation) yields electronically very promising few-layer graphene films. Growing similar monolayer graphene films would be desirable, but in this regime, a phase mixture with other surface phases is observed. The structure of some of the competing surface phases as well as the atomic structure of the SiC-graphene interface on this substrate is unknown. In this project, we will perform a massively parallel, exhaustive first-principles structure search (density functional theory including van der Waals effects) and use `first principles thermodynamics` to clarify: (i) the atomic structure of the most important unknown interface phase, (ii) the nature of the most favorable SiC-graphene interface structure, (iii) the phase competition that hinders the monolayer growth process. As a result, we expect a significantly improved understanding of why achieving large-area monolayer graphene films is difficult on SiC(000-1) with present methods, and how this difficulty caneventually be overcome. The project can only be performed with computational resources on the scale of PRACE-DECI, and, if successful, will result in a showcase for the power of first-principles computational materials predictions today.


Project Title: High-Throughput Materials Thermal Conductivity Computations
Project Leader: Chercheur Natalio Mingo, CEA-Grenoble, LITEN, Grenoble, France
Resource Awarded: 8 200 000 core hours on PSNC – Chimera and UIO – Abel

Dr Jesús Carrete Montaña – CEA-Grenoble, LITEN, Grenoble, France
The thermal conductivity of materials plays a fundamental role in many applications. Heat dissipation in microelectronics, for example, has been identified as the major bottleneck towards further miniaturization of electronic components. Finding ultra-low thermal conductivity semiconductor compounds is also a primary goal of current research on thermoelectric energy conversion, which is one of the most promising avenues for thermal energy scavenging. In the search for better compounds with low thermal conductivities, trial and error experimentation is costly, time consuming, and likely to miss out many potentially good material compositions. This is where high-throughput ab initio calculations can make a big difference. The same idea, when generalized to all properties of materials, is behind the creation of databases such as, managed by Prof. S. Curtarolo and coworkers, with which the present project is affiliated. However, when compared to other properties of materials that have been treated with high-throughput approaches, the computational cost of rigorously computing the thermal conductivity of even one compound is huge. The main limiting factor in the process is the calculation of the set of second- and third-order interatomic force constants needed as input, which can total 150,000 CPU hours even for a binary compound. We have developed and tested a software package (HiThruPack) able to run such calculations of the lattice thermal conductivity tensor of crystalline material compounds and alloys, end-to-end, in an automatic fashion and without any adjustable parameters. This package combines automatic usage of VASP as a calculation backend, custom code for harnessing the symmetries of the compound ensuring a minimum number of VASP invocations, a cutting-edge solver of the Boltzmann transport equation for phonons (wlsBTE) and a significant body of automation code to streamline the whole process. This makes it possible to perform systematic searches for new materials with specific thermal transport properties, rapidly exploring the chemical phase space to identify best material candidates. This in turn shall save precious experimental resources and accelerate the pace of novel materials discovery. The purpose of this project is to calculate the thermal conductivity of all hitherto synthesized binary and ternary cubic crystalline compounds with an entry in the ICSD database using this package, taking ab initio thermal conductivity calculations to a scale never attempted before and eventually delivering a major contribution to the field.


Project Title: Identifying the intermediate pathway for deNOx on the Rh {111} surface: (NO)2 or N2O?
Project Leader: Prof. Graeme W. Watson, Trinity College Dublin, Dublin, Ireland
Resource Awarded: 3 264 731 core hours on PDC – Lindgren

Global concerns over atmospheric pollution and the enhanced greenhouse effect has led to a search for new and improved technologies for pollution abatement. In particular, the emission of nitrogen oxides (NOx, x=1, 2) from combustion engines are known to be a large contributor to acid raid in the atmosphere and smog formation at ground level causing both environmental damage and human health concerns. Governmental regulations have thus been imposed to reduce NOx emissions, necessitating the development of technologies that decompose NOx (deNOx) into more environmentally benign N2. These technologies include, most notably, the three-way catalytic converter for automobile catalysis and selective catalytic reduction for industrial diesel engines. Rhodium, palladium and platinum metal clusters embedded on oxide supports are typically employed as the catalytic agent in these technologies, with Rh being the most effective metal for the deNOx process. As of 2007, 81% of the world’s Rh metal supply was used in deNOx technologies, severely depleting the ore, making it scarce and expensive. There is thus a need to understand why the deNOx properties of Rh metal are enhanced relative to other materials, in order to develop low cost catalysts that can emulate the properties of Rh. The purpose of this study is to investigate the most favourable pathway for deNOx on the Rh {111} surface and the intermediates involved. This will be achieved with density functional theory calculations. The study will begin by examining the most stable adsorption configuration for N, O, NO, NO2, N2O and (NO)2 on the Rh {111} surface and use these as initial points in determining their most stable co-adsorbed combinations. The energies obtained will be used to establish the elementary steps of the deNOx pathways. Through the use of nudged elastic band calculations,the energy barriers between these steps will be calculated to locate the lowest energy and hence most favourable pathway on the Rh surface. Identifying this decomposition pathway will answer a debate in the literature over whether N2O or (NO)2 is involved. By establishing the most favourable pathway, the ease of the deNOx reaction on Rh metal can be determined and compared to Pd and Pt. The results of this work will ascertain the desirable chemical properties for deNOx reactivity and hence help in the design of novel deNOx catalysts.


Project Title: Magnetic phenomena in molecular rings
Project Leader: Prof. Dr. Grzegorz Kamieniarz, University Poznan, Department of Physics, Poznan, Poland
Resource Awarded: 1 200 000 core hours on CASTORC – Cy-Tera

Molecular magnetism is a rapidly expanding field of research whose central theme is the design and study of magnetic molecules and materials with tuneable properties. Molecule-based metallic clusters exhibit many interesting features, which make them promising objects of theoretical and experimental research. They behave like individual quantum nanomagnets, displaying quantum phenomena on macroscopic scale. In view of potential applications of such materials in magnetic storage devices or in envisaged quantum computer processor as well as the low-temperature refrigerants, the simulation precision of their properties becomes the key issue. The goal of the project is: a) to calculate with high accuracy the total energy for all the symmetry allowed non-equivalent spin configurations for the chromium-based molecules and their twins with different spacers, using the DFT approach; b) to verify the reliability of the models, at least for the Heisenberg terms, using the first principle DFT electronic structure calculations and applying orbital effects and different exchange-correlation potentials or performing geometrical structure optimisation on isolated and/or on surface.


Project Title: NanoMaterials for Improved Li-ion Batteries
Project Leader: Prof. Dr. Fjelvaag Helmer, University of Oslo, Department of Chemistry, Oslo, Norway
Resource Awarded: 3 500 000 core hours on UIO – Abel

Dr. Vajeeston Ponniah – University of Oslo, Department of Chemistry, Oslo, Norway


Project Title: Understanding Surface and Interface Kinetics at the Atomic Scale in chalcopyrite solar cells, in particular the role of sodium
Project Leader: Dr. Guido Roma, Universität Mainz, Institut für Anorganische Chemie und Analytische Chemie, Mainz, Germany
Resource Awarded: 2 196 000 core hours on SURFSARA – Cartesius

Dr. Letizia Chiodo – Italian Institute of Technology, Rome, Italy
Elaheh Ghorbani – Universität Mainz, Institut für Anorganische Chemie und Analytische Chemie, Mainz, Germany
Dr. Janos Kiss – Universität Mainz, Institut für Anorganische Chemie und Analytische Chemie, Mainz, Germany
Dr. Hossein Mirhosseini – Universität Mainz, Institut für Anorganische Chemie und Analytische Chemie, Mainz, Germany
Thin film solar cells based on copper/indium/gallium/selenide chalcopyrite semiconductors as light absorbers (known as CIGS) are rapidly growing in the market. Further optimisation of their efficiency and cost is necessary and should be possible, but experimental approaches based on trial and error are lengthy and expensive. CIGS solar cells are made by stacking several layers of different materials, all of which play an important role. This proposal focuses on the metallic (molybdenum) back contact and its interface with the CIGS absorber. Its goal is to unravel the mechanisms at the atomic scale which influence the formation of the metal/CIGS interface during the CIGS deposition and in particular the catalytic role of sodium, which is known to be beneficial for the performances of the device, also by enhancing the formation of intermediate MoSe2 layer. The project will rely on computer simulations at the quantum level, starting from standard density functional theory to more advanced exchange-correlation functionals including van der Waals interactions and some hybrid functional and GW calculations. The goal is to study, on the one hand, the absorption and migration of relevant elements (Se, Na, O) on the metallic surface at various coverages; on the other hand, models of the Mo/MoSe2 interface will be proposed and investigated by the same tools. Following a multi-scale approach, the results of the quantum simulations will be used in a model for surface thermo-kinetics in order to determine the most important mechanisms in the growth process, those that to control the quality of the back contact, in particular with respect to the morphology of the MoSe2 layer.


Project Title: First-principle study of oxynitrides for white LED applications
Project Leader: Prof. Xavier Gonze, Universite catholique de Louvain, IMCN – NAPS, Leuven, Belgium
Resource Awarded: 5 200 000 core hours on CYFRONET – Zeus BigMem

Ir. Samuel Ponce – Universite catholique de Louvain, IMCN – NAPS, Leuven, Belgium
This work aims at improving our understanding of the physical phenomena that govern the temperature induced loss of emissivity in white-LEDs (wLEDs) devices. White light from wLEDs is produced by down converting the near-UV/blue light emitted by a InGaAs diode into visible light through the use of different layers of materials, called phosphors.
In this study, we focus on phosphors containing barium, silicon, oxygen and nitrogen, that can be used as a green-emitting layer. Such compound must be doped with rare-earth elements (here Europium) to allow the down converting process to occur. A high energy photon from the diode is emitted and will excite an electron from the doped valence band maximum to an exited state of the bulk system. This electron will lose some energy as heat and then return to its ground state while emitting a lower energy photon (in the visible range). For some materials the loss at working temperature (round 150°C) can be as high as half the nominal intensity at room temperature. The loss is believed to be determined by an ionization process from an energy level attached to the dopant, to the conduction band of the host, that happens if the excitation energy is small compared to the typical thermal energy.
To confirm this empirical concept, we propose to conduct a first-principle study on two phosphors with similar chemical composition but drastically different emission and thermal behavior properties. The most widely use technique is called Density Functional Theory and his based on the resolution of the many-body system in term of the density only. This technique has known limitations and in particular in its inability to describe excitated states properties correctly. We will therefore rely on a many-body scheme based on the Greens functions.
Probably the most widespread approximation is called GW where the self-energy is expressed as the product of the Green function and the screening. Nonetheless exciton (electron-hole interaction) cannot be describe with such an approximation. We will therefore use a modified approach using GW with an atomic pseudopotential containing a hole. This is the so-called “core-hole” approach that allows for the treatment of the electron-hole interaction. To achieve this goal we will use the free software (GPL license) “Abinit”. This software has its core developers located in the Université catholique de Louvain (UCL) in Belgium. It is a robust and long term code used by many scientist across the world (more than 2000 citations of the articles describing the project and its evolution).


Project Title: Surface Modified Metal Oxide Photocataysts
Resource Awarded: 2 799 360 core hours on WCSS – Supernova

The Photocatalyst project will undertake first principles density functional theory (DFT) simulations of new visible light activated photocatalyst materials based on semiconducting metal oxides, TiO2, ZnGa2O4 and Bi2WO6. Photocatalysis has the potential to be an extremely important technology for energy generation, through water splitting and CO2 reduction, and depollution, using only sunlight and cheap, widely available materials as a feedstock, while utilising earth abundant, non-toxic materials as catalysts. In the project work the DFT simulations will examine the formation of the interface formed between metal oxide nanoclusters deposited at surfaces of the parent oxides, which thmeselves are only active photocatalysts in the UV region and this work will allow us to design materials compositions that will shift light absorption into the visible region, while minimising charge carrier recombination that reduces efficiency. The simulations will be atomic models of surface modified TiO2 ZnGa2O4 and Bi2WO6 and will allow us to obtain a detailed understanding of how surface modification and interface formation determines the energy gap and changes photocatalytic activity. In addition, we will examine how the presence of defects and doping can be used to modulate the light absorption properties and charge transport. Comparison will be made with ongoing experimental work with the groups of Profs. Tada (Japan) and Gray (USA) on these novel materials systems.


Project Title: Spin-phonon coupling in multiferroic orthochromites
Project Leader: Dr. Sc. Igor Lukačević, University J. J. Strossmayer , Department of Physics, Osijek , Croatia
Resource Awarded: 1 750 000 core hours on PSNC – Chimera

Post-doc Sanjeev Kumar Gupta – Michigan Technological University, Department of Physics, Houghton, USA
Although ferroelectric ferromagnets, a very rare multiferroic materials, have an extremely interesting and useful property of simultaneous existence of both ferroelectricity and ferromagnetism, their spontaneous polarizations or magnetizations are smaller by a several orders of magnitude compared to those of ferroelectrics or ferromagnets. This way they are unable to compete with useful ferroelectrics and ferromagnets in applications. However, recently theoretical and experimental investigations had shown a way by which one could turn a material into a strong ferroelectric and ferromagnet. These studies propose strain as a parameter by which one could simultaneously control multiple order parameters. This method opens a route to high-temperature implementations of strong ferromagnetic ferroelectrics. The base of this approach lies in the spin-lattice coupling mechanism. The magnitude of this mechanism can be described by an spin-phonon coupling constant. On the other hand, despite the property of magnetic frustration, the magnetic spinels generally show magnetic ordering. This ordering comes from the coupling of spins to other degrees of freedom like orbital ordering or lattice distortions. Coupling can relieve the frustration, making the materials’ ordering dependent on the coupling strength with phonon displacements. Recent experimental and theoretical studies have confirmed these facts in some spinel oxides and showed that spin-phonon coupling is the dominant factor driving the material through a phase transition. Very recently Raman experiments demonstrated spin-phonon coupling in rare-earth orthochromites RCrO3 with magnetic R-ion (R = Y, Lu, Gd, Eu, Sm). Along with phonon frequency changes, a decrease in phonon lifetimes was observed, suggesting phonon mediated magnetic interactions between R3+ and Cr3+ ions. These interactions along with the softening of magnetic R-ion modes suggest that the origin of ferroelectric polarization in rare-earth orthochromites may be related to the displacement of the R-ion. Our results will confirm and make a further in depth understanding of the experimentally suggested mechanism in orthochromite family of materials and open the door to higher-temperature implementations of strong ferromagnetic ferroelectrics, which would allow for dramatic improvements in numerous electrically and magnetically controlled multifunctional devices and applications.


Project Title: Predicting Light-induced Processes in Artificial Photosynthesis from Light Absorption to Charge Separation
Project Leader: Dr. Francesco Buda, Leiden University, LIC (Leiden Institute of Chemistry), Leiden, The Netherlands
Resource Awarded: 2 014 200 core hours on ICM – Boreasz

MSc. Thomas Eisenmayer – Leiden University, LIC (Leiden Institute of Chemistry), Leiden, The Netherlands
MSc. Adriano Monti – Leiden University, LIC (Leiden Institute of Chemistry), Leiden, The Netherlands
Computational screening of artificial photosynthetic modules in order to assess their potential as light absorbers, charge separators or catalysts is an important tool towards the production of solar fuel. In photosynthesis, charges that are stable on catalytic timescales are generated from sunlight at near quantum unity efficiency.

In previous work we identified the mechanism and reaction coordinate of directional charge separation in the bacterial reaction center using ab-initio molecular dynamics (AIMD) in the ground and excited state. Vibrational coherences were found to induce partial charge transfer already in the ground state and to break the symmetry between cofactor chains in the excited state along an anisotropic normal coordinate. With constrained density functional theory (CDFT) we calculated the energy required for the formation of the first charge separated state and found it to decrease along the predicted anisotropic coordinate. Together these findings provide basic guiding principles for the in silico development of novel artificial photosynthetic devices. In a recent study we propose a molecular rectifier designed to achieve a stable charge separated state and characterized its electron transfer process with the time dependent DFT (TD-DFT) method at the CAM-B3LYP level and CDFT/B3LYP.

Through this analysis we were able to predict the optical behaviour of the system and describe in detail the electronic intermediate states that precede the formation of a stable charge separated state. In a subsequent study we explored the possibility to screen different artificial photosynthetic module in terms of the lifetime of the charge transfer state using the nonadiabatic restricted open-shell Kohn- Sham (ROKS) formalism with a Tully surface hopping algorithm. This led to the unexpected result that fixing certain degrees of freedom can increase the charge transfer lifetime by at least an order of magnitude without having to increase the donor-acceptor distance. Based on the achieved promising results and understanding, we aim in this project to extend our analysis to include also real time electron charge evolution initiated by light absorption and coupled to the nuclear motion. For this purpose we plan to perform Ehrenfest molecular dynamics with a TD-DFT propagation scheme to simulate in real-time the evolution of the electron density following a light-induced perturbation. However, this type of calculation even for systems of limited size is computationally very demanding and beyond the capabilities of our local computational facilities. Access to large supercomputing facilities is essential for the feasibility of this project.


Project Title: Spin-dependent thermal properties from first principles
Project Leader: Prof. Matthieu Verstraete, University of Liege, Physics, Liege, Belgium
Resource Awarded: 3 900 000 core hours on EPCC – Archer

The identification of alternative and renewable sources of energy is one of the most important challenges modern society faces, and has become more urgent and intense in the past few years. One of the most promising technologies is that of thermoelectric (TE) devices, which allow one to transform heat into electrical energy (an alternative energy source) or vice-versa (for refrigeration or heating). For the moment TE applications are restricted to niche products in domains such as aerospace, the military, or automobile industries. The efficiency of thermoelectric materials themselves will have to be roughly doubled before cost-effective, large-scale applications can be envisaged.

New perspectives on thermoelectrics have been opened recently by their structuring on the nanoscale. This has allowed experimentalists to obtain impressive efficiencies in thin film samples in the lab, but optimizing these materials and transferring these new ideas to a nanostructured bulk material suitable for industrial environments and mass production remains a challenge. Since the beginning of the nano revolution in thermoelectrics, theoreticians (in particular GD Mahan and MS Dresselhaus) have played a central role, proposing new material and device paradigms, and explaining how the intrinsic limits of bulk thermoelectric materials can be overcome or bypassed. Most of these theoretical works have however been performed at a semi-empirical level and typically include severe approximations such as a constant relaxation time for electrons and for phonons. The central challenge of the present project is to concentrate all of the tools necessary to achieve predictive characterization of thermoelectric materials from first-principles calculations.

This application pushes forward the advances in materials studies we have achieved over the past 3 years. It combines the competencies of the two groups in the Physics Department of the University of Liège working on the transport properties of metals and intermetallics. We will apply our novel methodology to calculate the thermoelectric properties of magnetic and heavy element systems. These include many unexplained physical phenomena and in particular the spin-Seebeck effect (appearance of a spin current due to a gradient in temperature). Using the same physics and some of the same elements, we will study disordered and complex Tellurium-based alloys, which are the industry standard materials for thermoelectric applications at room temperature (300-400K), and show great potential under nanostructuring and heavy-element doping.

The groups of M. Verstraete and J.-Y. Raty have complementary expertise, both in software development and in materials modelling, metallic, intermetallic, and nanostructured systems; they have specific competencies in the domains of the ab initio prediction of vibrational properties, as well as electron-phonon coupling, both of which are central to thermoelectric properties. A fundamental methodology has been developed for the fully first principles prediction of thermoelectric qualities of magnetic materials within density functional theory, in an effort led by Prof M. Verstraete. The methodology will be applied to nanostructured and disordered intermetallic systems, which are the specialty of Dr J.-Y. Raty.


Resource Awarded: 9 062 500 core hours on EPCC – Archer


Project Title: Unbiased Materials Design
Project Leader: Dr. Miguel Marques, Université Lyon 1 and CNRS, Institut Lumière Matière, Villeurbanne, France
Resource Awarded: 7 500 000 core hours on EPCC – Archer

Dr. Maximilian Amsler – University of Basel, Department of Physics, Basel, Switzerland
Dr. Silvana Botti – Université Lyon 1 and CNRS, Institut Lumière Matière, Villeurbanne, France
PhD student Tiago Cerqueira – Université Lyon 1 and CNRS, Institut Lumière Matière, Villeurbanne, France
Dr. Stefan Goedecker – University of Basel, Department of Physics, Basel, Switzerland
Prof. Fernando Nogueira – Universidade de Coimbra, Center for Computational Physics, Coimbra, Portugal
PhD Student Rafael Sarmiento – Université Lyon 1 and CNRS, Institut Lumière Matière, Villeurbanne, France
During the past years we have seen a tremendous development in the field of materials science. Not only we have better theories, capable of describing and predicting more accurately physical properties, we also have more sophisticated computer codes, and faster supercomputers. All these sparked a paradigm shift that is nowadays occurring: instead of studying the properties of materials we know, why not design new materials having the properties we desire. This new field, labelled Materials Design, usually starts from a database of known materials (we currently know around 30 000 well characterized inorganic solids) and uses artificial intelligence algorithms to devise new materials. In this project we propose a different approach, namely a completely unbiased optimization framework that allows for the design of new materials without any a priori knowledge. It is based on a combination of genetic algorithms with global structural prediction methods to obtain the ground-state crystal structure for given compositions. With these tools we will try to obtain new superhard materials, but the sheer amount of information stemming from our runs will advance considerably our current knowledge of inorganic solids. To perform our optimization, we estimate that we need around 5 million CPU hours, and the whole workflow involves commercial and home-grown codes.

Plasma & Particle Physics (3)


Resource Awarded: 6 250 000 core hours on ICHEC – Fionn-thin


Project Title: Full-f gyrokinetic simulations of LOC/SOC transition in Ohmic tokamak discharges
Project Leader: Timo Kiviniemi, Aalto University, Espoo, Finland
Resource Awarded: 4 000 000 core hours on PDC – Lindgren

Jan Westerholm – Åbo Akademi University, Turku, Finland
Fusion energy has been intensively investigated worldwide for decades and so far the tokamak magnetic confinement device has been the most successful concept for it. A major setback however for creating fusion conditions in a tokamak is caused by turbulent transport. During the last decade the first principal gyrokinetic code Elmfire has been developed to study turbulent transport from first principle basis with the Particle-In-Cell method while solving for the full gyrokinetic ion and drift kinetic electron distribution which allows for investigations of the complex interplay of the turbulent fluctuations with the large-scale flows driven by the background profiles and the fine-scale zonal flows driven by the turbulence itself.
In the present proposal, the Elmfire code will be used to study the toroidal rotation reversal observed in many tokamak machines after the electron density has been ramped up in an ohmic plasma. At low electron densities the core rotation is measured in the co-current direction and it reverses to counter-current when the density is increased above a certain threshold. At this same density threshold the poloidal asymmetry of the impurity density is observed to change from up-down symmetric to up-down asymmetric and a reduction in the turbulence density fluctuation level is found. The reversal density is strongly correlated to the density at which the Ohmic L-mode energy confinement changes from the Linear to the Saturated Ohmic Confinement regime (so called LOC/SOC transition). The Elmfire code is able to simultaneously evaluate turbulence, mean equilibrium flows and fine-scale zonal flows while taking impurities into account and is therefore able to incorporate all the above mentioned processes into one simulation.
To reduce the computational expenditure discharges from the smallest machine for which the LOC/SOC transition has been observed, Alcator C-mod, will be taken as the basis for this work. The simulation region will be limited to the core and the electrostatic assumption with a fixed magnetic field background will be employed. The radial grid resolution will be limited to the ion Larmor radius. Collisions are modeled by a momentum and energy conserving binary collision model between all particle species. Steady state profiles will be obtained by including realistic heat and particle sources and sinks. The simulations will be performed in two parts where first the neoclassical equilibrium will be obtained by removing modes with toroidal mode number unequal to zero after which the filtering will be switched off and turbulence will be allowed to develop. To obtain a density fluctuation level of the order of 1 per cent 3000 particles of each species per grid cell will be required on average.


Project Title: Global Gyrokinetic Simulation of Tokamak Plasmas
Project Leader: Prof. Hans Nordman, Chalmers University of Technology, Department of Earth and Space Sciences, Göteborg, Sweden
Resource Awarded: 5 000 000 core hours on FZJ – JuRoPA

Dr. Luis Fazendeiro – Chalmers University of Technology, Department of Earth and Space Sciences, Göteborg, Sweden
Fusion energy is one of the few resources which can provide reliable and inherently safe base load power. With the dangers of serious climate change becoming more evident and undoubtedly linked to anthropogenic greenhouse gas emissions resulting from fossil fuel usage, it is vital to fully explore all viable energy production alternatives. The next step on the way to achieving the goal of clean, almost unlimited fusion energy is the ongoing construction of the major experimental facility ITER in Cadarache, France. This is a major undertaking by the European Union, China, India, Japan, Russia, South Korea and the United States, which will test several of the technical features required for a power plant and demonstrate fusion as a sustainable energy source for the future. With the advent of the ITER device fusion research has entered a new major phase, in which increasing effort is being undertaken on understanding and modelling the complex nonlinear phenomena occurring in a hot fusion plasma. The aim is to achieve an improved predictive capacity. In particular, the characterization and control of plasma turbulence and turbulent transport mechanisms of heat and particles (both main ions and impurities) is one of the most central issues and needs to be validated in conjunction with existing experiments. The behaviour of impurities is of particular importance for ITER, whose divertor and first wall will very likely consist of some mixture of beryllium, tungsten and perhaps carbon. The two first materials have been tested through the new wall installed on the Joint European Torus (JET), the ITER-like wall experimental programme, which alread y produced many encouraging results. Following previous work done by our group [1-10], we will study turbulent transport in tokamak devices, via first principles gyrokinetic simulations, extending our approach from local (flux tube) to global simulations. Our main goals are to compare the results obtained in global simulations with experimental data from the new JET ITER-like wall discharges, in order to accurately predict turbulent transport in tokamaks and feeding into current and urgent design questions for ITER. This work will build on the group’s work in recent years with gyrokinetic simulations [1-10] and the comparison of this (kinetic) approach with computationally efficient fluid models for the better understanding of transport in fusion devices.