DECI 13th Call

Find below the results of DECI-13 (Distributed European Computing Initiative) call.

Projects from the following research areas:

Astro Sciences (8) Bio Sciences (7) Earth Sciences (2) Engineering (7)
Informatics (2) Materials Science (26) Plasma & Particle Physics (5)

Descriptions of projects follow.

Astro Sciences (8)


Project Title: Dynamics of Stellar Systems and Dwarf Galaxies in the Local Group
Project Leader: Prof. Rainer Spurzem, University of Heidelberg, Germany
Resource Awarded: 464,625 standard core hours

Dr. Shuo Li, Chinese Academy of Science, NAOC, Beijing, China
Dr. Yohai Meiron, Peking University, KIAA, Beijing, China
Dwarf galaxies are the most numerous galaxies in the Local Group and in the universe. Some of them belong to the most metal-poor, oldest and least luminous galaxies known. There is now a very good observational census of many Local Group dwarf galaxies for which global parameters have been measured, and it seems these galaxies extend in some way known scaling relations (like e.g. Faber-Jackson) for larger galaxies. Also for a couple of objects even high resolution star-by-star observations have been obtained for excellent color magnitude diagrams, with red and blue horizontal branches and even some blue straggler candidates. Also it has been detected that there is a smooth transition from dwarf galaxies (many of them dark matter dominated) to massive globular clusters, which have no dark matter. We want to study the dynamical history of the transition objects between globular clusters and dwarf galaxies by direct N-body simulation, which has been so far not be feasible for these massive objects. With the new GPU accelerated NBODY6++GPU code, in conjunction with the GPU accelerated series expansion code ETICS (both published) we can provide first fully self-consistent models for the lifetime of the objects. Our adaptation of the special software package GALEV allows us to generate observational data (colors in any pre-defined window, spectra) from our models just as if our supercomputer would be an astronomical instrument. Full simulations of the early formation phase using gas and stellar dynamics with feedback and star formation will be used in a later phase of the project to provide more sophisticated initial models with complex star formation histories.


Project Title: Disk galaxies with RAdiation and MAtter interactions
Project Leader: Dr. Karl Joakim Rosdahl, Leiden University, Leiden Observatory, Leiden, The Netherlands
Resource Awarded: 8,000,000 standard core hours

Dr. Valentin Perret, University of Zurich, Zürich, Switzerland
Prof. Joop Schaye, Leiden University, Leiden Observatory, Leiden, The Netherlands
Prof. Romain Teyssier, University of Zurich, Zürich, Switzerland
An understanding of the complicated problem of the formation and evolution of galaxies requires the use of three dimensional hydrodynamical simulations. Simulation work has added tremendous insight into the interplay of processes which regulate the growth of galaxies. Yet many problems remain unaccounted for, the most impending one being that the conversion from gas to stars is too fast in simulations compared to observations. The inclusion of radiation feedback, i.e. the interaction of radiation from stars with the gas that fills the galaxies and their environment, is thought an important piece of the puzzle in solving the problem and gaining further understanding. Yet, in part due to the complexity and computational cost involved, simulation treatments of radiation hydrodynamics (RHD), needed to model this interaction, have been largely out of reach (although recent simulation work has seen a rise in variously well motivated non-RHD approximations of radiation feedback).

The researchers have recently developed a fast and robust RHD version of the widely used hydrodynamical simulation code RAMSES. With the code, they will perform simulations of high-resolution star-forming galaxy disks in order to understand the role of radiation feedback, originating from dense sub-structures, in regulating the collapse of gas to stars and in generating galactic outflows.

In a previous project, the researchers have already performed studies of radiation feedback in disk galaxies, finding that stellar radiation has an important role in suppressing star formation via photoionisation, while radiation pressure plays a negligible role. With this proposed project, they want to follow up on the previous simulations with increased resolution, which allows them to study the possible impact of multi-scattering radiation pressure in dispersing dense and opaque sites of star formation, that were previously not properly resolved. In addition, they will study the impact of radiation feedback in starbursts associated with galaxy mergers, which were commonplace in the early Universe.

The output of this work will be important in understanding the processes that regulate the growth of galaxies. At the same time it will help in assessing the results of recent cosmological hydrodynamical simulation projects, which have been able to reproduce many observables of galaxies under the yet unconfirmed assumption that radiation couples very efficiently with inter-stellar gas and plays a major role in preventing it to form stars and in generating galactic outflows.


Project Title: Galaxy clusters in the virtual laboratory
Project Leader: Dr. Elke Roediger, University of Hull, Department of Physics and Mathematics, Hull, UK
Resource Awarded: 2,375,000 standard core hours

Prof. Brad Gibson, University of Hull, Department of Physics and Mathematics, Hull, UK
We propose for 950,000 CPU-hrs of simulation time for two projects in our research program on the growth and evolution of galaxy clusters: (i) We will perform a systematic study of triple galaxy cluster mergers and produce multi-wavelength mock observations from them. These mock observations will serve as a reference for various observed complex cluster and group mergers, and will form the basis of our future work on cluster galaxy evolution and particle acceleration in clusters. (ii) We will run tailored simulations of the motion of the elliptical galaxy NGC 1404 through the Fornax cluster with particular focus on its wake. Having pinned down the dynamic context of NGC 1404’s wake, we will study gas mixing in the wake or its suppression by viscosity. In direct comparison to our deep X-ray observations, we will measure the ICM viscosity in the Fornax cluster.


Project Title: Galactic Chemodynamics in the Era of Gaia
Project Leader: Prof. Brad Gibson, University of Hull, Department of Physics and Mathematics, Hull, UK
Resource Awarded: 8,640,000 standard core hours

Dr. Christopher Few, University of Exeter, School of Physics, Exeter, UK
Dr. Robert Grand, Heidelberg Institute for Theoretical Studies GmbH, Heidelberg, Germany
Dr. Jason Hunt, University College London, London, UK
Dr. Daisuke Kawata, University College London, London, UK
Dr. Chiaki Kobayashi, University of Herdfortshire, Centre for Astrophysics Research, Hatfield, UK
We propose to run two families of simulations of Milky Way-type systems, as a pathway to the goal of more realistic Milky Way simulations; the codes employed are our particle-based GCD+ and GADGET3, and our new chemistry-enhanced, grid-based code, RAMSES-CH. One family is based upon “zoom” style cosmological simulations, for which we will run 5 simulations with different merger histories and environments, and explore how these characteristics impact upon the formation and evolution of each galaxy’s sub-components, including their stellar and gaseous discs, bulge, and stellar and hot coronal halo. The other family is a controlled simulation of the Galactic disc; using high resolution simulations, we will study the formation processes of the detailed structures in Milky Way-type galaxies, such as spiral arms, and their effects on gas and stellar dynamics and star formation. Our codes are unique in terms of their self-consistent treatment of chemical evolution in non-linear dynamical simulations, which allow them to predict the distribution of elements in stars and gas as a function of position, kinematics, and age, allowing their comparison directly with observations. The associated chemical information will allow us to self-consistently generate mock observational data, using our stellar population synthesis package, linking directly to the anticipated returns from ongoing and future experiments, such as ESA’s Gaia and the Gaia-ESO Survey, and directly compare with the observed kinematical and chemical properties of stars in the Milky Way. The mock observational data and simulated outputs will be made publicly available.


Project Title: High-resolution cosmological hydro-simulations of galaxy clusters
Project Leader: Dr Veronica Biffi, INAF – O.A. Trieste, Italy
Resource Awarded: 3,600,000 standard core hours

With this project we propose to carry out hydrodynamical simulations of galaxy clusters, at intermediate-high resolutions, to study the formation and evolution of galaxies and their interaction with the hot intra-cluster medium. To this scope, we will employ the TreePM-SPH code Gadget3, in a modified version that includes an improved hydrodynamical scheme for better describing gas mixing. The simulations will also include the treatment of various physical processes, comprising models for star formation and chemical evolution, galactic SN-driven winds and a novel AGN feedback scheme.

As a first goal, we aim at constraining the effects of resolution on the novel SPH scheme and on the AGN feedback description, exploring the paramter space and identifying the best calibration for the improved resolution. As a second target, we want to use the final set of simulations to study in more detail the formation of the brightest central galaxy and its interaction with the surrounding ICM, at different epoch of the cosmic history. Ultimatly, this set of simulated clusters, and the galaxy

population residing in them, would constitute the first simulated sample at this resolution, taking into account a cosmological context and including an unprecedented variety of hydrodynamical processes. This will represent therefore a unique chance for comparing simulation results to the increasing amount of observational evidences provided by current and up-coming missions in various wavelenghts.


Project Title: Hybrid 3D simulations of turbulence and kinetic instabilities at ion scales in the expanding solar wind
Project Leader: Dr Luca Franci, Universita di Firenze, Italy
Resource Awarded: 3,200,000 standard core hours

Turbulence in magnetized collisionless plasmas is one of the major challenges of space physics and astrophysics. Both the anisotropic flow of energy toward smaller scales (cascade) and the energy damping of energy at dissipative scales are poorly understood. The solar wind is the only astrophysical turbulent plasma in which observations and simulations can be directly compared. However, solar wind turbulence evolves with distance because of its spherical expansion: wavevectors tend to align with the radial direction and the components of turbulent fluctuations are damped at different rates. Ultimately, expansion drives anisotropies of the cascade and of the velocity distribution functions of particles, the latter triggering kinetic instabilities. Moreover, both the characteristic proton scales and the plasma beta change with distance, so different kinetic effects may arise at different scales, modifying the nature of the cascade.

Magneto-hydrodynamics (MHD) simulations have shown that turbulence is not universal and may develop different anisotropies, and that expansion drives spectral and component anisotropies that match those observed in the solar wind. On the other hand, high-resolution 2D hybrid simulations, recently performed by the group, showed different spectral indices and different coupling of the primitive fields in the kinetic range compared to fluid scales, indicating a change of turbulent regime at sub-proton scales. The kinetic processes responsible for this change of the turbulent regime, as well as the scale at which it occurs (spectral break), are still strongly debated. Moreover, 3D simulations are mandatory to overcome the limitations of a 2D geometry.

We will perform high-resolution three-dimensional (3D) hybrid particle-in-cell numerical simulations of decaying turbulence, in order to characterize its properties and compare them with solar wind observation. We will vary the plasma beta and the proton temperature anisotropy in order to explore the “states` naturally assumed by the plasma as it expands into the heliosphere. Finally, by employing the Hybrid Expanding Box (HEB) model, we will follow the evolution of turbulence with distance as it is self-consistently driven by the spherical expansion.

This will be the first 3D study of the properties of turbulence capturing fluid-like and kinetic scales simultaneously. A large spectral range is indeed essential to establish an inertial range already at fluid scale, so that turbulence at kinetic scale is fed by a cascade and not by (arbitrary) large-scale initial conditions. Moreover, the 3D-HEB simulation will represent the first consistent study of the evolution of turbulence in 3D accounting for the effect of expansion and of kinetic physics.

We will determine how kinetic turbulence depends on large-scale conditions and the plasma beta. We expect to identify the physical process(es) responsible for the switch of turbulent regime(s) at sub-proton scales along with the associated scale(s). Finally, we will understand to which degree the expansion determines the observed properties of solar wind turbulence. Our data will help interpreting observations from forthcoming space missions, e.g., Solar Orbiter and Solar Probe Plus.


Project Title: Studying the IGM in Different Galactic Environments
Project Leader: Dr. Britton Smith, University of Edinburgh, School of Physics, UK
Resource Awarded: 8,625,750 standard core hours

Prof. Sadegh Khochfar, University of Edinburgh, School of Physics, UK
At the present time, only about 10% of normal matter actually resides in galaxies. The rest exists in the gas in between galaxies, known as the intergalactic medium (IGM). This material can be fully accounted for in the distant, early universe through observations of absorptions lines from hydrogen in the IGM. However, at present day, roughly 50% of all normal matter is in a nearly undetectable state and appears to be missing. Simulations have shown that this material is still in the IGM, but has been heated by shocks to temperatures high enough to ionize hydrogen, thus making it unobservable via hydrogen absorption. The only way to observe this material is through absorption of heavy elements that are capable of remaining not totally ionized in these high temperatures. We will use cutting edge cosmological simulations to study how these heavy elements evolve in the IGM (their temperatures, abundances, ionization states) and the degree to which they can be used to detect the missing material in the IGM. This work will provide a theoretical framework for current observational programs aimed at finding this missing matter using the Hubble Space Telescope.


Project Title: Planck LFI final data analysis
Project Leader: Hannu Kurki-Suonio, University of Helsinki, Finland
Resource Awarded: 20,000,000 standard core hours

Andres Curto, Institute of Physics of Cantabria, Santander, Spain
Elina Keihänen, University of Helsinki, Finland
Reijo Keskitalo, University of California, Berkeley, USA
Kimmo Kiiveri, University of Helsinki, Finland
Theodore Kisner, Lawrence Berkeley National Laboratory, USA
Ville Kosonen, University of Helsinki, Finland
Valtteri Lindholm, University of Helsinki, Finland
Luis Mendes, European Space Agency, SCIOPS, Madrid, Spain
Marina Migliaccio, University of Cambridge, UK
Paolo Natoli, University of Ferrara, Italy
Elina Palmgren, University of Helsinki, Finland
Martin Reinecke, Max Planck Institute for Astrophysics, Garching, Germany
Matti Savelainen, University of Helsinki, Finland
Anna-Stiina Suur-Uski, University of Helsinki, Finland
Daniele Tavagnacco, Trieste Astronomical Observatory, Italy
Maurizio Tomasi, IASF Milano, Italy
Jussi Väliviita, University of Helsinki, Finland

The aim of this project is to perform critical parts of the final analysis of the Planck satellite data. Planck is a space mission of the European Space Agency. It has produced the most important cosmological data set of the decade. Planck mapped the microwave sky at 9 frequencies, including polarization measurements at 7 frequencies. The 9 frequency maps are used to separate out the cosmic microwave background (CMB) from astrophysical sources. The resulting CMB map is analyzed to produce its angular power spectrum, from which cosmological parameters that describe the properties of our universe are determined. These include the age, overall curvature, and expansion rate of the universe, as well as the content of the universe: the amount of ordinary matter, radiation, neutrinos, dark matter, and dark energy.

The CMB reveals to us the structure of the early universe, showing us the primordial density perturbations, from which the galaxies later formed. From their properties we try to infer the nature of the mechanism that caused them. The favorite candidate is quantum fluctuations in the very early universe during a process called `inflation`. Many high-energy physics theories have been proposed to realize inflation. Planck will sort out, which, if any, of these theories agree with the observed properties of the universe.

Several rounds of Planck data analysis have already been performed, resulting in two data releases: the 2013 release based on the first 15 months of data and the 2015 release based on the full 4 years of data, but for which the analysis of the polarization data was still incomplete. PRACE resources were a significant contribution to both releases. The final round of data analysis is now beginning, leading to the final Planck results and the final data release by the end of 2016.

For accurate results, sophisticated data analysis is required. Simulation work dominates the computational load of Planck data analysis. Planck analysis codes either require simulations for self-calibration and validation of results, or depend critically on simulations for the results themselves. In many cases, codes require a Monte Carlo (MC) set (an ensemble of statistically consistent simulations that exhibit some well-specified and controlled degree of stochastic variation), usually in the form of frequency or component maps.

The aim of this application is to do three resource-intensive parts of Planck data analysis:

  1. Timeline-to-map MC simulation of Planck LFI data: thousands of realizations of instrument noise and CMB signal; also simulations astrophysical foreground radiation signal; from the maps angular power spectra, and their biases and error bars are estimated; also contributions to non-Gaussianity from systematic effects.
  2. Estimation of pixel-to-pixel noise covariance matrices of the LFI maps
  3. Cosmological parameter estimation for multi-field inflation cosmological models.

Bio Sciences (7)


Project Title: Computational study of the tetramer of amyloid-beta peptide
Project Leader: Prof. Mai Suan Li, Polish Academy of Sciences, Institute of Physics, Warsaw, Poland
Resource Awarded: 7,000,000 standard core hours

The project aims at providing molecular models to the community of scientists performing experiments with samples where amyloid peptides are in interaction with transition metal ions.
These peculiar interactions are important in connecting the molecular mechanisms underlying neurodegeneration to ions’ dishomeostasis involved in aging and in many pathologies. The tools and the methods employed in this project are of wide application as they can be useful in understanding how micro-environmental effects can impact the structure and the chemical reactivity of disordered macromolecules.
The specific example of the amyloid-beta peptide in the tetrameric form is considered.


Project Title: CArdiac MechanoELectrics
Project Leader: Dr. Gernot Plank, Medical University of Graz, Institute of Biophysics, Graz, Austria
Resource Awarded: 6,750,000 standard core hours

Dr. Edward Vigmond, University of Bordeaux, LIRYC, Pessac, France
Dr. Christoph Augustin, Medical University of Graz, Institute of Biophysics, Graz, Austria
Dr. Jason Bayer, University of Bordeaux, LIRYC, Pessac, France
Dr. Fernando Campos, Medical University of Graz, Institute of Biophysics, Graz, Austria
M.Sc. Caroline Costa, Medical University of Graz, Institute of Biophysics, Graz, Austria
Dr. Andrew Crozier, Medical University of Graz, Institute of Biophysics, Graz, Austria
Dr. Aurel Neic, Medical University of Graz, Institute of Biophysics, Graz, Austria
Dr. Ali Pashaei, University of Bordeaux, LIRYC, Pessac, France
Dr. Anton Prassl, Medical University of Graz, Institute of Biophysics, Graz, Austria
Dr. Caroline Roney, University of Bordeaux, LIRYC, Pessac, France
Advances in science continue to provide ever increasing amounts of data and detail. With respect to cardiac physiology, MRI and CT can provide sub 100 micron resolution for large hearts, providing uprecedented anatomical detail. Models of single cell function now incorporate many subsystems, including ion movement, tension development, energetics, and stochastic calcium regulation. These descriptions continue to grow in size as we unravel the working of the single cell. It is these molecular level events which give rise to organ level behaviour which we observe as large mechanical deformations which pump blood. The electrical and mechanical systems of the heart are intertwined and we must understand their interaction if we are to progress in the treatment of heart disease.
Cardiac modelling has progressed to the point where it can have direct clinical impact. The most complex human heart simulations use meshes with millions of degrees of freedom, have behaviour at each node described by up to 100 differential equations, and must solve linear and nonlinear systems at each time step. Computational requirements are formidable and have hindered full implementation of all the modelling detail possible. The human heart is the most relevant and useful model, yet its size limits the detail which can be put in. This project will greatly enhance our existing code to decrease runtime of high resolution cardiac electromechanical simulations while increasing the level of complexity and detail used in the model. We will achieve this through several avenues. Strong scalability of our code will be improved through the use of better preconditioners and optimal partitioning. Algorithms which show better stong scalability will be used.
We will use our simulator for three goals which are intertwined. The first is to study how electrical activation of the heart through the Purkinje system leads to optimal mechanical response. Working with electrocardiogram data, we will infer cardiac activation. Next, we will consider pathological cases where activation of the heart has been compromised and a pacing device needs to be implanted. We shall determine the ideal placement of leads for cardiac resynchronization therapy. Finally, we shall look at patients with aortic valve disease and aortic coarctation, conditions which both lead to pressure overload in the left vantricle. Using clinical datasets we will compute pressures which can be further exploited for determining blood flow and therapeutic targets.


Project Title: Cysteine-based redox switches in a mitochondrial chaperone
Project Leader: Elena Papaleo, Danish Cancer Society Research Center, Copenhagen, Denmark
Resource Awarded: 13,000,000 standard core hours

Richa Batra, Danish Cancer Society Research Center, Copenhagen, Denmark
Sanne Bergstrand-Poulsen, Danish Cancer Society Research Center, Copenhagen, Denmark
Miriam Di Marco, Danish Cancer Society Research Center, Copenhagen, Denmark
Giuseppe Filomeni, Danish Cancer Society Research Center, Copenhagen, Denmark
Matteo Lambrughi, Danish Cancer Society Research Center, Copenhagen, Denmark
Matteo Tiberti, University of London, UK


Project Title: The role of pigment packing in the flexibility of LHCII among different complexes: Insight from Molecular Dynamics Simulations
Project Leader: Assistant Professor Evangelos Daskalakis, Cyprus University Of Technology, Department of Environmental Science and Technology, Limassol, Cyprus
Resource Awarded: 2,180,250 standard core hours

Dr. Nikos Ioannidis, University of Crete, Department of Biology, Heraklion, Greece
Mr. Sotirios Papadatos, Cyprus University Of Technology, Department of Environmental Science and Technology, Limassol, Cyprus
Photosystem II (PSII) is known as the “engine of life” on earth. Photosynthetic antenna complexes of higher plants have a dual role: A) They absorb photons, transferring excitation energy to the reaction centers for photochemical utilization and B) With increasing light intensity they rapidly (within seconds) switch into a more dissipating state (energy trap), which safely converts excess energy into heat. The latter is associated with low lumen pH and high ionic gradient. The elucidation of the conformational changes upon this switch are vital in understanding the main photoprotective mechanism in plants. The identification of key components of a central mechanism that regulates the balance between photosynthesis and photoprotection can provide a basis for biochemistry that increase our understanding regarding plant sensitivity to stress. On this line, we will run trajectories from Classical Molecular Dynamic (CMD) and MetaDynamics Simulations of different LHC complexes in their lipid bilayer embedded forms. The protein conformational changes will be probed along with the role of the pigments and especially the carotenoids in the LHC flexibility at low lumen pH and high ionic gradient.


Project Title: Optogenetics: a first-principles investigation of Channelrhodopsin
Project Leader: Dr. Zeila Zanolli, University of Liege, Physics, Liege, Belgium
Resource Awarded: 8,347,500 standard core hours

Prof. Dr. Gianluca Lattanzi, Unversita degli studi di Bari Aldo Moro, Physics, Bari, Italy
Dr. Bruce Milne, Universidade de Coimbra, Center for Computational Physics, Coimbra, Portugal
Dr. Micael Oliveira, University of Liege, Physics, Liege, Belgium
Channelrhodopsin proteins have recently received a great deal of attention as useful tools for the analysis of neural networks in tissues and living organisms, a novel field of research going by the name of optogenetics. These proteins are light-gated ion channels, belonging to the family of type I opsins, comprising light-driven ion pumps. In this project we propose to study the channelrhodopsin-2 protein by combining molecular dynamics simulations and time-dependent density functional theory. The main objective of this Project is to improve our understanding of this protein by determining the key environmental factors that determine its visible light absorption characteristics.


Project Title: Mechanistics studies of human oxygen sensing enzymes
Project Leader: Carmen Domene, University of Oxford, UK
Resource Awarded: 1,700,000 standard core hours

Dr. Emily Flashman, University of Oxford, UK
Prof. Chris Schofield, University of Oxford, UK
The human body is able to sense changes in atmospheric oxygen levels and adjust its metabolic activities to suit the local environment. This is why we are able to live at a variety of altitudes ranging from below sea level to up on mountains high. How can cells detect and respond to oxygen levels? This project will lay the groundwork in order to exploit the basic science to artificially alter the activity of oxygenases, and to provide knowledge that will be useful for the pharmaceutical industry in targeting them for diseases.


Project Title: Large scale non-adiabatic QM/MM simulations for serial femtosecond x-ray crystallography structure refinement
Project Leader: Gerrit Groenhof, University of Jyväskylä, Finland
Resource Awarded: 15,000,000 standard core hours

A wide variety of organisms have evolved mechanisms to detect and respond to visible light. The biological response is mediated by structural changes that follow photon absorption. The initial step in such cases is the photo-isomerization of a highly conjugated prosthetic group. Examples include the photo-isomerization of the para-coumaric acid chromophore in the bacterial photoreceptor photoactive yellow protein (PYP),1 or of retinal in various rhodopsins. The isomerization occurs on ultrafast timescales, and is influenced by the protein environment of the chromophore.

Recently, we have directly determined the structural changes associated with the earliest steps in the photo-isomerization of the PYP chromophore by means of time-resolved serial femtosecond crystallography, using the femtosecond hard X-ray pulses emitted by the Linac Coherent Light Source. Although high quality diffraction patterns were obtained with time delays of 100 fs – 3 ps after initiating the photo-isomerization in the crystals with an 140 fs optical laser pulse, structural refinement of the transient structures populated on these ultrafast timescales turned out to be challenging, because these structures are very far from equilibrium and highly strained. Restraints in standard structural libraries for x-ray structure refinement are derived from structures at equilibrium and are therefore not applicable. In order to provide restraints appropriate to this refinement, which is essential to interpret the data of our experiment, we propose to perform excited state quantum mechanics/molecular mechanics (QM/MM) molecular dynamics simulations of the photo-isomerization.


Earth Sciences (2)


Project Title: High Resolution EC-Earth Simulations
Project Leader: Dr. Paul Nolan, National University of Ireland, Galway, Galway, Ireland
Resource Awarded: 3,403,400 standard core hours

The scientific objectives of the proposed project are two-fold;

i. to assess the improvements of the EC-Earth global coupled climate model in the representation of important climate processes with high-resolution global model resolutions (~39km) and

ii. to contribute to the preparation and running of the EC-Earth CMIP6 experiments.

The impact of greenhouse gases on climate change can be simulated using Global Climate Models (GCMs). Since 1995, the Coupled Model Intercomparison Project (CMIP) has coordinated climate model experiments involving multiple international modeling teams. The CMIP project has led to a better understanding of past, present, and future climate, and CMIP model experiments have routinely been the basis for future climate change assessments made by the Intergovernmental Panel on Climate Change (IPCC).
The CMIP5 simulations have demonstrated the added value of enhanced resolution when compared to output from the CMIP3 project. For example, the simulations showed significant improvement in the simulation of aspects of the large scale circulation such as El Niño Southern Oscillation (ENSO), Tropical Instability Waves, the Gulf Stream and its influence on the atmosphere, the global water cycle, extra-tropical cyclones and storm tracks and Euro-Atlantic blocking. In addition, the increased resolution enables more realistic simulation of small scale phenomena with potentially severe impacts such as tropical cyclones, tropical-extratropical interactions and polar lows. The improved simulation of climate also results in better representation of extreme events such as heat waves, droughts and floods.
This improvement in skill is expected to continue with the higher resolution CMIP6 simulations. It is envisaged that the proposed EC-Earth CMIP6 simulations, of the current project, will provide sharper and more accurate projections of the future global climate and lead to a better understanding, not only of the physical climate system, but also of the climate impact on societies. The high-resolution data will allow for an assessment of the impact of resolution on the accuracy of climate modelling. In addition, it is expected that the EC-Earth simulation data will be used as a basis for more-focused climate impact studies such as regional downscaling (e.g., boundary conditions provided to CORDEX). The EC-Earth simulations will likely be included for assessment in the expected U.N. Intergovernmental Panel on Climate Change (IPCC) AR6 reports.


Project Title: Near term climate projections in high resolution
Project Leader: Mr. Ralf Döscher, Swedish Meteorological and Hydrological Institute (SMHI), Rossby Centre, Norrköping, Sweden
Resource Awarded: 18,000,000 standard core hours

Martin Evaldsson, Swedish Meteorological and Hydrological Institute (SMHI), Rossby Centre, Norrköping, Sweden
Ulf Hansson, Swedish Meteorological and Hydrological Institute (SMHI), Rossby Centre, Norrköping, Sweden
Dr. Torben Koenigk, Swedish Meteorological and Hydrological Institute (SMHI), Rossby Centre, Norrköping, Sweden
The project ”Near term climate projections in high resolution” (NTCPROJ) is targeting the next level of near term climate projections for assessment of regional scale climate change by performing high resolution historical and near-term future climate scenario simulations. Analysis will initially focus on the Arctic area and the Northern Hemisphere.

Information on future climate is assessed with the help of future climate projections, carried out by numerical codes and forced by possible future greenhouse gas emission scenarios. This project ”Near term climate projections in high resolution” (NTCPROJ) is targeting the next level of climate projections by performing historical simulations of the recent climate and by near-term climate projections covering the time period of the coming decades up to 2050 in high resolution.

Typical climate projections, as for example analysed for the UN climate panel (IPCC) reports, are carried out over a time range of about 100 years with coupled climate models (including atmosphere, ocean, sea ice, land) in moderate grid resolution (150-250 km in the atmosphere). To enhance climate information for upcoming decades, atmosphere-standalone simulations in distinctly higher resolution (down to 40 km) are carried out, which use surface forcing information from previous fully coupled simulations. The forcing information provided by the coupled model includes ocean surface temperature, sea ice cover and snow cover on land. While coarse resolution coupled processes are incorporated in this set-up, enhanced atmosphere detail will regionally improve the representation of topographic influence and extreme weather and climate conditions.

There are large uncertainties in near-term changes of the large-scale atmospheric circulation; these arise from internal climate variability, model differences, and external forcing. The project will assess whether part of this uncertainty arises from the representation of the atmospheric response to sea-ice and snow-cover changes, and whether it can be reduced through better process representation. We will achieve this by performing a set of ensemble atmospheric model projections covering the next 30 years (i.e., near-term) with existing and improved model configurations.

The ensembles will further include runs using a range of surface boundary conditions that characterize uncertainties in the projected changes originating from coupled models, which are reflecting influences of internal climate variability and external scenario forcing. Results will illustrate the effects of high resolution and map the role of different sources of uncertainty, both of which is essential to build up of European climate services of high quality for various societal and economical sectors.

The model, EC-Earth, is particularly well suited to run in high resolution on PRACE systems due to its base in the Integrated Forecasting System (IFS) of the European Centre for Medium-Range Weather Forecasts (ECMWF), which has been optimized for high resolution. The proposed experiments cannot be carried out exclusively on national resources.


Engineering (7)


Project Title: Effect of rotation and surface roughness on heat transport in turbulent flow
Project Leader: Roberto Verzicco, University of Rome Tor Vergata, Italy
Resource Awarded: 17,000,000 standard core hours

Prof.Dr. Detlef Lohse, University of Twente, Faculty of Science and Technology, Enschede, The Netherlands
MSc. Richard Stevens, University of Twente, Faculty of Science and Technology, Enschede, The Netherlands
Dr. Yantao Yang, University of Twente, Faculty of Science and Technology, Enschede, The Netherlands
MSc. Xiaojue Zhu, University of Twente, Faculty of Science and Technology, Enschede, The Netherlands
“Turbulence” is often used as synonym for ”complexity”. It is a multiscale phenomenon, showing chaotic behaviour in space and time. Weather, climate, and ocean flow are all prime examples of turbulence and complexity [1-4]. The advantage in turbulence is that the underlying dynamical equations are well known. The most important way to drive turbulent flow is thermally – think of the atmosphere or of ocean flow. The paradigmatic example for such buoyancy driven flow is the Rayleigh-Bénard (RB) system [5-9], i.e., a fluid-filled cell heated from below and cooled from above. The RB system is mathematically well defined by the Navier-Stokes equations with appropriate boundary conditions, and exact global balance relations between the driving and the dissipation can be derived [10-13]. The system is also experimentally accessible with high precision, thanks to the simple geometry and high symmetries. All these features make RB the perfect model system for the development of the latest experimental and numerical techniques [8]. The main control parameters of the system are the Rayleigh number (the dimensionless temperature difference ∆ between the bottom and the top plate) and the Prandtl number (the ratio between kinematic viscosity ν and thermal diffusivity κ). In many industrial applications and natural phenomena the flow physics and heat transport are influenced by the addition of rotation and can also strongly depend on the surface roughness [5-8].

In contrast to a decade-old paradigm, even highly turbulent flows, e.g. rotating RB convection, display sharp transitions between different turbulent states [20-33]. In rotating RB convection the Rossby number, which indicates the ratio between the buoyancy and Coriolis force, indicates the rotation rate. Recent experimental results [31] at high Rayleigh numbers revealed the existence of a sequence of sharp transitions with increasing rotation rate towards the geostrophic regime [16,19,32-38], which is relevant to understand e.g. weather patterns and sea currents. We plan to study these different turbulent rotating flow regimes by using high-resolution direction numerical simulations (DNS) that will allow one-to-one comparisons with experiments performed by our collaborators. In the simulations we have access to the full flow field, which will allow us to unravel the physics of the different rotating turbulence regimes.

Various experiments [39-43] have revealed that roughness increases heat transfer, and therefore in many industrial applications surface roughness is used to increase the heat transport, e.g. in heating/cooling devices. Some experiments presumably reached the “ultimate state” of RB convection, where the Nusselt number (dimensionless heat transport) scales as Nu~Ra1/2 [42]. However, it is also speculated that this observation may just be a crossover regime [8]. More work is needed to understand this issue better. We plan to perform an extensive set of simulations to study the effect of different surface roughness patterns on the heat transport in RB convection to find out whether roughness can trigger the transition to “ultimate RB convection” [8,44-46] at lower Rayleigh (driving of the system), and to find out what the optimal roughness pattern is to increase heat transport.


Project Title: Coherent structures in isotropic turbulence
Project Leader: Javier Jiménez, ETSI Aeronáuticos UPM, Spain
Resource Awarded: 1,560,000 standard core hours

Understanding the dynamics of turbulent cascades is essential for developing appropriate and accurate tools for the prediction and control of turbulent flows. The study of coherent structures has recently become a new tool in this research, particularly since well-resolved simulations at realistic Reynolds numbers have begun to be possible. Our group has pioneered these techniques and currently hosts around a 1PB of well-resolved turbulence data (both homogeneous and wall-bounded), which are routinely used by us and other groups, both remotely and through dedicated visits.

The most important recent development has been the possibility of storing time-resolved simulations, allowing us to keep track of the evolution of these structures and of other aspects of the flow, and to study their dynamics. However these novel techniques are not so much limited by computational power as by the amount of data to be stored and post-processed. In collaboration with Nvidia, we have lately ported some of our simulation codes to gpu clusters. They are very efficient, but a few thousand node-hours can easily create several Terabytes.

The first purpose of this proposal is to supplement our library of time-resolved isotropic turbulence with a higher resolution case (1024^3 instead of 512^3), allowing us to extend our analyses to a more realistic Reynolds number. This “reference” data set would be very useful by itself, but the second goal is to use it to explore ways to reduce the time required for turbulence direct simulations, especially with reference to future ‘Pascal-’ and ‘Sierra-class’ gpu clusters. These promise to be very fast, ‘unfortunately’ producing even more data. At some problem size it will become more practical to extract all the flow information on-the-fly during the simulation rather than to store the results, particularly if otherwise idle resources can be used for this purpose, and to recompute the case when a new question arises. We have already used this procedure in channel simulations on classical machines (Lozano-Duran et al, 2014).

The critical problem size can be increased using single precision arithmetic and coarser grids, which are avoided now mostly because their effect on the physics is unknown. The idea is to accompany the above reference case by a series of downgraded simulations using restart reference files as initial conditions. Turbulence is a chaotic phenomenon for which any two simulations diverge at a rate given by the highest Lyapunov exponent. We have computed these exponents for isotropic turbulence, and the proposed procedure is to determine which numerical simplifications can be implemented while keeping the rate of divergence within its natural Lyapunov value. The present simulation will give us the opportunity to do this on a problem size of the order expected to be useful for the next one or two generations of chips.

The result is expected to be not only useful for the present isotropic code but for other codes such as turbulent channels (which we have also ported to gpus).


Project Title: Fractal generated turbulence in industrial applications: direct numerical simulations
Project Leader: Dr. George Papadakis, Imperial College London, Department of Aeronautics, London, UK
Resource Awarded: 12,992,000 standard core hours

The central aim of this proposal is to use high fidelity direct numerical simulations (DNS) to improve fundamental understanding of fractal generated turbulent flows applied to 3 industrial cases from the aeronautical and process industries. Recent experimental evidence has demonstrated that fractal grids can improve process performance in all cases, but the fundamental mechanisms that explain this behaviour are still not clear. Improved understanding is expected to lead to further performance enhancements. The presence of curved boundaries plays an important role in the separation and mixing processes, thus it is important to use body fitted grids, as opposed to Cartesian grids used up to now. More specifically, we plan to perform high fidelity simulations to elucidate the mechanisms that affect lift and drag in airfoils with fractal trailing edges, to characterise power consumption and mixing characteristics in stirred vessels with fractal and standard impellers and to investigate heat transfer enhancement in flows around a cylinder immersed in the wake of fractal grids. The elucidation of the underlying mechanisms for each case will demonstrate how fundamental research can be applied to provide solutions to practical industrial flow problems. We expect that the results will stimulate new ideas on how fractal turbulence can be used in other applications. The ultimate aim is to generate enough knowledge as to be able to provide guidelines for the design of bespoke grids tailored to specific applications.


Project Title: Complex fluid simulations for future revolutionary vehicles, from aquatic to aerial: highly deformable geometries and high-speed compressible flows
Project Leader: Research Associate Asimina Kazakidi, Forth, Institute of Computer Science, Greece
Resource Awarded: 4,158,000 standard core hours

Prof John Ekaterinaris, Foundation for Research and Technology, Institute of Applied and Computational Mathematics, Heraklion, Greece
Dr Konstantinos Panourgias, University of Patras, Department of Mechanical and Aerospace Engineering, Patra, Greece
Dr Dimitris Tsakiris, Foundation for Research and Technology, Institute of Applied and Computational Mathematics, Heraklion, Greece
This project seeks to utilize two disparate CFD codes (CURVIB and hpDG), for the solution of highly deformable geometries, and chemically reacting and plasma compressible flows, respectively:

(a) The flow dynamics around highly deformable geometries involving intense motions, which apply to cases such as cephalopod appendages, is of high importance for understanding the energetics of swimming and exploit this to robotic underwater vehicles. The difficulty in flow simulations around time-varying geometries lies in the solution accuracy. The use of fixed-grid methods has been the golden standard for such simulations, in which a moving (immersed) boundary is defined on a stationary domain; thus, these methods are capable of handling arbitrarily large deformations and allow effective solutions of fluid-structure interaction problems.

(b) The high-order accurate discontinuous Galerkin (DG) method for unstructured meshes is used to numerically solve high-speed flows with discontinuities, chemically reacting flows, and plasma flows over complex three-dimensional geometries. Automatic adaptive refinement/derefinement capabilities, both h-type (local mesh refinement) and p-type (order of polynomial expansion), have been implement. It was demonstrated that exploitation of these h/p refinement can yield high resolution numerical solutions required for large eddy simulations. High order explicit and implicit time stepping methods have been implement. All these features have been efficiently parallelized. As a result, the DG method we implemented provides high-resolution solutions, which can lead to better understanding of the flow fields by performing large eddy simulations over the complex geometries of interest. The DG method is well suited for large eddy simulations (LES) over complex configurations since filters can be used to ensure grid independence of the LES realizations and by exploiting the h/p refinement capabilities the backscattered energy of unresolved scales can be minimized. Applications of major interest are the flow control of plasma flows using electromagnetic fields, and high-speed combustion and re-entry flows. To render the DG method suitable for large scale problems, efficient numerical techniques have been implemented in the framework of a highly scalable parallel environment.


Project Title: Multiscale simulations of nanoparticle suspensions
Project Leader: Prof Pietro Asinari, Politecnico di Torino, Italy
Resource Awarded: 3,400,000 standard core hours

Nanofluids are fine dispersions of nanoscale particles in water or oil. Because of their excellent thermophysical properties, nanofluids are currently investigated in several research fields, especially for energy or biomedical applications. For example, nanofluids can be employed as more performing direct solar absorbers in solar collectors, or as coolants in more efficient heat exchangers for automotive applications. Moreover, suspensions of magnetic and photothermal nanoparticles offer less invasive alternatives to the traditional cancer imaging and treatment.

However, the multiscale nature of nanosuspensions represents the main issue in defining a proper fluid model which takes into consideration both nanoscale phenomena and resulting macroscopic properties. The dynamics of the solvent layer around nanoparticles, the Brownian motion as well as the particle clustering and stabilization are just some examples of the complexities involved in these physical systems. Despite a wide range of experimental observations and empirical theories, the nanofluids modelling still presents some criticalities. Hence, a comprehensive, multiscale physical understanding for designing and tailoring nanofluids to specific applications is still missing.

A possible bridge between classical Molecular Dynamics (MD), which deals with nanoscale phenomena, and continuum fluid models is the Coarse Grained (CG) modelling approach. In this project, we aim to define and develop a CG model able to reproduce particle dynamics in nanofluids, in order to predict their overall macroscopic properties. The basic idea behind coarse graining is to combine several atoms into one group (CG bead). Then, by mimicking atomistic interactions, bonded and nonbonded potentials between CG beads are introduced, therefore allowing to simulate the beads dynamics.

In this project, particle-particle potentials will be first computed by means of the “constraints” MD algorithm, which consists in progressively moving two nanoparticles along an interacting direction in order to evaluate the reciprocal potential of mean forces. The role of polymeric surfactants will be also investigated, and the particle-polymer interactions included in the description of the force field for the considered nanofluids. Finally, the dynamics of the CG nanofluid will be simulated, thanks to the beads’ geometry and force field obtained from the preparatory MD simulations. Both Brownian dynamics and fluctuating hydrodynamics will be included in the model, for a better representation of the overall thermophysical properties of nanofluids.

We expect that the CG method implemented in this project for nanofluids would also have the potential to describe other multiscale physical, chemical, biological and engineering systems.


Project Title: Large scale simulation of subcritical transition in pipe flow
Project Leader: Dr. Philipp Schlatter, KTH, Department of Mechanics, Stockholm, Sweden
Resource Awarded: 20,000,000 standard core hours

M.Sc. Jacopo Canton, KTH, Department of Mechanics, Stockholm, Sweden
Bent pipes are ubiquitous in both man-made devices and biological systems. Their application in industry ranges from junctions between sections of straight pipes, to exhaust tubes, and heat exchangers just to name a few examples. The reasons for their vast employment are, amongst others, their enhanced cross-sectional mixing and improved mass and heat transfer coefficients which make them an attractive alternative to traditional mechanical systems. Organic beings contain curved `pipes` which constitute, for instance, respiratory and vascular systems. In these cases the interest lies in understanding the influence that the vortical structures, formed due to the secondary flow, have on the circulation of fluid. In addition, bent pipes constitute a relevant flow case for the research on hydrodynamic stability and transition to turbulence. In fact they exhibit marginal differences with respect to the most canonical and studied wall-bounded flow case of the past century, i.e. the straight pipe, but very little is known about the mechanisms responsible for the onset of turbulence in curved pipes.

Accordingly, the understanding of the physics of this flow has a direct and substantial impact on everyday life and an adequate knowledge of such a problem will help in finding scientific methods to reduce drag and the like. The Navier-Stokes equations govern the dynamics of turbulent flows. This set of equations, when properly non-dimensionalised, includes the Reynolds numbers (Re) and the curvature as unique characterising parameters for bent pipes. Re is by far the most important nondimensional number in fluid mechanics, and can be considered as a measure for the “speed” of the flow inside the pipe. Turbulence is a characteristic state of flows with sufficiently high speeds, or high Reynolds numbers. Most, if not all, fluids observed in nature are indeed turbulent, or become so when Re is high enough. Of particular importance to the scientific and engineering communities is the understanding of the modifications, and the causes behind them, occurring in the flow as turbulence sets in.

The aim is to study the route to turbulence in low-curvature, bent pipe flow as a function of the Reynolds number through direct numerical simulations (DNS). DNS allows us to resolve all relevant scales of the turbulent flow. These will be carried out using the massively parallel DNS code available at KTH Mechanics, nek5000, which is based on an accurate and efficient spectral-element discretization.


Project Title: Optimizing sails by supercomputing
Project Leader: Mikko Brummer, WB-Sails Ltd, Helsinki, Finland
Resource Awarded: 601,024 standard core hours

This application is directly related to the WB-Sails Ltd SHAPE project `CFD simulations of sails and sailboat performance` awarded by PRACE in April 2015.

We wish to implement in the project a HPC-based CFD-analysis of sails and sailboats, with the XFlow dynamic simulation code. The code allows 6-DOF (degrees of freedom) motion of the objects, for a realistic simulation of forces and moments around a boat moving in a seaway. Coupled with the Friendship Framework’s code CAESES, XFlow can be used for optimizing sail shapes for a number of parameters. XFlow’s ability to predict free surface flows is explored against known model test cases by the Delft model Basin.

The HPC provider allocated in the SHAPE project was CSC in Finland. However, it has turned out that CSC cannot provide free HPC access to a private company. Therefore we are looking for another provider through DECI who can.


Informatics (2)


Project Title: Accelerating Spike Exchanges in Parallel NEURON Simulations
Project Leader: Prof. Zeki Bozkus, Bogazici University, Istanbul, Turkey
Resource Awarded: 892,500 standard core hours

NEURON ( is a popular simulation environment for building and using computational models of neurons and networks of neurons. One of the future aims of computational neuroscience is to simulate a whole human brain consisting of billions of neurons.
A parallel version of NEURON package is available. Each neuron is represented by a separate process and spikes are exchanged between processes using communication facilities provided by MPI. NEURON simulator uses MPI Allgather and MPI Allgatherv calls for spike exchanges.
In this project, we aim to optimize communication in NEURON simulations. One of the things we try is to use METIS for graph partitioning on NEURON models to optimize the communication based on communication topology. Graphs to be used as input for METIS are generated from the communication matrix extracted from the simulation. After the partitioning, a mapping from virtual processes
to real hardware is made. Our method maps frequently communicating processes to the same machine if possible in order to decrease the overhead of communication. Another optimization method involves the use of persistent communication facilities of MPI.


Project Title: A middleware for data staging control and coordination for the HPC software storage I/O stack.
Project Leader: Florin Isaila, Universidad Carlos III de Madrid, Madrid, Spain
Resource Awarded: 800,000 standard core hours

On current large-scale HPC platforms the data path from compute nodes to final storage passes through several networks interconnecting a distributed hierarchy of nodes serving as compute nodes, I/O nodes, and file system servers. Although applications compete for resources at various system levels, the current system software offers no mechanisms for globally coordinating the data flow for attaining optimal resource usage and for reacting to failure or overload.

In this project, we develop CLARISSE, a middleware designed to enhance data-staging coordination and control in the HPC software storage I/O stack. CLARISSE exposes the parallel data flows to a higher-level hierarchy of controllers, thereby opening up the possibility of developing novel cross-layer optimizations, based on the run-time information. To the best of our knowledge, CLARISSE is the first middleware that decouples the policy, control, and data layers of the software I/O stack in order to simplify the task of globally coordinating the data staging on large-scale HPC platforms.

There are four main topics for which we would need PRACE resources for empirically proving the scalability of our approach: elastic load-aware collective I/O, cross-application parallel I/O scheduling, data path resilience, in-situ and in-transit locality exploitation.


Materials Science (26)


Project Title: Ab initio Design of Perovskite Photovoltaics
Project Leader: Prof. Feliciano Giustino, University of Oxford, Department of Materials, Oxford, UK
Resource Awarded: 14,000,000 standard core hours

Dr. Fabio Caruso, University of Oxford, Department of Materials, Oxford, UK
Dr. George Volonakis, University of Oxford, Department of Materials, Oxford, UK
Owing to the increasing global demand for energy and the environmental impact of fossil fuels, the development of efficient strategies for solar energy conversion has become a key scientific priority. Among realistic alternatives to conventional silicon-based photovoltaics, solar cells based on hybrid organic-inorganic metal-halide perovskites have recently attracted enormous attention due to their record-high energy-conversion efficiencies (up to 20.1%), low materials costs, and highly scalable manufacturing processes. Despite the extraordinary performance of such devices, there are still many fundamental open questions, which need to be addressed in order to further optimize these devices and improve the energy-conversion efficiency. For example one of the key priorities is to find suitable replacements for the environmental issue created by Pb in the lead tri-iodide perovskite network. Within the ‘Ab initio Design of Perovskite Photovoltaics’ project, we will apply first-principles numerical simulations — i.e., atomic-scale simulation methods based on the fundamental equations of quantum mechanics – to the description of the key microscopic processes that underpin the solar-energy conversion in hybrid perovskites. We will proceed to a comprehensive computational screening of novel perovskites for photovoltaic applications. This search will predict — among a large library of perovskites — the materials with the optimal properties for the realization of novel record-breaking perovskite solar cells. Furthermore, the opto-electronic properties of the most-prominent candidates will be assessed using state-of-the-art first-principles methods. Finally, we will address the interfacial loss mechanisms that limit the energy-conversion efficiency in perovskite solar cells providing an unprecedented first-principles description of the electronic structure at the hetero-junction between the light absorbing layer (perovskite) and charge extraction layers.


Project Title: Simulating Bismuth-containing compounds for oxide ion conductivity
Project Leader: Prof. Graeme W. Watson, Trinity College Dublin, Dublin, Ireland
Resource Awarded: 8,160,000 standard core hours

In recent times, environmental concerns have driven research into more efficient, less polluting, and renewable energy sources. Solid oxide fuel cells (SOFCs) represent one such potential technology. However, current fuel cells have limited practical use due to their high operating temperatures, and thus limited lifespan and expense of producing, running, and maintaining them. The component which currently limits the efficiency is the solid electrolyte which allows the flow of oxygen across the SOFC. This requires temperatures which are prohibitively high for practical use to allow sufficient oxide ion conduction. Much research has been focused on developing electrolyte materials which can conduct oxide ions across the SOFC efficiently at lower temperatures, which is required for the cells to generate electricity. Two such materials containing Bismuth as a component are the subject of a computational investigation in this project. We are examining two recently synthesised materials that have shown exceptionally high conductivity by modelling the dynamics of the oxide ions under different environments. The first is based on layering the best-known oxide ion conductor, δ-Bi2O3, between another oxide ion conductor to stabilize the structure, and enhance the oxide ion conductivity, of δ-Bi2O3. The second material, a perovskite-structured system containing Bismuth, has recently been shown to conduct oxide ions remarkably well at temperatures lower the current operational temperatures of SOFCs. We will examine the dynamical behaviour of the oxide ions within these two systems, under different conditions which mimic operational conditions of a SOFC. This will be accomplished by modelling the behaviour of the atoms in the material using state of the art quantum mechanical methods run on high-performance computers. The aim is to obtain a detailed understanding of the behaviour of the oxide ions within these materials under simulated operating conditions to aid in the development of new, robust, and efficient materials for more environmentally friendly fuel cell devices.


Project Title: CHARge TransfER dynamics by time dependEnt Density functional theory
Project Leader: Dr. Biplab Sanyal, Uppsala University, Department of Physics and Astronomy, Uppsala, Sweden
Resource Awarded: 23,604,000 standard core hours

Dr Carmine Autieri, Uppsala University, Sweden
Raghuveer Chimata, Uppsala University, Sweden
Soumyajyoti Haldar, Uppsala University, Sweden
Dr. Banerjee Rudra, Uppsala University, Sweden
In recent times, there has been a tremendous interest in the research community in developing suitable routes for realizing alternative energy sources for the need of human mankind in near future.
The vast abundance of sunlight gives us the opportunity to convert solar energy to electricity and chemical energy through hydrogen production by water splitting, photocatalysis and photosynthesis. In all these areas, ultrafast charge transfer process plays an important role. Though the experimental field has progressed quite significantly in the last decade in studying charge transfer dynamics by femtosecond pump probe techniques coupled with core-hole clock method in the realm of x-ray spectroscopy, femtosecond transient absorption spectroscopy, time dependent fluorescence spectroscopy etc., the theoretical understanding of the ultrafast charge transfer processes via quantum mechanics is still inadequate. The complexity lies in the time dependent description of coupled electron and ion dynamics that occurs non-adiabatically. The adiabatic charge transfer process is not suitable to describe photoactive systems where a transition between different electronic levels occurs due to photon absorption. In this proposal, we aim to study ultrafast charge transfer processes by time dependent density functional theory and non-adiabatic molecular dynamics simulations and apply in a variety of problems. Ab initio molecular dynamics simulations will be performed by VASP code followed by time dependent density functional theory calculations using OCTOPUS code. Finally, non-adiabatic molecular dynamics simulations will be carried out using PYXAID code.


Project Title: Fundamental optoelectronic properties of ZnO-X alloys
Project Leader: Prof. Clas Persson, University of Oslo, Department of Physics (FYS), Oslo, Norway
Resource Awarded: 13,351,040 standard core hours

M.Sc. Gustavo Baldissera, KTH, Sweden
Dr. Kristian Berland, University of Oslo, Department of Physics (FYS), Oslo, Norway
M.Sc. Rongzhen Chen, KTH, Sweden
M.Sc. Alexander Hupfer, University of Oslo, Department of Physics (FYS), Oslo, Norway
Dr. Oleksandr Malyi, University of Oslo, Department of Physics (FYS), Oslo, Norway
ZnO is a very good example of a functional material suitable for many different types of applications. Bulk ZnO and cation alloying of ZnO are today intensively investigated. Surprisingly however, very little attention has been paid to understanding more advanced alloy structures based on ZnO. In the project we will further explore the rather unconventional type of ZnO-based materials, that is (ZnO)1-yXy where X is an isovalent alloy compound, for instance X = SiC. By substituting both the cations and anions in ZnO (e.g. SiZn and CO) one can significantly alter and control the material properties, while ensuring relatively small disturbance on the crystalline structure since the binary constituents are isovalent with matching bond lengths.


Project Title: FRICtion SIMulations in transition metal dichalcogenides systems
Project Leader: Dr. Paolo Nicolini, Czech Technical University in Prague, Department of Control Engineering, Prague, Czech Republic
Resource Awarded: 7,700,000 standard core hours

Dr. Antonio Cammarata, Czech Technical University in Prague, Department of Control Engineering, Prague, Czech Republic
Dr. Benjamin Irving, Czech Technical University in Prague, Department of Control Engineering, Prague, Czech Republic
Prof. Dr. Tomas Polcar, Czech Technical University in Prague, Department of Control Engineering, Prague, Czech Republic
Friction is a hundreds-of-billions-of-dollars problem. It is responsible for a lot of inconveniences, such as wear, wastefulness of mechanical energy and production of heat and noise, in a way that in many cases is unwanted. Since ancient times, humans have made use of lubricants for reducing friction. Most lubricants are liquid, but there are several applications where liquid lubricants are not suitable or may even be impossible to use, e.g. where they could be easily expelled from gaps between moving parts of a device, and/or in high-temperature or vacuum conditions. In such cases, one is obliged to resort to dry lubricants. Substances that belong to this class of lubricant include graphite, polytetrafluoroethylene and transition metal dichalcogenides (TMDs). The latter group of compounds is particularly interesting due to their excellent frictional behavior in dry air or vacuum conditions (molybdenum disulfide, MoS2, is one of the first materials showing superlubricity in ultra-high vacuum). However, low hardness, high porosity and low adhesion to the substrate are major drawbacks to tackle. In addition, although mechanisms explaining superlubricity were proposed as long ago as the early 90s, a comprehensive understanding of the phenomena occurring during the sliding process under load is far from certain.

Experimental discovery of new and useful TMD-based materials with optimal characteristics is challenging and time-consuming, requiring development cycles that include candidate material identification, testing, and further structural optimization. Computational techniques expedite the process by narrowing the composition-structure phase space for experimental exploration to only the most favorable compounds, efficiently driving the design of novel TMD materials with improved or new properties.

The main goal of the project is to get a deep understanding of the phenomena that take place during the sliding process of TMD-based materials under load (i.e., tribological conditions) with particular attention to the effect of different environmental conditions on the frictional properties of such materials. In particular, we will study how the interlayer interactions affect the local atomic topology (atom rearrangement and/or plane reorientation) during tribological motion of TMDs. Ab initio and molecular dynamics (MD) simulations, in combination with group theory techniques, will allow us to have a better understanding of the origin of macroscopic friction at the nanoscale. Expected outcomes of the study include also improvements in parameterization of existing empirical force fields.


Project Title: Ferroelectric Relativistic Intrinsic Solar Cells
Project Leader: Dr. Aron Walsh, University of Bath, Chemistry, Bath, UK
Resource Awarded: 23,298,048 standard core hours

Ferroelectric photovoltaics (photoferroelectrics), are currently attracting significant attention, due to their unique combination of a spontaneous photocurrent and a switchable photovoltage [1-3]. Materials engineering has enabled materials which had previously been considered poor absorber materials, such as oxide perovskites, to be applied to solar-energy conversion. The halide perovskites, in particular CH3NH3PbI3 (MAPI), have provoked a revolution in photovoltaics [2]. The ability to induce symmetry breaking, removing crystal inversion symmetry (e.g. with temperature, pressure, or defects) and allowing phase transitions from centro-symmetric to polar space groups, opens new possibilities for the application of ferroelectric materials as PV devices.

Moreover, manipulation of the spin-degrees of freedom, moving electrons by means of electric fields, has also reinvigorated the interest of the scientific community in the development of inversion-asymmetric structures [4]. This can be achieved through the combined effect of spin-orbit coupling and potential symmetry-breaking, resulting in a momentum-dependent splitting of spin bands – the so-called Rashba effect [5,6].

This project will target two key topics focusing on SbXY (X = chalcogen, Y = halide) materials, which are prototype photoferroelectrics, and will allow for the opportunity to probe spin-electric coupling mechanisms, derived between the spin-orbit Rashba effect and ferroelectricity.

(i) Thermal properties: Materials modelling has traditionally focused on characterising equilibrium structures; however, systems perturbed by temperature, pressure and light can undergo significant changes in structure and properties, which need to be accounted for to better model behaviour under operating-device conditions. The temperature dependence of a number of material properties can be probed through lattice-dynamics calculations, as demonstrated for the lead chalcogenides [7] and perovskite systems [8,9]. Using similar calculations, we will model the phase equilibria and obtain, among other things, the structure and optical properties as a function of temperature on the Density-Functional Theory (DFT) free-energy surface, by employing the Quasi-Harmonic Approximation (QHA), for a series of SbXY materials.

(ii) Ferroelectric properties: Ferroelectrics can be characterized as having a nonpolar reference structure which is related to the ferroelectric ground state through a polar distortion. The transition path corresponds to one or more unstable phonon modes, which can be identified through lattice-dynamic calculations, and the path can thus be characterised by following one (or more) negative modes at the Brillouin-zone center in a suitable supercell. By taking the non-polar configuration as the reference structure, the atoms are displaced along the collective eigenvectors of the negative modes, allowing the evolution of the structure and properties through the transition to the ferroelectric phase to be probed. This technique sometimes also allows new polar phases to be identified.

[1] V. M. Fridkin, Photoferroelectrics, (1979). [2] K. T. Butler, et al Energy Environ. Sci. 8, 838 (2015). [3] J. Im, et al J.Phys.Chem.Lett. 6, 3503 (2015). [4] A. Manchon, et al, Nature 14, 871 (2015). [5] D. Di Sante et al, Advanced Materials 25, 509 (2013). [6] S. Picozzi, Frontiers in Physics 2, 1 (2014). [7] J. M. Skelton, et. al, Phys. Rev. B, 89, 205203 (2014).


Project Title: Engineering the electronic properties of epitaxial graphene on SiC(0001) via intercalation and molecular transfer doping.
Project Leader: Dr. Nuala Caffrey, Linköping University, Department of Physics, Chemistry and Biology (IFM), Linköping, Sweden
Resource Awarded: 7,000,000 standard core hours

Prof. Igor Abrikosov, Linköping University, Department of Physics, Chemistry and Biology (IFM), Linköping, Sweden
Dr. Rickard Armiento, Linköping University, Department of Physics, Chemistry and Biology (IFM), Linköping, Sweden
The thermal decomposition of silicon carbide (SiC) is one of the most promising methods to produce high-quality epitaxial graphene on a wafer scale, directly on a semiconducting surface. However, the electronic properties of the resultant graphene has been shown to depend intimately on the chosen SiC surface. When graphene is grown on the Si-rich SiC(0001) surface, the first carbon layer is covalently bonded to the surface Si atoms, with only subsequent layers displaying the characteristic electronic features of graphene. Furthermore, these graphene layers are heavily doped, due to charge transfer from the surface, and have a considerably reduced electron mobility compared to free-standing graphene.

Our aim is to compensate and eliminate the structural and electronic influence of the interface with the substrate. We will do this by intercalating an electronegative nitride layer at the interface or by functionalizing the surface with charge accepting organic molecules. We plan to use automatized largescale ab-initio computation, in line with the emerging trend of high-throughput computation for engineering new materials, to screen a vast number of molecules to find those capable of inducing charge neutrality in graphene/SiC(0001).


Project Title: Structural and dynamical properties of complex interfaces
Project Leader: Dr. Jorge Iniguez, Luxembourg Institute of Science and Technology, Materials Research and Technology, Belvaux, Luxembourg
Resource Awarded: 6,588,449 standard core hours

Prof. Matthieu Verstraete, University of Liege, Physics, Liege, Belgium
INTERPHON will study from first principles the vibrational properties of materials with interfaces. The latter represent a very important contribution to thermal transport in particular, and are still a challenge to characterize, requiring nanoscopic, quantum, and chemical detail on one hand, and large system sizes (mesoscopic, from 100s of nm to micrometers) on the other. INTERPHON is the first step in our endeavour, and will characterize two classes of complex interface systems, i.e., domain walls in ferroelastic perovskites and incommensurate interfaces in chalcogenide superlattices. Both axes have intrinsic novelty and scientific interest, and will further be a basis for fitting effective models that can then tackle realistic system sizes.


Project Title: Design principles for solid-state linkage-isomer systems: a computational survey of known materials
Project Leader: Prof. Steve Parker, University of Bath, Chemistry, Bath, UK
Resource Awarded: 15,750,000 standard core hours

Ms. Lora Da Silva, University of Bath, Chemistry, Bath, UK
Dr. Jonathan Skelton, University of Bath, Chemistry, Bath, UK
Dr. Aron Walsh, University of Bath, Chemistry, Bath, UK
Linkage isomerism is a change in the binding mode of a ligand in an organometallic complex in response to external stimuli such as heat or light. These microscopic structural changes can lead to significant changes in macroscopic properties such as colour and non-linear optics. When the switching is retained in the crystalline form, linkage isomerism is a canonical example of solid-state single-crystal-to-single-crystal phase transitions. Given suitable properties (e.g. a mechanical or photochromic response), solid-state linkage-isomer systems have potential applications as functional materials for e.g. sensing or data storage. Several families of linkage-isomer systems have been identified, based on e.g. NO, NO2, SO2 and N2 ligands, but at present the structure-property relationships which control the isomerisation and the associated changes in material properties are not well understood. Such an understanding is essential to enable targeted discovery of new systems, as well as the optimisation of properties for specific applications.

We have developed a workflow which combines the strengths of periodic and molecular quantum-chemical calculations for modelling linkage-isomer systems. The calculations provide a complete set of data on the isomerisation and resulting changes in properties, including how these are influenced by the crystal packing in the solid state. Previous proof-of-concept studies have demonstrated that the combination of techniques yields results in excellent agreement with state-of-the-art experimental work, and that the microscopic insight obtained from the simulations is highly valuable in interpreting experimental results and informing further studies.

In this project, we aim to apply our method systematically to a selection of know linkage-isomer systems from all four ligand families. We will build up a database of high-quality calculation results, which will be used to identify and understand trends in structure and properties. This insight will be used to develop design principles for identifying novel linkage-isomer systems, and for tuning the properties of existing ones for specific applications. Following the recent open-data initiative, we will aim to make the data freely available for other researchers in the field to benefit from. Finally, this study will also allow us to benchmark our methodology across a broader range of systems, and we expect the results of this to be applicable to other classes of functional molecular materials.


Project Title: Computational Modelling of Magnetic Nanostructures at Surfaces
Project Leader: Dr Zeljko Sljivancanin, Vinca Institute of Nuclear Sciences, Department of Theoretical Physics, Belgrade, Serbia
Resource Awarded: 10,500,000 standard core hours

Jelena Pajovic, Vinca Institute of Nuclear Sciences, Department of Theoretical Physics, Belgrade, Serbia
Dr Zoran Popovic, Vinca Institute of Nuclear Sciences, Department of Theoretical Physics, Belgrade, Serbia
Srdjan Stavric, Vinca Institute of Nuclear Sciences, Department of Theoretical Physics, Belgrade, Serbia
Solid state disks (SSD) and conventional hard disks drives (HDD) are the most common types of data storage devices used in modern computers. The HDDs still dominate the market of storage devices due to the highest capacity and the lowest price per byte of stored data. The HDD technology is based on tiny magnetic particles on the surface of the disk used to store information in digital form. Two physical states, utilized to represent zeros and ones, correspond to two different magnetic polarities of the particles. During last twenty years the areal density of magnetic data storage increased more than 1000 times reaching the value of 10 Tbits/cm2. The increase in the areal density is accompanied with continuous miniaturization of the magnetic particles, which drives data storage technology to the nanometer size. Yet, the scaling of the magnetic grains is limited by thermal instability since for very small particles the magnetic anisotropy energy (MAE), responsible for preserving the direction of their magnetization, becomes comparable with the thermal energy. The random flip of the magnetization in small particles under the influence of temperature is commonly referred as superparamagnetism. According to Arrhenius law, for MAE of 40 meV per atom and a typical stability requirement of ~10 years, the minimal number of atoms within magnetically stable nanoparticles is approximately 30000. It turns out that the MAE per atom sharply increases when the dimensionality of the magnetic structures is reduced. Hence, the engineering of low-dimensional magnetic structures at well defined crystalline surfaces emerges as a promising approach towards design of novel data storage media which would provide the storage capacity far superior compared to those presently available at the market. Due to an enormous growth in computing power and advances in numerical methodologies during the last decade, the computational modelling of materials has matured to the stage which allows description of the properties of studied structures with time and spatial resolution not always accessible even with the cutting-edge experimental methods. Computational material science reached the stage where novel materials with tailored properties are about to be designed by computer. The main scientific aim of this project is to combine world class expertise and experimental infrastructure of research group from EPFL, Lausanne, Switzerland, with high level competence in ab-initio modeling of materials based on density functional theory (DFT), acquired by the research team from Vinca Institute of Nuclear Sciences, Belgrade, Serbia, to investigate magnetism in metal nanostructures grown on 2D materials (graphene and hexagonal boron-nitride (h-BN) or ultrathin MgO films) supported by metal surfaces [Ag(100), Ru(0001), Pt(111), Ir(111)]. The synergy between experiment and theory will provide insights into physical properties of the studied structures, with the level of details which is beyond capabilities of any of applied methods alone. In addition to a close collaboration with experimental group from EPFL we will examine several relevant model systems where computational simulations will enable understanding of fundamental mechanisms giving rise to peculiar magnetic properties experimentally observed in nanostructures at crystalline surfaces.


Project Title: Magneto Electric Governance for Applications and Polarization Amplifications versus Structures and Temperature in ABO3
Project Leader: Prof. Philippe Ghosez, University of Liege, Physics, Liege, Belgium
Resource Awarded: 9,625,000 standard core hours

Dr. Eric Bousquet, University of Liege, Physics, Liege, Belgium
Nickelates (ANiO3 where A is a rare earth) are materials which are famous for their Metal-Insulator phase transition. Indeed, the temperature at which the transition appears depends on the size of the Acation (until nearly 600 Kelvin for small A cation). This phase transition is linked to a structural distortionfrom the Pbmn to the P21n phase which corresponds to the appearance of an oxygen octahedral cages breathing mode in a rock-salt configuration. Due to this distortion, a gap opens which leads to the Metal- Insulator transition.

The main goal of our study is to control the temperature of this phase transition tweaking the modes. We decided to focus on YNiO3, PrNiO3, NdNiO3, GdNiO3 and LaNiO3. However, LaNiO3 is the only one that is always metallic and belongs to another phase (R-3c). Combining these compounds in heterostructures, we induce a polarization which is directly and linearly coupled to the modes responsible for the gap. Our first calculations on YNiO3/LaNiO3 show the possibility of controlling these modes thanks to an electric field. Moreover, coupling between the magnetic modes and the structural ones suggest the possibility of changing the magnetic properties acting on the structural ones.

The same kind of studies applies for ferrites (AFeO3) in addition with magneto-electric properties. As a consequence, electrical properties driven by magnetic constraints or at the contrary magnetic properties driven an electric field are expected. As for nickelates, the electronic structure of ferrites may be tunable controlling the lattice dynamic.

Finally, starting with calculations at zero temperature by means of first principles calculations, the goal is to validate our conclusions against the temperature. The approach is based on effective potential fitted on our first principles simulations to combined the lightness of an empirical potential (compared to first-principle) and the precision of first-principles calculations.


Project Title: Methanol Synthesis at the Terminal Interface of the Cr2O3/ZnO catalyst: Surface Islands or selective doping, and Fe2O3 as a new candidate material?
Project Leader: Dr. John Carey, Tyndall National Institute, Cork, Ireland
Resource Awarded: 5,099,218 standard core hours

The rapidly increasing world population and urbanisation over the past 100 years has led to huge developments and modernisation of industrial and automotive technologies. This demand has placed considerable strain on the non- renewable world oil reserves, and general consumption at the current rate is expected to exhaust supply of natural oil in the next 50 years. Current research efforts have been largely focused on developing materials for renewable energy applications in the areas of solar energy, geothermal energy, wind power, hydropower and biofuels as a result of the emerging energy crisis.

The research conducted under the European seventh framework funded BIOGO project (grant no: 604296) (Catalytic partial oxidation of biogas and reforming of pyrolysis oil for synthetic gas production and conversion into fuels will examine the conversion of biodegradable matter (biomass) to create biogas, which is mainly methane gas, to develop synthesis gas (syngas, CO + H2) as a precursor for developing methanol and alkanes as essential biofuels. Methanol is a useful product as methanol fuel cell technologies are already implemented in industry, while short and long chain alkanes can be useful to mix with current octane fuels (petrol) to ensure longevity, or produce desirable synthetic petrol. The ongoing challenge is the development of materials that are cost effective and efficient at carrying out the conversion process from syngas to biofuel. Through simulation lead catalysis design, BIOGO is investigating catalyst materials to carry out this process in a more cost efficient manner than traditional approaches. The traditional methanol synthesis catalyst uses a composition of ZnO/Cr2O3 for syngas production and Cu-ZnO/Al2O3 for methanol production as first proposed by BASF in 1923. The presence of Cr2O3 may result in formation of Cr(VI) in the catalyst, which is carcinogenic and banned within the EU; however extensive investigations on the ZnO/Cr2O3 catalysts for the syngas to methanol reaction has not been carried out to date.

This study will use density functional theory calculations to investigate the conversion of syngas to methanol on various mixed ZnO/Cr2O3, Cr2O3/ZnO composites by detailing the lowest energy key intermediates along the reaction pathway to provide a reaction energy profile. The energy profiles for the varying structures will be compared to elucidate the catalyst composition that is most effective for methanol synthesis. The structure of the mixed metal oxide catalyst which provides the lowest energy pathway will allow development of novel catalyst structures by substitution of the Cr2O3 component with Fe2O3. The methanol synthesis pathway will then be investigated on the ZnO/Fe2O3 catalyst for comparison to the traditional catalyst.


Project Title: Mixed Metal Oxides for Preferential CO Oxidation
Project Leader: Dr. Michael Nolan, Tyndall National Institute, Cork, Ireland
Resource Awarded: 5,099,694 standard core hours

Hydrogen-fueled proton exchange membrane fuel cell (PEMFC) technology is a leading, pollution-free and energy- saving power source for stationary and portable applications, that only produces water as a by-product. Efficient operation of the PEMFC requires high-purity hydrogen, where even ppm concentrations of CO severely poison the existing Pt anode catalyst. There is thus a pressing need for efficent H2 purification to obtain CO-free H2. In this regard, catalytic preferential CO oxidation (PROX), i.e., CO + H2 + 0.5 O2 -→ CO2 + H2, is a key reaction to remove traces of CO from the H2-rich stream and prevent catalyst poisoning.

The leading, and well studied catalysts for PROX are supported noble metal particle (Pt, Rh, Ru, Ir) catalysts. However, due to the high price (Which will only increase) and the limited availability of noble metals, which places this group of metal on the EU’s critical materials list, their long term application in high volume PEMFCs is uncertain. Therefore, recent years have seen increasing attention turning to noble-metal-free alternative catalysts. Leading materials in this regard are transition metal oxides, such as cobalt oxide and cobalt oxide-cerium oxide which show potential as catalysts for PROX.

However, the current understanding of PROX over these catalysts is rather limited and empirical, with no firm understanding of the fundamental properties of these materials for PROX. To improve our understanding of these catalysts and ensure their incorporation into PEMFC technology the fundamentals of their performance in PROX have to be understood, which will be the target of this PRACE project. We explore key questions, such as: what are the electronic and geometric structure of Co and Ce during PROX, what are the active sites and reaction pathways and what is the role of CeO2 in activating further Co3O4, together with experimental work with Prof. Rupprechter in Vienna. Boost our understanding of PROX over cobalt based oxide catalysts will enable rational design of the next generation of PROX catalysts.


Project Title: Multiscale Modelling of Ionic Conductors
Project Leader: Prof. Natalia Skorodumova, KTH, Sweden
Resource Awarded: 13,727,000 standard core hours

Johan Nilsson, KTH, Department of Materials Science and Engineering, Stockholm, Sweden
PhD assistant Prof. Igor Pasti, University of Belgrade, Faculty of Physical Chemistry, Belgrade, Serbia
Prof. Andrei Ruban, KTH, Department of Materials Science and Engineering, Stockholm, Sweden
Dr. Olga Vekilova, KTH, Department of Materials Science and Engineering, Stockholm, Sweden
Pjotrs Zguns, Uppsala University, Department of Physics and Astronomy, Uppsala, Sweden
Ion conducting materials are in the heart of many modern clean energy technologies, such as fuel cells, lithium batteries, catalyst etc. Better understanding of mechanisms operating in these materials starting from the atomic level is vital for future progress of these technologies. Oxide-ion conductivity in these materials shows complex behaviour depending on dopant concentration, preparation method and sample history. This makes it difficult to draw reliable conclusions about the mechanisms and decisive factors of optimal ionic conductivity based only on experimental data. Presently there is little information about the dopant–oxygen vacancy distribution in bulk oxides (apart from small dopant concentrations) and virtually nothing about their redistribution at various interfaces, and its impact on conductivity. The present project will substantially extend the current knowledge regarding the physics of ionic conductors by revealing the relation between dopant–oxygen vacancy ordering and conductivity, and the relevant ab initio based multiscale methodology beyond the current state-of-theart will be developed.


Project Title: Computational modelling of functionalization of 2D systems with nanomagnets
Project Leader: Prof. Dr. Ivan Stich, Slovak Academy of Sciences, Institute of Physics, Department of Complex Physical Systems, Bratislava, Slovak Republic
Resource Awarded: 15,000,000 standard core hours

Dr. Jan Brndiar, Slovak Academy of Sciences, Institute of Physics, Department of Complex Physical Systems, Bratislava, Slovak Republic
Dr. Rene Derian, Slovak Academy of Sciences, Institute of Physics, Department of Complex Physical Systems, Bratislava, Slovak Republic
Prof. Dr. Jaroslav Fabian, Universität Regensburg, Institut I – Theoretische Physik, Regensburg, Germany
MSc. Tobias Frank, Universität Regensburg, Institut I – Theoretische Physik, Regensburg, Germany
Dr. Kamil Tokar, Slovak Academy of Sciences, Institute of Physics, Department of Complex Physical Systems, Bratislava, Slovak Republic
Dr. Robert Turansky, Slovak Academy of Sciences, Institute of Physics, Department of Complex Physical Systems, Bratislava, Slovak Republic
Experimental realization of graphene by mechanical exfoliation has triggered unprecedented research activity. The discovery has paved the way to study of an entire family of two-dimensional crystals with interesting properties that were mostly unexplored so far. Nevertheless, pristine graphene has several drawbacks. First, it is a zero band-gap material, which hinders its potential applications for nanoelectronics. Second, magnetic properties are absent from graphene. Therefore, the proper functionalization of graphene, in particular charge and spin doping on atomic level, has recently received much attention. Despite the huge present effort, full control of charge and spin doping processes has not yet been achieved. The main difficulty comes from strong quantum correlations between electrons especially for nano-magnets. The most common approach to predict properties of magnetic nanostructures is the density functional theory (DFT) in various approximations, or via perturbation theories based on DFT. However, these methods have well-known limitations, especially for systems with open-shell electronic configurations, such as nanostructures composed of magnetic dand f- atoms. To overcome such problems use of Quantum Monte Carlo (QMC) techniques for describing strongly correlated electronic systems are proposed. Their prediction power is comparable to standard quantum chemistry techniques albeit at much lower computational cost and represents a completely different computational paradigm compared both to DFT and quantum chemistry methods. QMC does not attempt to model the electronic wavefunction directly, but rather only statistically samples the many-body wavefunction to collect the information on the underlying observables. The statistical nature of the method is the reason for its comparative numerical effectiveness, and makes calculations very tolerant to platform heterogeneities. Based on all above aspects we propose use of QMC techniques to calculate properties of magnetic nanostructures composed of the first raw transition metal atoms on two dimensional graphene and/or related 2D substrate.


Project Title: New routes to describe the electronic correlation in solid state systems
Project Leader: Dr. Dario Rocca, Université de Lorraine, Nancy, France
Resource Awarded: 4,050,000 standard core hours

The goal of this project is to develop and apply new powerful computational tools to compute the properties of materials from first principles by using quantum mechanics. The need for a breakthrough in this field is particularly urgent to accurately understand and predict the microscopic mechanisms and interactions involved in complex experimental and technological applications. For example, the weak dispersion forces necessary to understand the interactions in molecular crystals or in surface catalysis applications are poorly described by the commonly used ground state density functional methods. This proposal will apply new methodologies based on the concept of linear response function (i.e. dielectric matrix or polarizability) to treat with unprecedented accuracy important problems in materials science. The outcomes of this proposal based on large scale numerical calculations will have profound implications for science and society.


Project Title: Nitride-based quantum wells: From atoms to vertical transport
Project Leader: Dr. Stefan Schulz, Tyndall National Institute, Cork, Ireland
Resource Awarded: 2,040,000 standard core hours

Heterostructures based on the nitride semiconductor family InN, GaN, AlN and their related alloys form key building blocks of modern light emitting devices. Despite intensive research work worldwide, several open questions still remain, not only from a device perspective but also regarding the fundamental properties of nitride-based nanostructures. From a theoretical modeling viewpoint these structures present a variety of challenges. For instance, alloy fluctuations in nitride-based structures lead to strong carrier localization effects, which affect electronic, optical and transport properties of these systems as demonstrated clearly in experimental studies. However, in the theoretical modeling random alloy fluctuations and their consequences are widely neglected or have to be introduced artificially in continuum-based calculations.

NIQUFATRAN aims for a fully three-dimensional atomistic modeling of the electronic transport through multiple nitride- based InGaN quantum wells, accounting for alloy fluctuations and corresponding carrier localizations effects. Starting from atomistic calculations, alloy fluctuations and their impact on the transport properties in multiple quantum well systems are naturally included in our approach. To achieve this, NIQUFATRAN combines atomistic tight-binding models for the electronic structure calculations with non-equilibrum Green’s functions techniques to gain insight into the transport properties of heterostructures relevant for device applications. We target here both the analysis of electron and hole transport properties and how this dependents on the quantum well indium content as well as the barrier thickness between the quantum wells. In connection with industry partners, the theoretically obtained results will be compared and benchmarked against experimentally achieved data. Thus, the outcome of NIQUFATRAN is of significant interest to both answering questions related to basic properties of nitride-based systems as well as to device applications.


Project Title: Nonequilibrium molecular dynamics simulations for the development and parameterization of a continuum model for the rheological properties of polymer nanocomposite melts
Project Leader: Prof. Vlasis Mavrantzas, University of Patras, Department of Chemical Engineering, Rio, Patras, Greece
Resource Awarded: 5,960,906 standard core hours

Emannouil Theodoros Skountzos, University of Patras, Department of Chemical Engineering, Rio, Patras, Greece
Dimitris Tsalikis, University of Patras, Department of Chemical Engineering, Rio, Patras, Greece
Flora Tsourtou, University of Patras, Department of Chemical Engineering, Rio, Patras, Greece
We propose to carry out large-scale nonequilibrium molecular dynamics (NEMD) simulations of PEO-silica nanocomposite melts in full atomistic detail in order to understand the interplay between polymer microstructure and flow. To enable a deep understanding of the role of particle pair interaction potential, we will restrict ourselves to unentangled PEO melts where complications arising from inter-chain entanglements and long relaxation times are absent. We will study several polymer nanocomposite melts corresponding to: a) different volume fractions of silica NPs in the PEO matrix, b) different MWs of the host polymer chains, and c) different silica NP diameters. Our simulations will allow us to understand the role of polymer-surface interactions on particle stability and dispersion, as well as the effect of shearing or annealing on the state of particle adsorption and particle dispersion. They will further enable us to determine the conditions under which the viscosity of the nanocomposite diverges exhibiting a solid like response, to compute the effective hard-hard sphere diameter of the NPs as a function of PEO molar mass, to understand the respective role of thermodynamic and hydrodynamic stresses to viscosity enhancement, and to examine how the microstructure of polymer chains between particles resembles that of chains confined between plates. Finally, the outcome of the NEMD simulations for the viscometric functions and the overall chain conformation will be used to parameterize a recently introduced nonequilibrium thermodynamics continuum model [1] describing the flow behaviour of unentangled PNCs.


Project Title: Molecular wires based on tautomeric proton transfer
Project Leader: Prof. Liudmil Antonov, Bulgarian Academy of Sciences, Institute of Organic Chemistry, Sofia, Bulgaria
Resource Awarded: 1,200,000 standard core hours

Information technology has in the last few decades changed our daily life dramatically. The resulting exponential increase of the amount of processed information requires new concepts and new devices for data processing and storage. The development of molecular devices is an exciting and promising idea in this direction, which might establish the necessary ground for needed technological jump in the information technology of the Future. The concept is based on the use of single molecules as building “hardware” elements (wires, switches, logic gates, rectifiers, etc.) and their further suitable assembly into working devices by using chemical bonding and/or nonbonding interactions. The quest for inventing such molecular level “hardware” has increased dramatically over the last decade and the main emphasis is given to organic and hybrid systems, because the wide range of molecular propensities can be combined with the versatility of synthetic chemistry to alter and optimize molecular structure in the direction of desired properties.

In the current proposal we suggest a new idea for molecular wires, based on proton transfer over various distances. The overall concept is based on achieving controlled tautomeric proton transfer between two terminal units linked by a conjugated pi-electronic system. Using computational abilities of PRACE a molecular design of a number of possible systems will be performed by using quantum chemistry.


Project Title: Quantum Monte Carlo Description of van der Waals Crystals
Project Leader: Dr. Sam Azadi, University College London, London, UK
Resource Awarded: 22,093,655 standard core hours

Molecular van der Waals (vdW) crystals, including organic and inorganic, play vital role in understanding the physics and chemistry of the Earth and planets. They are also of intense technological interest. Low-Z molecular systems are among the most abundant in the solar system, as represented by planetary gases and ices. Their behaviour at high pressures is crucial for modelling the structure, dynamic and evolution of the large planets. Moreover, compression of molecular systems provides the opportunity of forming new materials, possibly with novel properties such as high-temperature superconductivity, disordered and amorphous materials. One of the simplest organic molecular solids is crystalline benzene with aromatic rings stacking. Because of its simplicity, high symmetric and rigid molecular structure, crystalline benzene has become the model structure for calculating the lattice-model vibrations in molecular crystals.

Unfortunately standard electronic structure methods, i.e. standard functionals within density functional theory (DFT), do not well-treat systems with significant dispersion forces. Molecular vdW crystals are some of the worse systems for standard DFT. The crystalline benzene phase diagram is a big challenge for theoretical and computational methods. To distinguish between crystal phases and competing low-energy polymorphs, the lattice energy calculations require very high accuracy. The energy difference between crystalline benzene and its low-energy polymorphs under pressure is less than few kJ/mol [1]. The most successful energy method based on density functional theory (DFT) formalism is reliable to only ~10 kJ/mol [2]. Recently, it has been shown that to tackle this problem using ab initio many-electron wave function methods is essential [3] .

In this project, we will compute properties of benzene crystals including dispersion interactions with Quantum Monte Carlo (QMC) methods [4]. We will test modern van der Waals density functionals against the accurate QMC results. We will tune DFT functionals to give better results for molecular vdW crystals. We will use the diffusion quantum Monte Carlo (DMC) [5] method to determine the ground-state enthalpies and hence the phase diagram in the pressure range up to 200 GPa.


Project Title: π -interactions in the Phenalenyl dimer and in its derivatives through quantum Monte Carlo and the Jastrow Antisymmetrized Geminal Power ansatz
Project Leader: Dr. Emanuele Coccia, Université Pierre et Marie Curie, Paris, France
Resource Awarded: 7,596,712 standard core hours

Dr. Matteo Barborini, INFM-CNR S3, Modena, Italy
Julien Toulouse, Université Pierre et Marie Curie, Paris, France
Quantum Monte Carlo (QMC) methods are becoming a powerful and reliable alternative to wavefunction methods (post Hartree-Fock and perturbative approaches) and density-functional theory (DFT) for quantum chemical calculations, thanks to their favorable scaling with the system size and to the extremely good suitability to high performance computing infrastructures, such as petascale architectures.

Variational Monte Carlo (VMC) represents the simplest method, which combines together Monte Carlo integration for computing the energy as the expectation value of the electronic Hamiltonian H and the variational principle for the ground state. The diffusion Monte Carlo (DMC) technique is one of most used Green’s function Monte Carlo approaches, allowing one to extract the exact ground-state energy within the fixed-node approximation. VMC and DMC scale as N^3-4 (N is the number of electrons), similar to the DFT scaling. The main drawback of any QMC approach, namely the very large prefactor in the scaling preventing the systematic use of QMC in quantum chemical calculations of medium- and large-size systems, has been dramatically alleviated by the introduction of supercomputers, such as Blue Gene clusters, on which massive parallel calculations are made possible. Thanks to the embarrassing parallelism of the algorithms, QMC calculations provide an excellent speedup, linear up to thousands of MPI tasks.

In this project we will apply the Jastrow Antisymmetrized Geminal Power (JAGP) ansatz to the study of radical and diradical systems, as the the phenalenyl radical and its dimer and the corresponding anionic and cationic species, by means of the QMC TurboRVB package.

The phenalenyl is a vey popular neutral pi-radical that has been recently used to construct organic conductors. It can form molecular crystals with electric, magnetic and optical properties.

The JAGP has been seen very suitable to the investigation of systems in which both the static and dynamic electronic correlation play an essential role. For this reason, first we will perform a VMC and DMC investigation on the phenalenyl radical, and the corresponding anion and cation to study the interplay between electronic and structural features. Second, we will apply our computational protocol to the calculation of the binding energy of the dimer of the three systems, focusing the attention on the specific properties of the JAGP wave function along the potential energy curve.


Project Title: Electron Transport through Spin Crossover Molecules: Adsorption, Transport and Gating Mechanisms from First Principles
Project Leader: Dr. Sergi Vela, Université de Strasbourg, Institut de mécanique des fluides et des solides de Strasbourg, Strasbourg, France
Resource Awarded: 2,251,959 standard core hours

Dr. Jordi Ribas-Arino, Université de Strasbourg, Institut de mécanique des fluides et des solides de Strasbourg, Strasbourg, France
Prof. Vincent Robert, Université de Strasbourg, Institut de mécanique des fluides et des solides de Strasbourg, Strasbourg, France
In the last decades, spintronics has emerged as an alternative to traditional silicon-based transistors in the ever-lasting quest for (further) miniaturization. It is expected that breakthrough advances towards this goal will be achieved by using the spin of individual magnetic molecules trapped between electrodes in a controlled way, in what is called molecular spintronics.

Among the different families of molecules that have been studied to that purpose, spin crossover (SCO) materials have recently attracted attention because of their unique properties. Along this project, we will inspect the main features governing the anchoring process, kinetic trapping scenarios and gating mechanisms of the molecular junction within the electrodes. This includes studying a variety of Fe(II)-based SCO molecules attached to metallic electrodes such as Gold, Silver or Copper.

The performance of different functional groups, designed to be used as anchoring tails, will be analyzed, as well as the adsorption energy of those groups in different coordination sites (hollow, top, bridge), and different surfaces.


Project Title: Synergetic Effects of H and He Interacting with Dislocations in Tungsten
Project Leader: Dr. Dmitry Terentyev, SCK-CEN, Institute of Nuclear Materials Science, Mol, Belgium
Resource Awarded: 637,500 standard core hours

Retention of plasma gas components such as hydrogen (H) isotopes and helium (He) is one of the limiting factors in selection of plasma facing materials for future thermonuclear fusion devices. Tungsten (W) is one of the promising candidates for such materials and was chosen for the divertor armor for International Thermonuclear Experimental Reactor (ITER) and as the first wall material for design of demonstrational fusion power plant – DEMO. During the runtime of the thermonuclear installation, such materials undergo the exposure to plasma components, such as isotopes of helium and hydrogen. Although the majority of these elements is released back from the material, a certain amount of them is stored in the subsurface region which inevitably weakens the strength of the considered components.

This project is dedicated to the analytical estimation of simultaneous accumulation of H/He components in tungsten, which is essentially important to understand the relevant physical mechanisms of their trapping in the material and for a further thoroughly parameterization of them numerically.

The latest experiments of retention of plasma components in tungsten at the temperature below 500 K have pointed to the significant amount of retained helium and hydrogen. The latter. unlike helium, does not agglomerate in the form of clusters in the bulk defect-free material. While the low temperature in the experiments excludes the influence of thermally activated vacancies, the lattice defects such as dislocations and grain boundaries are considered as a possible traps for hydrogen and helium.

Up to now the magnitude of interaction energy between dislocation and plasma components was estimated for screw dislocations and H/He isotopes only using electronic structure numerical calculations. The introduction of the values obtained into the rate theory models of retention of hydrogen isotopes has shown a good agreement between the retention depth profile obtained by the model and the experimental data. Unfortunately, the consideration of only one type of dislocations (unlike screw, edge dislocations were not studied) represents a too rough approximation of the exploitation process of a real thermonuclear installation.

In this project the electronic structure calculations using density functional theory (DFT) are proposed in order to evaluate numerically the possible synergetic affinity of hydrogen and helium to screw and edge dislocations. For this, we calculate the interaction energy of a hydrogen/helium atoms and their small clusters with different types of dislocations.

The results of this project are essentially important for the parameterization of upper scale models and/or benchmarking of cohesive models (e.g. interatomic potentials for a W-H-He system) and rate theory methods of retention of plasma components. Moreover, we expect that the obtained results can shed some light on the well known phenomenon of the synergy between He and H interaction in a W matrix which contains lattice imperfections such as dislocations.


Project Title: Specific adsorption from ionic liquid electrolytes on carbon surfaces
Project Leader: Vladislav Ivanistsev, University of Tartu, Estonia
Resource Awarded: 1,350,000 standard core hours

In the present project, we aim to apply molecular dynamics method to investigate the transport in dense ionic electrolytes and adsorption behaviour at carbon electrodes. Adsorption, intercalation and desolvatation process of Li+, Na+, Cl− and I− ions solvated in ionic liquid-based electrolyte will be described by the free energy profiles of solvated ions near a model electrode surface and the interfacial structures visualized at molecular-level. As the adsorption of halide anions and the intercalation of alkali metal ions in carbon electrode are the key features utilized in applications ranging from liquid purification to energy storage, Our main objective is to find the trend in lowering the free energy barriers while varying the chemical composi­tion of the electrolyte.


Project Title: Ordering of Water in Aqueous Solutions: Insights from Atomistic Simulations of Second Harmonic Scattering
Project Leader: Dr. Chungwen Liang, EPFL, Lausanne, Switzerland
Resource Awarded: 5,443,200 standard core hours

Dr. Gabriele Tocci, EPFL, Lausanne, Switzerland
Recently submitted work performed in the laboratory of Prof. Sylvie Roke (EPFL) on elastic second harmonic scattering (ESHS) in dilute electrolyte solutions reveals the presence of extremely longrange (tens of nanometers) correlations in the structure of water. Conventional (e.g. Debye-Huckel) mean-field models do not offer a consistent explanation of experimental findings, so we endeavor to carry out large-scale molecular dynamics (MD) simulations to investigate structural correlations in water induced by ions and calculate the second harmonic signals of the systems, which can be directly compared to ESHS experiments.


Project Title: Water on metal-oxide surfaces from methods beyond DFT
Project Leader: Adam Foster, Aalto University, Espoo, Finland
Resource Awarded: 16,000,000 standard core hours

Understanding the interface of water and metal-oxides is of crucial importance to future applications in clean energy production and CO2 processing. Unfortunately, studying these solid-liquid interfaces experimentally has proven very challenging. On the theoretical side, standard methods used to approach the problem disagree with experiment and are sensitive to the exact simulation setup. Using advanced quantum-mechanical simulation methods, we will reduce uncertainty regarding this pivotal scientific and technological problem.

Plasma & Particle Physics (5)


Project Title: Particle content in a model of a composite Higgs bosons
Project Leader: Prof. Benjamin Svetitsky, Tel Aviv University, School of Physics and Astronomy, Tel Aviv, Israel
Resource Awarded: 4,914,000 standard core hours

Prof. Thomas DeGrand, University of Colorado, Physics, Boulder, USA
Dr. Yigal Shamir, Tel Aviv University, School of Physics and Astronomy, Tel Aviv, Israel
The recent discovery of a light Higgs boson completes the Standard Model of particle physics, as proposed and elaborated over the last fifty years. A consistent picture of the dynamics of particle physics now describes a basic set of particles—quarks, leptons, gauge bosons. At first glance, the simplest mechanism for giving them mass via spontaneous symmetry breaking has now been confirmed. The Higgs boson itself, however, presents us with serious difficulties stemming from a conceptual gap in the Standard Model. Foremost among these is the question of why the Higgs is so light compared to the energy scales that must contribute to its structure. The Standard Model demands extension. Such extensions can be tested, to a degree, by future runs at the Large Hadron Collider and at future accelerators.

We will explore an extension of the Standard Model in which the Higgs boson is composed of more fundamental constituents, just as most particles created in today’s experiments are composed of fundamental quarks. The interaction binding the constituents of the Higgs is similar to quantum chromodynamics, which binds quarks, except that the number of fields and their symmetries are different; moreover, the new interaction is expected to operate on a distance scale smaller by about 10,000 than the characteristic scale of quantum chromodynamics. The Higgs boson should first emerge as a massless bound state of the new strong dynamics. It is subsequently given mass by its interaction with the known particles of the Standard Model, especially the W and Z gauge bosons and the top quark. At the same time, this theory offers a description of the top quark as a partially composite particle, which would explain why it is so heavy compared to other quarks and, indeed, somewhat heavier than the Higgs particle itself.

Like quantum chromodynamics, the theory is a quantum field theory with strongly coupled constituents. This severely limits the ability to study it by analytical methods. We will run simulations with the Hybrid Monte Carlo algorithm to generate configurations of the fields in space-time, and use these to determine the particle content of the theory. The configurations will be useful in later work to study the coupling of these particles to the Standard Model.


Project Title: Kinetic instabilities mediating particle acceleration in collisionless 3D magnetic reconnection
Project Leader: Prof. Dr. Jörg Büchner, Max Planck Institute for Solar System Research, Göttingen, Germany
Resource Awarded: 11,970,000 standard core hours

Dr. Patrick Kilian, North-West University, Centre for Space Research, Potchefstroom, South Africa
Patricio Alejandro Munoz Sepulveda, Max Planck Institute for Solar System Research, Göttingen, Germany
Magnetic reconnection is an essential process for conversion from magnetic into kinetic energy in many space environments. Although it has been studied extensively in the fluid (MHD) framework and in 2D, that it is not true for a fully kinetic 3D case. The kinetic approach, justified due to the collisionless nature of many space plasmas, allows to study, i.e.: electron acceleration generated from micro-instabilities, with important observational signatures such as X-ray emissions. Previous works have shown that magnetic reconnection can be an efficient source of non-thermal electrons in the relativistic regime, although the conditions for that in non-relativistic cases, more suitable for spaces plasmas in the Solar Sytem, are not clear yet.

On the other hand, a 3D geometry allows current aligned instabilities able to produce enhanced micro-turbulence and influence the electron dynamics in magnetic reconnection. However, most of the simulation studies of electron acceleration in magnetic reconnection have remained 2D due to computations constraints, but it is uncertain if the same mechanisms will operate in realistic 3D configurations.

For these reasons, our aim is to study electron acceleration in non-relativistic magnetic reconnection with fully kinetic Particle-in-Cell 3D numerical simulations. Note that these simulations are computationally more expensive than their relativistic counterparts. We propose to clarify the mechanisms generating these non-thermal populations, by relating them with the signatures of kinetic micro-instabilities expected under several parameter ranges. This will allow to improve the understanding of the micro-physics of these essential kinetic plasma processes, with far-reaching consequences for phenomena like solar flares, magnetic substorms and similar processes in the solar wind.


Project Title: Particle diffusion and dispersion in magnetohydrodynamic turbulence
Project Leader: Dr. Yue-Kin Tsang, University of Exeter, Mathematics, Exeter, UK
Resource Awarded: 13,500,000 standard core hours

Magnetised turbulence, which remains a great challenge in theoretical physics, pervades the universe and plays a key role in a variety of astrophysical processes as well as laboratory plasma experiments. The simplest theoretical framework for understanding magnetised turbulence is that of incompressible magnetohydrodynamics (MHD). In this project, we purpose to study numerically transport processes, such as particle diffusion and particle dispersion, in MHD turbulence. Particle transport has application in cosmic rays propagation in the solar wind and the interstellar medium. It is also of relevance to the thermal conduction in galaxy cluster plasmas. Transport processes are equally important in the design and interpretation of fusion experiments. Furthermore, from a theoretical point of view, a Lagrangian analysis following particle trajectories such as the one proposed here may be able to differentiate between competing theories in MHD turbulence that give identical prediction on the Eulerian inertial range scaling of the energy spectrum even though they involve fundamentally different mechanisms.


Project Title: Optimization of nanostructured targets for efficient generation of energetic ions by ultrashort laser pulses
Project Leader: Dr. Ondrej Klimo, Czech Technical University in Prague, Faculty of Nuclear Sciences and Physical Engineering, Prague, Czech Republic
Resource Awarded: 3,015,000 standard core hours

Dr. Edwin Chacon Golcher, Academy of Sciences of the Czech Republic, ELI Beamlines Project Division, Prague, Czech Republic
MSc. Martin Jirka, Czech Technical University in Prague, Faculty of Nuclear Sciences and Physical Engineering, Prague, Czech Republic
Dr. Jan Psikal, Czech Technical University in Prague, Faculty of Nuclear Sciences and Physical Engineering, Prague, Czech Republic
MSc. Jiri Vyskocil, Czech Technical University in Prague, Faculty of Nuclear Sciences and Physical Engineering, Prague, Czech Republic
With conventional technology, kilometer-sized accelerators are required for high-energy physics research and very large installations are also required for medical applications such as particle treatment of tumors. Laser-plasma-based acceleration techniques provide much stronger electric fields (more than four orders of magnitude), leading to the hope that particle accelerators could become much cheaper and more versatile.

Efficient acceleration of ions to high energies is achieved by tightly focusing of an intense laser pulse on the thin solid target [1]. The laser target interaction is essentially three-dimensional in such case, because of strong radial convergence of the laser beam. The dimensionality of plasma expansion has also an effect on maximum ion energy and ion acceleration efficiency [2]. Moreover, a 3D surface microstructure can be used to increase laser absorption and maximum ion energy [3].

Up to now, 3D microstructures employed in experiments were composed from the layer of closelypacked nanospheres on the laser-irradiated side. The nanosphere target is relatively simple for optimization due to only a single free parameter. However, the drawback of such targets may be their relatively large thickness, which reduces the efficiency of laser-ion acceleration [4]. Therefore, we propose to use hollow targets to further improve the efficiency. The possibility of manufacturing these targets using FIB (focused ion beam) has been already demonstrated by our collaborators. Enhanced acceleration has been demonstrated by our numerical simulations in 2D geometry. In the frame of this project, we would like to study the interaction of ultrashort intense laser pulse with such targets and the following ion acceleration in full 3D geometry, which is, however, very demanding for computational resources.

The results of our simulations will enable us to better understand the processes of laser absorption and ion acceleration in this kind of targets and optimize them for future experiments. This is very important because the target fabrication and the experimental investigation on petawatt class lasers are both very costly nowadays. The results of this project will also help us to better prepare for future experiments in terms of diagnostics methods and equipment required.

[1] R. A. Snavely et al., Physical Review Letters 85, 2945 (2000).
[2] A. Henig, dissertation thesis, Ludwig–Maximilians–Universitat Munchen (2010).
[3] D. Margarone et al., Physical Review Letters 109, 234801 (2012).
[4] T. Ceccotti et al., Physical Review Letters 99, 185002 (2007).


Project Title: Magnetic reconnection and turbulence in laboratorial and astrophysical plasmas
Project Leader: Dr. Nuno Loureiro, Massachusetts Institute of Technology (MIT), Department of Nuclear Science & Engineering (NSE), Cambridge, MA, USA
Resource Awarded: 4,812,500 standard core hours

Luis Fazendeiro, Instituto Superior Técnico, Portugal
Prof. Nuno Loureiro, Instituto de Plasmas e Fusao Nuclear (IPFN), Lisbon, Portugal
Magnetic reconnection, the topological reconfiguration of the magnetic field lines in a plasma, accompanied by explosive energy release, is a fundamental unsolved problem that is of vital importance in astrophysical, space and laboratorial phenomena. It can be observed in phenomena as diverse as solar flares, storms in the Earth’s magnetosphere and sawtooth crashes in tokamaks (magnetically confined fusion devices), which can disrupt the whole experiment.

In the context of solar physics, reconnection events can largely influence space weather and thus the safety of satellites and space missions. Therefore, this is a key science topic of the Magnetospheric Multiscale Mission, launched in March 2015, the Solar Probe Plus, to be launched in 2018 (both by NASA) and ESA’s Solar Orbiter (with launch date by 2017). The reconnection process often takes place in a turbulent background, as in the case of the Sun, and the interplay between these two phenomena is very poorly understood. In this work we will study how this background can affect the rate at which reconnection occurs and the energy dissipation mechanisms.

Both magnetic reconnection and plasma turbulence possess an intrinsic multiscale nature and the complexity of the equations describing them poses severe difficulties to existing and foreseeable computational resources, thereby strongly limiting the level of detail of such investigations. There is therefore great interest in exploring reduced models which start from a kinetic model but greatly simplify the resulting equations, whilst maintaining most of the relevant physics.

For this project we shall use a novel, state-of-the-art massively parallel code, Viriato, which numerically solves such a reduced kinetic model. We will identify which fraction of magnetic energy is released into kinetic energy of the bulk flows or heating, the rate at which the reconnection is taking place in 3D systems, as well as the region of the plasma where most heating takes place. With this work we thus aim to provide the broadest understanding of 3D reconnection in a turbulent background yet.