DECI 6th Call


Find below the results of DECI-6 (Distributed European Computing Initiative) call.

Projects from the following research areas:


Astro Sciences (10)


Project Title: Applying Radiation Hydrodynamics to understand Core Collapse supernovae (II)
Project Leader: Hans-Thomas Janka – Max Planck Institute for Astrophysics, Garching, Germany
Resource Awarded: 1,500,000 on FZJ – JuRoPA

Supernova explosions of massive stars are among the most powerful cosmic events. They give birth to neutron stars and stellar black holes, produce strong neutrino and gravitational wave signals, and are the prime source candidates of chemical elements from iron to plutonium. The details of the physical mechanism that leads to the final explosion of the star are not yet fully understood. In this project we perform the currently most advanced simulations of the supernova evolution of massive stars and treat the neutrino-matter interactions in the supernova core with unprecedented accuracy. In this project we aim at continuing our two-dimensional models from the DECI-5 project, and to investigate first steps with three-dimensional models of core collapse supernovae with detailed neutrino-transport.


Project Title: CLUES to Galaxy formation
Project Leader: Gustavo Yepes – Universidad Autónoma de Madrid, Spain
Resource Awarded: 480,000 on HLRS – Laki

Collaborators: Stefan Gottlöber – Astrophysikalisches Institut Potsdam, Potsdam, Germany
Yehuda Hoffman – Hebrew University of Jerusalem, Israel
Steffen Knollmann – Universidad Autónoma de Madrid, Spain
Cecilia Scannapieco – Astrophysikalisches Institut Potsdam, Potsdam, Germany
This is a continuation of a previous DECI project with acronym SIMUGAL-LU. In that project we proposed to perform high resolution N-body plus gas-dynamical re-simulations of the Local Group of galaxies, selected from a large N-body simulation with constrained initial conditions which reproduce, on large scales, the observed matter distribution around our Milky Way galaxy. This simulation to date is the most realistic representation of the formation and evolution of our galaxy and our local environment in the Universe. In this new project we want to go a step further and re-simulate galaxy groups selected from different realisations of the constrained initial conditions to increase the sample size and gain a better understanding of the formation process of the Local Group of galaxies. The simulations use the currently best defined cosmological model and implement very detailed and sophisticated models for the physical processes involved in star formation, like, e.g. cooling, feedback and metal enrichment.


Project Title: Jet Activity in Cosmological Simulations
Project Leader: Dr. Volker Gaibler – Max Planck Institute for Extraterrestrial Physics, Garching, Germany
Resource Awarded: 1,000,000 on LRZ – HLRB II

Collaborators: Dr. Sadegh Khochfar – Max Planck Institute for Extraterrestrial Physics, Garching, Germany
Prof. Joe Silk – University of Oxford, Astrophysics, UK
Cosmological simulations seem to require interaction of extragalactic jets with giant galaxies to explain the observed properties of these galaxies. The jets are extremely powerful and can have an enormous impact on the energy budget of their host galaxy and its environment. Many signs of this feedback can be observed nowadays, but theoretical understanding of it is still poor – expectations range from suppressed star formation and quenched cooling flows to triggering of star formation. To correctly model the effects of jet activity on cosmological simulations, the different gas phases present within and around galaxies have to be taken into account for the jet feedback since star formation is related only to the properties of the cold gas component. In our project, we perform the currently most advanced simulations of spatially resolved jet activity within the dynamical environment of galaxy formation and evolution at high redshift. This will enable us to measure jetinduced changes in both the cold and the hot gas as well as the effects on the galaxy and its environment.


Project Title: Cosmological simulations on many-core architectures
Project Leader: Prof. Dr. Simon Portegies Zwart – University of Leiden, Leiden Observatory, Leiden, The Netherlands
Resource Awarded: 355,000 on CINECA – CNE-SP6, CSC – Louhi XT and FZJ – JUGENE

Collaborators: Prof. Douglas Heggie – University of Edinburgh, School of Mathematics, UK
Kelly Holley-Bockelmann – Vanderbilt University, Nashville, USA
Tomoaki Ishiyama – Center for Computational Astrophysics, Japan
Prof. Jun Makino – Center for Computational Astrophysics, Japan
Prof. Steve McMillan – Drexel University, USA
Keigo Nitadori – RIKEN, Advanced Science Institute, Yokohama, Japan
In astronomy large cold dark matter simulations are used to study the formation of large-scale structures in the universe. Such simulations help us understand details of the universe, such as how galaxies form.

Since we want to study these problems in more detail, we plan to optimize our simulation code to allow runs of 81923 particles on more than 10^5 cores. Simulations of such size require a highly scalable code that is able to use large amounts of processing power (i.e. CPU cores) and computer memory. Current bottlenecks in our code therefore need to be removed. Our objectives are to study and improve the scalability of the code, to determine the performance on different architectures and if possible to conduct some experiments with larger scale meta-computing (current experience with 3 machines).


Project Title: Driven Supersonic Turbulence: Beyond Magnetic Flux Freezing
Project Leader: Dr Turlough Downes – Dublin City University, School of Mathematical Sciences, Dublin, Ireland
Resource Awarded: 990,000 on FZJ – JUGENE

Collaborators: Dr Michael Browne – ICHEC, Galway, Ireland
Virtually all stars, including ones like our own sun, form in cold, dense clouds of gas and dust known as molecular clouds. These clouds are turbulent with the motions of the gas and dust in the cloud being observed to be supersonic. This turbulence influences the way in which stars condense under gravity from the molecular clouds it can stir up the cloud material and disrupt star formation or, perhaps counter- intuitively, it can create regions in a molecular cloud which actually encourage star formation. We know that magnetic fields are present in these molecular clouds, and also that they must have a significant effect on the behaviour of the turbulence in the clouds. Until now most researchers have made the assumption that the magnetic field is perfectly tied to the cloud material: if the cloud material moves it will drag the magnetic field with it and vice versa. This assumption is known as the flux freezing approximation. However, magnetic fields only interact with charged particles in the cloud and so different physics is important for charged particles than that which is important for neutral particles: the molecular cloud behaves like a multifluid with each fluid (charged or neutral) obeying different physics and the flux freezing approximation is, as a result, rendered invalid.
In this project we will study the properties of turbulence in molecular clouds including the complex physics associated with magnetic fields and multifluid effects. We will study the types of structures produced by such turbulence with a view, ultimately, to understanding more about the evolution of molecular clouds and about how stars like our own sun form.


Project Title: The dynamics of black holes: testing the limits of Einsteins theory
Project Leader: Vitor Cardoso – Instituto Superior Técnico, Portugal
Resource Awarded: 1,540,000 on LRZ – HLRB II

Collaborators: Leonardo Gualtieri – Universita di Roma la Sapienza, Department of Physics, Italy
Carlos Herdeiro – Universidade do Porto, Porto, Portugal
Ulrich Sperhake – California Institute of Technology, Pasadena, USA
From astrophysics to high-energy physics and quantum gravity, black holes (BHs) have acquired an ever increasing role in fundamental physics. From an astrophysical perspective, it has been established that supermassive BHs lurk at the center of many galaxies and provide fertile ground for stellar growth and evolution. Millions of stellar-mass BHs populate the galaxies, power violent processes such as gamma-ray bursts and represent the strongest source of gravitational waves to be observed with interferometers such as LIGO, VIRGO and the planned space mission LISA. In high-energy physics, the gauge-gravity duality has created a powerful framework for the study of strongly coupled gauge theories and lead to applications in connection with the experimental program on heavy ion collisions at the Relativistic Heavy Ion Collider (RHIC) and CERNs Large Hadron Collider (LHC), among many others. BHs play a special role in the duality; for instance, the confinement/deconfinement phase transition in QCD may be related to the Hawking-Page phase transition of BHs in anti-de Sitter space-times. Given the central role that BHs have been claiming in physics, a major task for theoreticians is to understand processes in which they are involved. With the advent of techniques to evolve BH spacetimes numerically,the field is undergoing a phase transition from a promising branch of general relativity to one of the most exciting fields in 21st century research, one that will open up unprecedented opportunities to expand and test our understanding of fundamental physics and the universe. The goal of this project is the numerical evolution of BHs in generic backgrounds, in a fully non-linear framework. We focus our investigations on high-energy collisions of BHs and on the evolution of astrophysical spinning binaries in quasi-circular orbits. The former are of fundamental importance in high-energy particle collisions, specifically for event generators, which will use data from accelerators to test TeV-scale gravity. Accurate, long-term waveforms of astrophysical binaries of spinning BHs, on the other hand, are important for data analysis of gravitational-wave detectors. The list of further studies facilitated by our simulations ranges from tests of the cosmic censorship to an understanding of the stability and phase diagrams of black objects in generic spacetimes


Project Title: Magneticum Pathfinder, towards the next generation of cosmological structure formation simulations
Project Leader: Klaus Dolag – Max Planck Institute for Astrophysics, Garching, Germany
Resource Awarded: 1,200,000 on SARA – Huygens P6

Collaborators: Stefano Borgani – University of Trieste, Astronomy Unit, Italy
Joseph Mohr – Ludwig-Maximilians Universität München, Sternwarte, Germany
This project aims to follow the formation of cosmological structures in a so far unaccomplished level of detail by performing one large scale plus one highresolution simulation, taking into account many physical processes to allow detailed comparison to a variety of multi-wavelength observational data. We view this as a pathfinder simulation towards what will be needed as the ultimate theoretical counterpart of upcoming, large volume and multiple wavelength astronomical surveys from instruments like Planck, SPT, Pan-STARRs, LOFAR, eROSITA and many more. In combinations with datasets from these surveys, our proposed simulations will enable forefront studies of the formation and evolution of galaxies and galaxy clusters as well as the study of dark energy and the origin of cosmic acceleration.


Project Title: Pathways to A Realistic Milky Way
Project Leader: Brad Gibson – University of Central Lancashire, Preston, UK
Resource Awarded: 1,295,000 on BSC – MareNostrum and CINECA – CNE-SP6

The formation of disk galaxies such as our own Milky Way is one of the major problems facing the current cosmological paradigm, in which the Universe is dominated by dark energy and cold dark matter (CDM). Models which are based on structure formation within the CDM cosmologies have invariably failed to reproduce some of the most important properties of disk galaxies. Recent advances in our ability to implement and resolve star formation and supernova explosions have been shown to improve our ability to model low mass disk galaxies. Extending these tantalizing results to a simulated galaxy the size of the Milky Way is only possible using supercomputers like those available in the DEISA infrastructure. Such a successful simulation of the Milky Way will allow direct comparison with the myriad of observations on the Milky Way, and provide an interpretive framework for understanding its formation and evolution.


Project Title: Modelling solar flares and active regions
Project Leader: Prof. Åke Nordlund – University of Copenhagen, Niels Bohr Institute, Denmark
Resource Awarded: 2,400,000 on FZJ – JuRoPA and HLRS – Laki

Collaborators: Remo Collet – Max Planck Institute for Astrophysics, Garching, Germany
Prof. Fernando Moreno-Insertis – Instituto de Astrofisica de Canarias, Tenerife, Spain
Prof. Göran Scharmer – Instutute for Solar Physics, Stockholm, Sweden
Prof. Robert F. Stein – Michigan State University, East Lansing, USA
The Sun is a fascinating astrophysical object, given its closeness to us, its remarkable level of activity, and the intricate physics of its different layers. The study of the Sun and heliosphere is necessary to understand the structure and evolution of the stars; it is also of practical importance to know the environment in which the Earth moves and the perturbations to which its magnetosphere is subjected (space weather). Right now Europe has a unique opportunity for progress thanks to a new generation of computational tools and observational facilities with unprecedented capabilities and to recent advances in theory.

The overarching aim of this research is to understand the dynamics of solar flares and solar active regions – sunspots and the neighbourhoods of sunspots – and their interaction with the solar convection zone, on scales that range from less than 10 km to more than 100,000 km, using both 3-D MHD simulations and 3-D relativistic charged particle (particle-in-cell code) simulations to realistically model solar active region physics. The results from these simulations are to be used for modelling and interpreting direct observations of solar active region phenomena, including solar flares and the general heating of the solar corona, and also to function as test beds for rigorous validation of helioseismic inversion methods that allow observational diagnostics of sub-surface solar layers.

The project uses mainly two well-proven MPI- (and hybrid MPI/OpenMP-) codes that parallelize well to thousands of CPU cores; a staggered mesh code (the Copenhagen StaggerCode) for MHD-simulations, and the Copenhagen PhotonPlasma Code, which is a relativistic particle-in-cell (PIC) code with modular provisions for particle-particle interactions (Coulomb collisions, Bremsstrahlung, Compton scattering, etc.).


Project Title: Turbulence and Angular Momentum Transport in Accreting Magnetized Disks
Project Leader: Gianluigi Bodo – INAF Osservatorio Astronomico di Torino, Italy
Resource Awarded: 3,750,000 on FZJ – JuRoPA and RZG – RZG P6

Collaborators: Fausto Cattaneo – University of Chicago, Chicago, USA
Andrea Mignone – Università degli Studi di Torino, Italy
Accretion is a fundamental process for many astrophysical objects and the associated release of gravitational energy can power some of the most energetic phenomena in the universe. Accretion mainly occurs in the form of a disk and a basic problem of the theory is to determine how the material in the disc can effectively accrete losing its angular momentum. This process of angular momentum transport in the accretion disk is currently interpreted as the result of magnetohydrodynamical turbulence driven by the magnetorotational instability. Recent results have however questioned the effectiveness of such mechanism, showing a decrease of the transport with the increase of the Reynolds number, leading to the possible conclusion that in astrophysical conditions it cannot explaine the needed angular momentum transport. With this project, by using numerical simulations of unprecedented resolution and accuracy and introducing the new effect of stratification. we aim at answering the crucial question whether there is a universal regime of magnetorotational turbulence, whose properties are independent of dissipation and external condition and whether the resulting transport is effectively of astrophysical significance.


Bio Sciences (13)


Project Title: Protein systems spanning two membranes
Project Leader: Prof.Dr. Ulrich Kleinekathöfer – Jacobs University, Bremen, Germany
Resource Awarded: 2,800,000 on HLRS – Laki

Collaborators: Prof.Dr. Paolo Ruggerone – Università degli Studi di Cagliari, Italy
The study of membrane proteins embedded in their native environment, i.e., lipid bilayers and water is numerically still quite computer demanding and requires some effort especially when longer simulation times are needed. Extremely challenging are the simulations of protein complexes that span over two membranes. For example, gram-negative bacteria have an outer and an inner membrane and some efflux pump systems, the protein machineries responsible for the occurrence of multidrug resistance, do work across both membranes. Bacterial resistance against antibiotics is becoming a real threat in treating several diseases these days. The availability of the atomic structure of several efflux pumps or reliable models via homology opens the possibility to study the structure-function relationship computational via large-scale molecular dynamics (MD) simulations. A simulation of the whole complex incl. patches of the outer as well as the inner membrane is needed to understand the cooperative effects of the individual proteins. Some functions of the individual proteins are at least partially understood by now, but it is clear that the function of the pump needs the interplay of the individual proteins. The large size of the system and the complexity of the interaction among the different parts of the pump make the investigation challenging and very time-consuming as well. The Italian as well as the German groups have experience in modelling the influx of ions and antibiotics through cell membranes and furthermore in studying parts of the efflux system. The function of one transporter involved in the efflux was recently studied in a joint effort. Now, the complete efflux system consisting of more than half a million atoms shall be attacked using the MD code NAMD developed by the Schulten group. This code has excellent parallel scaling properties making it an ideal tool for making use of the DEISA infrastructure in an European project which would be otherwise impossible.


Project Title: Building bigger dinosaurs: the mechanics of terrestrial giants
Project Leader: Dr William Sellers – University of Manchester, Manchester, UK
Resource Awarded: 2,400,000 on CINECA – CNE-SP6

Collaborators: Dr Lee Margetts – University of Manchester, Manchester, UK
Size is hugely important in biology. Almost any feature you care to measure about an organism has some sort of correlation with body size. The mechanics of animals is no exception. As animals get bigger their legs get longer and their muscles get stronger as you would expect. Thanks to the work of various engineers over the centuries we have a very good understanding of how mechanics change as things get bigger. We know that as things get larger they get stronger but not as quickly as they get heavier. That means that the strength to weight ratio of big things is lower than the strength to weight ratio in small things and we need to take this into account when scaling up from models to the real thing. It is no different in animals. If we consider the group of animals that have skeletons and live on the land we see that the adult size varies from 2 g to 7000 kg in modern animals, and perhaps up to as much as 100 tonnes if we include dinosaurs. This is a huge range when we consider that the materials that the animals are made from are largely identical and is particularly impressive for the very largest animals where the strength to weight ratios are lowest. What we want to know is whether these largest animals were limited in size by their mechanics. They had to move around so there would have been enormous forces acting on the skeletons and it is entirely possible that the largest of the dinosaurs were actually as big as they could be in terms of mechanics. Access to the DEISA Supercomputing Grid will mean that we can test this proposition directly.


Project Title: High Throughput in-silico screening in HPC architectures for new inhibitors for treatment of blood diseases
Project Leader: PD Dr. Wolfgang Wenzel – Karlsruhe Institute of Technology, Karlsruhe, Germany
Resource Awarded: 1,900,000 on EPCC – HECToR QC2 and EPCC – HECToR XT6

Collaborators: Dr. Horacio Pérez-Sánchez – Karlsruhe Institute of Technology, Karlsruhe, Germany
Based on the stochastic tunneling method (STUN), a novel strategy for high-throughput in-silico screening of large ligand databases has been developed, named FlexScreen. Each ligand of the database is docked against the receptor using an all-atom representation, and the ligands with the best evaluated affinity are selected as lead candidates for drug development. Using the thymidine kinase inhibitors, the shortcomings of rigid receptor screens in a realistic system were documented. A gain in both overall binding energy and overall rank of the known substrates when two screens with a rigid and flexible (up to 15 sidechain dihedral angles) receptor are compared, is demonstrated. STUN suffers only a small loss of efficiency when an increasing number of receptor degrees of freedom is considered. FlexScreen thus offers a viable compromise between docking flexibility and computational efficiency to perform fully automated database screens on hundreds of thousands of ligands. Enrichment rates of rigid, soft and flexible receptor models for 12 diverse receptors were also investigated. A flexible sidechain model for up to 12 aminoacids increased both binding propensity and enrichment rates: EF_1 values increased by 35% on average with respect to rigid-docking. Using this flexible receptor model we are interested in the investigation of new inhibitors for the treatment of blood diseases like blood coating. Antithrombin is the main target in our study, as it regulates clotting when interacting with Thrombin. At the moment mainly heparin-based inhibitors have been designed, but they have some undesirable side effects that can complicate the disease treatment. Using our approach we intend to screen large ligand databases, obtain hits, and refine them in order to design new inhibitors against blood clotting and with no side effects, therefore we need the use of HPC resources in order to carry out all the necessary docking calculations and beyond we seek to use infrastructures services provided by DEISA in terms of UNICORE and OGSA-DAI.


Project Title: Scaling of cortex simulations on large cluster computers
Project Leader: Prof. Anders Lansner – KTH, Sweden
Resource Awarded: 700,000 on FZJ – JUGENE and RZG – Genius

Collaborators: Prof. Markus Diesmann – Forschungszentrum Jülich, Germany
Even the honey bee brain has on the order of a million neurons communicating via a thousand times as many synapses and it has a very intricate structure. The human brain is about 10000 times larger and much more complex. The size and complexity of the brain simulation models are severely limited by the restrictions imposed by the capacity of personal and small-sized parallel computers. This proposal concerns running brain simulations on a very large Blue Gene/P cluster supercomputer.

Very large-scale simulations are necessary in order to better understand the neural dynamics and information processing going on in the brain, since these are to a large extent globally coordinated phenomena. Such simulations will also help understand and diagnose diseases and dysfunctions of the brain and give knowledge important for drug development and other therapy.

Neural network simulations have a potential to scale very well on cluster architectures since most computations are local and long-range communication is event based. Our goal is to establish a workflow for very large scale to full-scale brain simulations. We aim to simulate mouse cortex sized networks (20 million neurons, 100 billion synapses) comprising biophysically detailed components as well as a realistic cortical architecture on 64K cores of the Blue Gene/P. These brain simulations will be among the largest ever done. They will feature more realism in terms of components and network structure than other simulation models and they will perform as an associative memory, which is an important brain function. European research in the area of brain simulation and neuromorphic engineering is internationally leading and our proposal is well connected to current European funded activities. Moreover, plans are currently in place for proposing an EC flagship project on brain science including simulation and it is therefore important to initiate these activities now.


Project Title: Unravelling the molecular mechanisms of quantum coherence preservation in biological systems.
Project Leader: Dr. Aurélien de la Lande – Université Paris-Sud 11 – CNRS, Laboratoire de Chimie Physique, Orsay, France
Resource Awarded: 660,000 on PDC – Ekman

Collaborators: Dr. Jan Rezac – Institute of Organic Chemistry and Biochemistry, Prague, Czech Republic
The preservation of coherences between entangled quantum states is a major issue in Quantum Informatics, Material Sciences and Chemistry. When chemical systems are concerned one has to consider a quantum system, composed of electrons, embedded in a classical bath of atomic nuclei. When quantum-classical couplings are neglected the above separation stems at the root of the Born-Oppenheimer approximation, a widely accepted paradigm in chemistry. On the other hand, non Born-Oppenheimer reactions such as Spin Crossing reactions (SCR), Long Range Electron Transfers (LRET) or Exciton Energy Transfers (EET) represent important classes of chemical processes for which the explicit coupling between the quantum and the classical systems need to be taken into account (open quantum systems). This is a question of fundamental interest that certainly hasn’t received enough attention in the past, as testified by, the most recent literature devoted to biological EET. Indeed several systems have been shown to preserve quantum coherences between electronic quantum states for hundreds of femtoseconds at room temperature.
Our project is to better understand the molecular mechanisms that allow such long lived coherences and their impact on chemical observables (such as chemical rate constants). We will resort to state-of-the-art DFT-Born Oppenheimer Molecular Dynamics simulations that we have modified to model such effects. The computational approach will be applied to electron transfer reactions between proteins.


Project Title: Computer-aided drug design of cyclophilins ligands by massively parallel free energy calculations
Project Leader: Dr Julien Michel – University of Edinburgh, (Institute of Structural and Molecular Biology, Edinburgh, UK
Resource Awarded: 2,180,840 on LRZ – HLRB II

Collaborators: Dr Frank R. Beierlein – FAU Erlangen-Nürnberg, Computer Chemistry Center, Erlangen, Germany
Dr Christopher J. Woods – University of Bristol, School of Chemistry, Bristol, UK
A crucial step in preclinical pharmaceutical research is the design of small molecules that bind with strong affinity and high specificity to a target protein implicated in a disease. In this context, inexpensive computer models are routinely used to screen databases of thousands of chemicals and identify putative small-molecule ligands. However, the methodologies used to estimate the binding affinity of a trial molecule are currently very limited in accuracy, and lack the reliability necessary to predict binding selectivity.


Atomically detailed, thermodynamic simulations of small molecules binding to a protein have the potential to deliver high accuracy predictions of binding affinity and selectivity. We have recently developed free energy calculation methodologies to reliably compute binding affinities for datasets of diverse drug-like molecules. The methods are promising as they show superior accuracy over standard techniques. We are now in the position to assess the true predictive power of our methodologies. However, because they are exceptionally computationally intensive, state-of-the art case studies have been limited to small datasets.

We will use the computing resources offered via DEISA to scale-up our software and develop workflows so we can robustly predict binding affinities for large datasets of small molecules across multiple protein targets. This will allow us to not only validate the power of our methodology to predict binding affinities, but also to demonstrate the prediction of binding selectivity. Our efforts will focus on a family of proteins called cyclophilins, which are the subject of structure-based drug design efforts in our experimental collaborator`s lab. Our simulations will be run in parallel, and will be tightly coupled with experiment. Use of DEISA resources may therefore contribute to the discovery of novel ligands targeting these proteins, potentially useful for novel drug therapies.


Project Title: Molecular dynamics simulations of target-site recognition in protein-DNA vomplexes
Project Leader: Dr. Petra Imhof – University of Heidelberg, Germany
Resource Awarded: 700,000 on CSC – Louhi XT and RZG – Genius

Biological function of site-specific DNA-binding proteins largely depends on the proteins´ ability to quickly scan the DNA chain and reach the target site. The current view of the sequence searching mechanism is a mixture of one-dimensional diffusion of the protein along the DNA (sliding) and three-dimensional movement, i. e. jumps to other DNA sites. However, the underlying mechanism for sliding and location of the target site is yet unknown. In particular, the influence of the DNA sequence on the search mechanism of the protein is not fully understood. Here, we propose atomistic computer simulations to investigate the location and recognition of a target site on the DNA, i.e. a specific sequence, by the restriction enzyme EcoRV. Interactions between the restriction enzyme and different DNA sequences will be examined using molecular dynamics (MD) simulations. In order to obtain a comprehensive picture of the recognition process, step-wise movement along the DNA (sliding) as well as binding and unbinding at different sequences will be simulated.


Project Title: Free energy calculations of enzymatic reactions for biotechnological applications
Project Leader: Pietro Vidossich – Universitat Autònoma de Barcelona, Unitat de Química Física, Barcelona, Spain
Resource Awarded: 200,000 on FZJ – JuRoPA

Collaborators: Bernd Ensing – University of Amsterdam, Van Hoff Institute for Molecular Sciences, Amsterdam, The Netherlands
Understanding structure-function relationships in heme proteins is a key step towards the development of biotechnological applications. Molecular modeling may contribute valuable insight in this respect providing the atomic and electronic rearrangements that take place during catalysis and the forces that guide them. Our project targets the enzyme chlorite dismutase (Cld), the catalyst of the last step of the bacterial perchlorate respiratory pathway i.e. the conversion of chlorite to chloride and dioxygen (ClO2- → Cl- + O2, reaction 1). Perchlorate-reducing bacteria have recently become important subjects of research because of their potential use in remediation of contaminated water. Our objective is to pinpoint the structural determinants that make Cld an efficient and selective catalyst for reaction 1 (as opposed to other heme proteins). To the scope we will use ab initio molecular dynamics simulations, based on DFT/MM interacting potentials, and a new free energy method based on metadynamics aimed at reducing the computational cost of these type of calculations.


Project Title: Improving ligand binding free energy estimate through enhanced sampling algorithms
Project Leader: Andrea Cavalli – Italian Institute of Technology (IIT), , Genova, Italy
Resource Awarded: 1,560,000 on PDC – Ekman

Collaborators: Michele Parrinello – ETH Zurich, Switzerland
An accurate estimation of the protein-ligand binding free energy is of paramount importance in drug discovery, as it can provide the affinity constant of a small organic molecule (i.e., a drug candidate) towards its biological counterpart according to the following equation: dG = –RTlogK. Once the binding affinities (K) are computationally predicted, the molecules can be proposed for chemical synthesis and pharmacological evaluation, thus facilitating and enhancing the drug discovery process. Unfortunately, due to the logarithmic relationship between free energy and binding affinity, the acceptable error must be less than 1 kcal/mol in the dG estimation. Currently available docking algorithms routinely used at both academic and industrial levels are usually unable to provide such accuracy because i) the protein is often treated as a rigid body, ii) empirical scoring functions are implemented to estimate the interaction energy between ligands and proteins, and iii) the entropy is neglected or roughly parameterized. With the present project we propose to develop and exploit a novel computational protocol that combines enhanced sampling (ES) methods with molecular dynamics (MD) simulations. In brief, the main objective of the project is to devise an ES/MD-based procedure able to provide a reliable (error less than 1 kcal/mol) estimate of protein-ligand binding free energy. This tool will therefore be able to predict the affinity of small but complex organic molecules for purposely selected pharmacologically relevant target proteins. . Such an objective will certainly have a great impact for both the academic and industrial drug discovery community.


Project Title: Metabolomics through multi-genetic associations
Project Leader: Prof. Karsten Suhre – Ludwig-Maximilians-Universität München, Department of Biology II, Germany
Resource Awarded: 330,000 on CSC – Louhi XT and EPCC – HECToR QC2

Collaborators: Dr. Christian Gieger – Institute for Epidemiology (EPI), GSF – National Centre for Environment and Health, Germany
Dr. Bernet Kato – Kings College, London, UK
In a previous DEISA project, we have successfully used high-performance computing for an innovative genome-wide association study in metabolomics [Gieger et al., Illig et al.]. Due to the availability of the DEISA computing resources, we were able to initiate a paradigm shift – a hypothesis-free approach test a large number of potential hypothesis, the set of which represents all possible combinations of metabolite ratios (>20,000) on 500,000 genetic loci for thousands of individuals. We are presently preparing the follow-up study based on new original dataset that has been funded by the German Diabetes Centre. In biological terms we are expecting to discover a new range of genetically determined metabotypes (see our previous report for details), as a totally new metabolite panel shall be tested. In terms of computational requirements it will be at least comparable to our previous project: the number of metabolites will increase from 163 to over 250. The number of SNPs may be over 1,000,000. Moreover, we plan to test new methods for hypothesis testing.


Project Title: Kinetics of signalling proteins
Project Leader: Prof.Dr. Peter Bolhuis – University of Amsterdam, Van Hoff Institute for Molecular Sciences, Amsterdam, The Netherlands
Resource Awarded: on SARA – Huygens P6

Collaborators: Dr. Jocelyne Vreede – University of Amsterdam, Van Hoff Institute for Molecular Sciences, Amsterdam, The Netherlands
For living organisms it is essential that they can respond to external stimuli. At the molecular level these stimuli often induce functional changes in the structure of a protein. An important example of such receptor proteins is the Photoactive Yellow Protein (PYP) that causes bacteria to escape from harmful irradiation. UV-A/ Blue light causes the protein to change structure through a chain of events. The transition toward the signalling state in the photocycle of PYP is a poorly understood reaction network of conformational changes. Molecular dynamics simulation provides in principle a mechanistically detailed picture of the functioning of a signalling protein in action, but is hampered by the millisecond time scale of the process. Therefore we will employ powerful advanced path-sampling methods that enable the study of this network in molecular detail. In particular, we aim to combine the replica exchange method with the Transition Interface Sampling technique to elucidate the rate constants, as well as the free energy landscape, and the relevant reaction coordinates. In this way we will obtain more insight in the functioning of PYP, and of signalling proteins in general. In addition, this work will open op the way for studying the kinetic of large-scale conformational changes in proteins.


Project Title: Parallel-Scalable Single-Particle Analysis in 3D Electron Microscopy
Project Leader: Carlos Sánchez – Centro Superior de Investigaciones Científicas (CSIC), National Centre of Biotechnology, Spain
Resource Awarded: 100,800 on BSC – MareNostrum

Collaborators: Slavica Jonic – Institut de Minéralogie et de Physique des Milieux Condensés, Paris, France
Structural biology aims at the elucidation of the three-dimensional (3D) structure of biological macromolecular complexes in order to fully understand its function in the live cell. A widely used approach to collect such structural information is imaging tens of thousands of copies of the same complex in an electron microscope. This is called Single-Particle Analysis. These images are extremely noisy and deformed due to the imperfect optical system. Therefore, they have to be intensively processed to obtain the 3D structure at high resolution that is essential for studying the relation structure-function. The whole field is usually called 3D Electron Microscopy (3DEM).
Currently, solving a structure at high (subnanometer) resolution typically takes 50000 CPU-hours. Note that this time would be even longer without the parallelization of the most time consuming steps. Such long elapsed CPU time is mostly due to the fact that the code is not maximally optimized but also to the fact that the same sequence of image analysis steps has to be repeated several times for the same structure (e.g., the procedure has to be repeated with new images collected for a new specimen that is purer or more stable biochemically or the procedure has to be repeated with new values for the parameters involved in different image analysis steps). Although not much can be done as a remedy to the latter problem, there is still an important gain in the speed that can be obtained if the code is optimized.. Also, one should always have in mind that future structural biology problems will require to process even larger data sets, which will necessitates applications with high throughput (large amount of work performed within a given time). For instance, one has to process 50000-100000 images of isolated particles in each run for a currently typical high-resolution structure (0.8-1 nm resolution). We estimate that one will need 5-10 times more images to achieve quasi-atomic resolution (0.4 nm) in the case of completely asymmetric structures.
Over the last 20 years, we have developed at the National Centre of Biotechnology (CSIC) an open-source software package (Xmipp, to address this 3D reconstruction problem. The package is publicly available and it is used by structural biology groups all over the world. The package includes parallel capabilities to reduce the computing time. In this project we aim at optimizing the steps involved in the 3D analysis of single-particles, which shall transform the XMIPP into a high throughput application. Indeed, the computing time with optimized XMIPP will be further reduced. This will increase the productivity of structural biologists and optimize the use of available.
The group at the National Centre of Biotechnology (CSIC) has been appointed to lead the Image Processing Centre for Microscopy associated to INSTRUCT (the European Infrastructure for Structural Biology), thanks in part to the image processing developments that are collected in the aforementioned software package. The objective of such centre is to provide support to structural biologists all over Europe in performing image analysis steps towards the computation of high-resolution (0.5-0.8 nm) 3D structure of macromolecular assemblies of biological interest. The high-resolution structure is essential for studying the function of these macromolecules in living cells. Typically, the atomic-resolution structures (0.1-0.2 nm) of the domains of a macromolecular complex computed by other experimental methods (X-ray crystallography or NMR) or theoretical methods (prediction) will be positioned into the structure obtained by electron microscopy and the regions corresponding to the interface between the neighbouring domains will be studied to understand what allows these domains to assemble and what is their role in the complex.


Project Title: Simulations of Influenza Viral Entry
Project Leader: Prof. Erik Lindahl – Stockholm University, CBR – The Stockholm Center for Biomembrane Research , Stockholm, Sweden
Resource Awarded: 4,200,000

Collaborators: Dr. Peter Kasson – Stockholm University, CBR – The Stockholm Center for Biomembrane Research , Stockholm, Sweden
Membrane fusion, the process by which neuronal exocytosis and infection by enveloped viruses occur, has been notoriously difficult to characterize at a molecular level. Part of the problem is that the underlying reaction that fusion proteins catalyze is not fully understood. The development of robust predictive models for the mechanism of lipid membrane fusion and its catalysis by viral fusion proteins will greatly aid in the understanding of the underlying physical process and how to effectively target it with antiviral agents. We have developing high-performance simulation methods to analyze membrane fusion. In our work thus far, we have simulated vesicle fusion at atomic resolution, yielding novel insight into structure and mechanism of fusion intermediates. We are now extending these simulations to generate high-fidelity models of fusion in a experimental model systems, and predict the catalytic mechanism of influenza fusion proteins. Simulations are performed using the Gromacs software package that we and our collaborators develop, one of the fastest in the world.


Earth Sciences (4)


Project Title: Global and local Air pollution transport over the Mediterranean basin and the Middle East
Project Leader: Dr. Andrea Pozzer – The Cyprus Institute, Energy, Environment and Water Research Center, Cyprus
Resource Awarded: 2,400,000 on PDC – Ekman

Collaborators: Dr. Marina Astitha – The Cyprus Institute, Energy, Environment and Water Research Center, Cyprus
Dr. Alexander De Meij – The Cyprus Institute, Energy, Environment and Water Research Center, Cyprus
Dr. Kirsty Pringle – University of Leeds, School of Earth and Environment, UK
Dr. Holger Tost – Max Planck Institute for Chemistry, Germany
Air quality is a very important issue in the Mediterranean basin, due to the strong urbanisation and industrialisation of the surrounding countries. Atmospheric chemistry models are used to assess and estimate the impact of natural and anthropogenic gas and aerosol emissions on air quality and climate change.
The region is not only influenced by local pollution (such as industry, traffic and ship emissions), but also by the transboundary transport of air pollution. To assess the impact of both processes on the atmospheric composition of the Mediterranean region, the model is required to perform global simulations at high spatial resolution. In addition, the complex chemical interactions of the aerosol components and their importance in the region entail the usage of a highly complex model which can cover different time and spatial scale processes. These requirements hence lead to the need of an extremely high demanding simulation in terms of computational power and data storage capability. The DEISA grid infrastructure offers an ideal environment to make such computation and analysis.
The proposed project aims in investigating the regional and global interactions between air pollution, long range transport and perturbations of the water cycle with a focus on the Mediterranean region and the Middle East. A high resolution simulation of an atmospheric chemistry climate model resolving complex aerosol processes will enable us to study the mechanism governing the Mediterranean weather and air quality.


Project Title: Global unstructured simulations
Project Leader: Dr. Martin Käser – Ludwig-Maximilians-Universität München, Department of Earth and Environmental Sciences, Geophysics, Germany
Resource Awarded: 960,000 on CINECA – CNE-SP6 and SARA – Huygens P6

Collaborators: Dr. Martin Mai – King Abdullah University of Science and Technology, Thuwal, Kingdom of Saudi Arabia
The major discoveries of the interior of our planet have been achieved through the study of seismic waves. To this end, the discipline of global seismology is analyzing the translational and recently also the rotational movement of the ground occurring globally after an earthquake. Through an inversion process, tomographic images of the earth interior can be produced and converted into petrophysical properties. The key ingredient in this process is the simulation of wave propagation through planetary scale models with 3-D variations of anisotropic, viscoelastic properties and density. The main objective of this proposal is to apply a well-advanced simulation tool based on the discontinuous Galerkin method (SeisSol) to the problem of global wave propagation and inversion. This project will be tightly linked to the recently initiated ITN QUEST ( network which has a focus on seismic imaging tools on all scales.


Project Title: Impact of the Indonesian Throughflow on the Atlantic Meridional Overturning Circulation
Project Leader: H.A. Dijkstra – Utrecht University, IMAU, The Netherlands
Resource Awarded: 2,745,000 on EPCC – HECToR XT6 and IDRIS – BABEL

Collaborators: Ir. Michael Kliphuis – Utrecht University, IMAU, The Netherlands
MSc. Dewi Le Bars – Utrecht University, IMAU, The Netherlands
In the ITAMOC project, we aim to study the connection between the Indonesian Throughflow (ITF) and the Atlantic Meridional Overturning Circulation (MOC) using a very high-resolution global ocean model. The strength of the ITF may affect the transport through the Mozambique Channel and hence that of the Agulhas Current more southward. The strength of the Agulhas Current controls the leakage of heat and salt from the Indian Ocean to the Atlantic Ocean through Agulhas rings. This interbasin salt transport is thought to be a crucial factor in controlling the strength of the MOC. We will investigate the connection between ITF and MOC using the Parallel Ocean Program (POP) model with a horizontal resolution of 0.1 degrees and 45 vertical levels. In addition to a control type simulation of 20 years under climatological forcing, two basic simulations are planned under observed daily atmospheric forcing over the period 1990-2010: one with realistic representation of the Indonesian Archipelago and one with a closure of the Indonesian Throughflow.


Project Title: Performance optimisation and reproducibility analysis of distributed simulations of the Earth climate system
Project Leader: Dr. Patrick Jöckel – DLR (German Aerospace Center), Institute for Atmospheric Physics, Oberpfaffenhofen, Germany
Resource Awarded: 830,000 on CINECA – CNE-SP6, CSC – Louhi XT and HLRS – NEC SX-9

Comprehensive Earth System Models (ESMs) are powerful scientific tools for the investigation of the processes and feedback mechanisms in the climate system and for assessing the system response to natural and anthropogenic perturbations. The ECHAM/MESSy Atmospheric Chemistry (EMAC) model system is specifically designed to allow a wide variety of model configurations with tailor made complexity for different scientific tasks. The requirements on the high-performance computing (HPC) system vary with the degree of complexity and the spatial and temporal resolution. In consequence, it is expected that the optimal (i.e., most efficient) HPC system differs for different model configurations. For an efficient exploitation of computational resources, in particular within a heterogeneous HPC grid, it is therefore desirable to select the best suited HPC system for each simulation setup. Further advantages for ESM simulations from a HPC grid arise for distributed ensemble simulations, distributed sensitivity simulations, and the distributed piecewise simulation of long time scales. The prerequisite is, however, the independence of the simulated climatology from the HPC system – a reproducibility characteristic, which is a-priori not guaranteed, due to the (deterministic) chaotic nature of the ESM and the peculiarities of the different HPC systems. The DEISA grid computing infrastructure provides an ideal frame for the efficient analysis (and potential optimisation) of the run-time performance of different EMAC setups on different HPC systems, and further to test the statistical independence (i.e., their reproducibility) of the results on the HPC system.


Engineering (9)


Project Title: Acoustic Coupling in Engineering
Project Leader: Prof. Dr.-Ing. Sabine Roller – German Research School for Simulation Sciences GmbH, Aachen, Germany
Resource Awarded: 537,528 on CSC – Louhi XT

Direct Aero-Acoustic simulations are a challenging computational task, as many different scales have to be included in a single computation.The phenomena generating the noise in the flow as vortices are small scale structures, whereas the propagation of the sound waves needs to be observed over a large distance. However direct simulations have the advantage, that they require the least modelling and respect interactions between the acoustic and the flow. However a decomposition of the simulation into specialized parts is attractive in order to reduce the computational effort and allow different models and scales. As the acoustic noise generation and the sound propagation are often separatable in space, it is possible to apply a spatial decomposition with an appropiate coupling. With this scheme it is feasible to compute more realistic models, incorporating many physical phenomena, yet the computing time needed for such simulations is still very large. This project has the aim to enable acoustic simulations in the engineering context for complex geometries with the help of modern computer science developments. One of the deployed kernels can take advantage of CoArray Fortran.


Project Title: Multi-Scale Simulation of Aeroacoustic Turbulent Flows over Porous Surfaces
Project Leader: Prof. Dr.-Ing. habil. Manfred Krafczyk – Technische Universität Braunschweig, Braunschweig, Germany
Resource Awarded: 750,000 on CINECA – CNE-SP6 and RZG – Genius

The study of aerodynamic flows generating noise pollution is very important for many classes of applications, one being flows around aeroplanes during landing and take-off. The central hypothesis of the present project is that noise generated by such high-Reynolds-number flows around moving or rigid structures, can be significantly reduced through permeable (porous) materials partially covering specific areas of the wings and tail units or of the jet engines. The main research objective of the project is, thus, to gain detailed insight into the aero-acoustics and the dynamics of turbulent flows inside and around porous surfaces, with the aim to increase the aerodynamic efficiency of planes and to decrease the acoustic pollution. This project proposes direct numerical simulations (DNS) of flows in porous media coupled to large-eddy simulations (LES) in the near-wall region of the boundary layer. In order to analyze the behaviour of different porous materials, the acoustic far field is computed by a standard CFD solver based on a RANS model of turbulence. This will be coupled to a DNS-LES solver based on a Lattice Boltzmann method. In order to reflect the coupled physics at all disparate scales involved, it is necessary to conduct a series of multi-billion-DOF simulations. To that end, It is crucial to find optimal HPC platforms and utilize them most effectively, whereby the variety of massively parallel platforms and the accompanying support provided by DEISA are indispensable.


Project Title: Boiling Fluids
Project Leader: Luca Biferale – University of Rome Tor Vergata, Dipartimento di Fisica, Italy
Resource Awarded: 3,600,000 on CINECA – CNE-SP6 and RZG – Genius

Collaborators: Prof.Dr. Federico Toschi – Technical University of Eindhoven, Department of Applied Physics, Eindhoven, The Netherlands
Roberto Verzicco – University of Rome Tor Vergata, Italy
Rayleigh-Benard (RB) convection the buoyancy of a fluid heated from below is a classical problem in fluid dynamics. From an applied view-point, thermally driven flows are of utmost importance. Examples are thermal convection in the atmosphere, in the oceans, in metal-production processes in Earth`s mantle, in stars, and the reversal of Earth`s magnetic field. In this classical problem, the flow is determined by the container geometry, the material properties of the fluid, and the top-down temperature difference. A fundamental physical question is the dependence of the heat transfer rate for a given temperature difference between the bottom and top plates, for a given fluid, and given geometry. In the last two decades, considerable progress has been achieved in the understanding of global and local properties for the flow organization of turbulent convection through a combination of experimental, numerical and theoretical works. All this was done especially under the approximation due to Oberbeck and Boussinesq, assuming that the transport and expansion coefficients are constant throughout the fluid and that the temperature dependence of the density is linearized in the buoyancy force. This is justified when one tends to restrict the convection regime to sufficiently small intervals of the temperature drops. Much less is known when compressible and/or multi-phase flows are concerned. We propose here to make fully resolved simulations, using a new Lattice Boltzmann Method for Thermal Flows [1], addressing thermal convection in 3 dimensional compressible non-ideal flows, close and in presence to phase coexistence (boiling). The main goal consist in assessing the effects of droplet condensation/evaporation on the global heat flux and on the multiscale structure of the flow close to the boundary and in the bulk.


Project Title: High Resolution DataBases for CFD modeling
Project Leader: Dr. Vincent Moureau – Université de Rouen – CNRS – INSA Rouen, CORIA, Rouen, France
Resource Awarded: 2,118,600 on CINECA – CNE-SP6

Abstract: The CFD-DB project aims at the building of high-fidelity databases of turbulent reactive flow simulations. These databases are dedicated to the understanding and the validation of turbulence models in complex geometries. The simulations are performed with the finite-volume solver YALES2, based on innovative technologies to handle very large meshes of billions of cells and to solve efficiently in parallel large linear systems. This research code, developed at CORIA, is designed to model turbulent reactive flows and atomization of liquid fuels with body-fitted unstructured meshes. Such databases have already been successfully built in 2009 for an industrial swirl burner with 2.6 billion cells. The objective of the current proposal is to make a big step forward, breaking the 10 billion-cell mesh and 100k cores limits, i.e. approximately four times bigger than the present databases. Such a resolution is needed to resolve all the scales of turbulent flows at realistic Reynolds numbers.


Project Title: Control Of Separated Unsteady Flow
Project Leader: Christian Engfer – German SOFIA Institute, Stuttgart, Germany
Resource Awarded: 1,990,000 on LRZ – HLRB II and SARA – Huygens P6

Collaborators: Sebastian Illi – Universität Stuttgart, Institut für Aerodynamik und Gasdynamik, Germany
NASA and the DLR, the German Aerospace Center are working together to create and operate the Stratospheric Observatory for Infrared Astronomy (SOFIA), a reflecting telescope located inside a Boeing 747SP aircraft. During observation in the stratosphere, a door in the fuselage will be opened to expose the infrared telescope to the atmosphere. In general, the flow over cavities such as the SOFIA telescope port is characterized by unsteady flow phenomena associated with prominent pressure fluctuations caused by amplified acoustic resonances. In the present case, this phenomenon causes unwanted vibrations of the telescope structure and reduces its pointing stability. Flow control in the case of SOFIA is performed by a slanted rear wall that stabilizes the shear layer’s impingement point and prevents it from oscillating violently.
The objective of the present numerical investigations is to estimate the potential to increase the performance of the system by active and passive flow control methods. These methods allow changing the shear layer’s characteristics and permit to shift the acoustic resonant frequencies away from resonance modes of the telescope structure.


Project Title: Distributed particle transport simulation in a Grid-like HPC CFD environment
Project Leader: Prof.Dr. Hans-Joachim Bungartz – Technische Universität München, Institut für Informatik, Garching bei München, Germany
Dr. Tobias Weinzierl – Technische Universität München, Institut für Informatik, Garching bei München, Germany
Resource Awarded: 520,000 on FZJ – JUGENE

Collaborators: Dinesh K. Kaushik – King Abdullah University of Science and Technology, Thuwal, Kingdom of Saudi Arabia
David E. Keyes – King Abdullah University of Science and Technology, Thuwal, Kingdom of Saudi Arabia
Dr. Ioan Lucian Muntean – Technical University of Cluj-Napoca, Cluj-Napoca, Romania
The DiParTS project (Distributed Particle Transport Simulation in a Grid-like HPC CFD Environment) numerically studies particles dispersed in non-stationary fluids within tube-like geometries on the micro-scale, where the fluid and, as a consequence, the particles are stimulated by an oscillating pressure. The particles’ long-time behaviour due to the pressure oscillations, i.e. their averaged movement on the long-term time-scale, allows us to draw conclusions, for example, on the causes of particle sedimentary deposition and centrifugal particle separation in several applications, as the particles exhibit a drift along the stimulation amplitude. Here, classical fluid-structure interaction phenomena interplay with Brownian motion and particle-wall interaction. In a preceding DEISA project, we already studied simplified experimental setups on the short-time time-scale. Despite some promising and interesting insights from a fluid-dynamics point of view, the full simulation of the situation described above however proved to be far from solvable with today’s computing power. Due to this proposal, we nevertheless will broaden the horizon of computability, as we switch from a fully coupled system to an approach where the fluid simulation without particles on an extremely fine spatial and temporal resolution is cut into small time intervals, these chunks of computational challenges are deployed to supercomputers, and the fluid fields are coarsened spatially before the supercomputer streams the data back to the scientist’s local workstation where it is post-processed, i.e. the Brownian motion and the particles’ effect are remotely added to the flow field after the fluid dynamics time step has terminated. The extreme computing power spent on this waterfall process – in particular on the fine-scale fluid dynamics simulation – will yield new insights on the long-time behaviour of the overall simulation setup, while the approach is validated simultaneously by a comparison of a fully-coupled fluid-interaction setting with the decoupled simulation for several small time steps.


Project Title: Extreme Computing for Advanced Methods of Solving Partial Differential Equations
Project Leader: Dr Lee Margetts – University of Manchester, Manchester, UK
Resource Awarded: 2,300,000 on BSC – MareNostrum, CINECA – CNE-SP6, HLRS – Laki and RZG – Genius

Collaborators: Prof. Stephane Bordas – University of Cardiff, Cardiff, UK
Prof. Marc Duflot – Cenaero, CFD and mutiphysics group, Gosselies, Belgium
Prof Oubay Hassan – University of Swansea, Swansea, UK
Prof. Guillaume Houzeaux – Barcelona Supercomputing Center (BSC-CNS), Barcelona, Spain
Dr Paul Mummery – University of Manchester, Manchester, UK
Prof. Timon Rabczuk – University of Weimar, Weimar, Germany
Today`s global challenges in the energy, aerospace and biotechnology sectors require extreme engineering approaches that take into account the complex interplay of different physical processes that operate at multiple length and time scales. The effective collaboration of domain scientists who have access to a European infrastructure for supercomputing applications is essential to accelerate research in these areas, providing the simulation tools that will enable European industry to be highly innovative and internationally competitive.

The project has two broad aims. Firstly, it will foster a network of European scientists who are researching advanced methods for solving partial differential equations. The methods include, but are not limited to, the finite element method, the extended finite element method and meshfree methods. Secondly, using the DECI resources requested, the team will implement and deploy a �Virtual Laboratory� for materials characterisation that will be used to evaluate the response of exemplar high performance engineering materials to a range of insilico tests. The materials will be scanned at a high resolution using a state of the art X-ray tomography scanning facility at the University of Manchester. The resulting 3D images will be converted into micro-structurally faithful computer models. Finally, stress, thermal, vibration and fracture mechanics tests will be carried out, not one after another, but at the same time using multiple supercomputers simultaneously.

This work is particularly exciting because extremes of temperature, pressure and vibration, which are experienced by materials in the energy, space and aerospace sectors, are difficult and costly to recreate in the laboratory. In contrast, with the necessary computing infrastructure, they are easier to simulate.

After the DECI funded project, the newly established EC4aPDEs network will continue to recruit new members, with the aim of developing and sharing a common application software infrastructure that will make use of future European extreme computing resources.


Project Title: Fluid Dynamics of Film Cooling Investigated by Large-Eddy Simulation
Project Leader: Prof. L. Kleiser – ETH Zurich, Institute of Fluid Dynamics, Switzerland
Resource Awarded: 600,000 on HLRS – NEC SX-9 and HLRS – NEC SX-8

Collaborators: Prof. U. Rist – Universität Stuttgart, Institut für Aerodynamik und Gasdynamik, Germany
High blade temperatures limit the performance of gas turbines. Film cooling is used to lower the blade temperature by ejecting cold gas through holes in the blade surface directly into the boundary layer. The resulting flow configuration is very complex with respect to the geometry and the involved physical phenomena. Numerical simulations of film cooling are mostly done using Reynolds averaged Navier-Stokes (RANS) simulations which by their nature are incapable of capturing the full physics of this problem. Therefore, we investigate this flow by large-eddy simulations (LES) which are able to resolve the essential flow structures and the large-scale turbulence in space and time. For modelling the effect of subgrid scales we use our approximate deconvolution model (ADM).

We aim at simulating film-cooling configurations with an accurately represented geometry, increasing the level of reality step by step. This will allow us to investigate in detail flow regimes (e.g. high temperatures) that are hard to capture with experiments. Although LES need only about one percent of the computational time of corresponding direct numerical simulations (DNS), they still require exceptionally large computing resources.


Project Title: Inertial Particles in developing wall turbulence
Project Leader: Dr. L. Brandt – KTH, Sweden
Prof. Carlo M. Casciola – Universita di Roma la Sapienza, Dipartimento di Meccanica e Aeronautica, Italy
Resource Awarded: 2,100,000 on EPCC – HECToR XT6 and EPCC – HECToR QC2

Collaborators: J.G.M. Kuerten – Technical University of Eindhoven, The Netherlands
Prof.Dr. Federico Toschi – Technical University of Eindhoven, Department of Applied Physics, Eindhoven, The Netherlands
Traditionally, direct numerical simulations (DNS) of turbulent flows are performed in simplified computational domains that are characterised by periodic boundary conditions in all three directions. However, real applications in nature and technology often involve the interaction with a solid wall and are thus inhomogeneous is space. Here, we study the flow case of a spatially evolving turbulent boundary layer and focus on the combined advection of inertial particles. Such a computational study based on highly resolved DNS and not yet attempted in the literature, bears many interesting and relevant physical effects due to the growing boundary layer; for instance the non-dimensional number characterising the particle-wall accumulation is gradually changing with the downstream distance. The raw scientific data is planned to be shared with the scientific community (iCFDdatabase,

From a computational point of view, spatially developing flows are necessarily investigated in very long domains in order to accurately capture the whole streamwise extent of the (physical) flow; another problem is the reliable and trustworthy generation of inflow turbulence in a non-periodic setting. It is thus only recently that the DNS of turbulent boundary layers has become feasible. The additional complexity of coupling the advection of inertial particles leads to large computational demands on massively parallel computers.


Materials Science (10)


Project Title: Azobenzene photoisomerization in liquid crystal mesophases
Project Leader: Prof. Claudio Zannoni – Università di Bologna e INSTM, Italy
Resource Awarded: 2,880,000 on CINECA – CNE-SP6 and RZG – RZG P6

Azobenzene (AB) molecule is the prototype of a class of materials whose shape and properties can be rapidly modified, upon illumination, thanks to a trans-cis interconversion. This photo-responsive process is key to a number of applications, in which either the mechanical effects generated by the conformational change or the changes in physical properties are exploited [1-3]; however the actual conversion mechanism in condensed phases is still largely unknown. It is particularly important to clarify the mechanism when AB is dissolved in a low molar mass or elastomeric liquid crystal [2,3], since their order and molecular correlation amplify this relatively weak molecular effect into mesoscopic or even macroscopically observable opto-mechanical effects [2]. We have recently developed a molecular dynamics based methodology suitable for the problem [5] and we propose to apply it to provide the first attack on this important problem. More in detail, the project is aimed at extending a recent simulation study of azobenzene isomerization in vacuum and in various solvents [3] to the more complex environment provided by liquid crystalline phases, namely the nematic and smectic phase of the well known mesogen 8CB (4-n- octyl-4’-cyano biphenyl), whose properties have already been well reproduced by atomistic simulations [5]. We plan to use HPC- parallel molecular dynamics (MD) simulations and classical force fields to describe both the ground state of azobenzene and 8CB, and the excited state of azobenzene.
The expected results are twofold: on one hand we aim to predict the shift of transition temperatures between the different phases of the LC solvent induced by the two isomers of azobenzene solute observed experimentally [6]; on the other we will assess the effect of the anisotropy of the medium and of the smectic formation on the isomerization extent, on the quantum yield and on the photoisomerization mechanism.


Project Title: Quantum-Mechanical Prediction of Gas Phase Biomolecular Secondary Structure
Project Leader: Dr. Volker Blum – Max Planck Society, Fritz-Haber-Institute, Berlin, Germany
Prof. Matthias Scheffler – Max Planck Society, Fritz-Haber-Institute, Berlin, Germany
Resource Awarded: 1,500,000 on CINECA – CNE-SP6 and RZG – RZG P6

Collaborators: Prof. Ville Havu, Helsinki University of Technology, Institute of Mathematics, Finland
The accurate computational prediction of peptide / protein behaviour is among the most competitive research areas today, but fully QM predictions (here: van der Waals corrected density functional theory) for such systems are still hampered by the computational demands for the necessary molecular dynamics timescales. DEISA will allow us to advance these limits significantly, by predicting conformational equilibria and ensembles of two benchmark peptides, Ac-Ala19-Lys-H+ and (Ac-Lys-Ala19-H+)2 (220 and 440 atoms, respectively). Key parts of the results can be directly compared to accurate gas-phase experimental data (vibrational spectroscopy), and will provide a key check and corroboration of the current ideas of secondary structure formation and its driving force in such molecules.


Project Title: Rational Design of TiO2 Surfaces with Novel Functionalities: DFT Study
Project Leader: Prof. Dr. Herbert Over – Justus-Liebig-Universität Giessen, Giessen, Germany
Resource Awarded: 1,200,000 on CSC – Louhi XT and SARA – Huygens P6

Collaborators: Dr. Ari Paavo Seitsonen – CNRS & Université Pierre et Marie Curie, France
Recently the Japanese company Sumitomo Chemical has discovered an efficient and stable Deacon catalyst on the basis of RuO2 supported on TiO2, a true breakthrough in current catalysis research. In order to identify a promising alternative catalyst for the Sumitomo Process, rational material design on the basis of density functional theory (DFT) and nudged elastic band (NEB) calculations will be applied. In DFT the electronic structure is solved explicitly, leading to a vast increase in the required computing time, but also increasing the reliability and predictability of the results. Motivated by experimental data we are investigating TiO2, a simple and inexpensive material, and the selective substitution of its anionic sublattice surface sites by chlorine, sulphur and nitrogen. The complex mechanism of the selective replacement reactions will be studied starting from molecular precursors containing hydrogen, such as HCl, H2S and NH3. Hydrogen-containing molecular precursors are required to remove bridging oxygen atoms from the surface in the form of water as the leaving group and subsequently fill in these vacant sites. The activity and stability of the modified TiO2 surfaces will be tested in the Deacon process, i.e. the heterogeneously catalysed oxidation of HCl by molecular oxygen.
The proposed simulations will allow the numerical optimisation of the chemical reaction paths implied above. The forces on the ions are obtained by solving the Kohn-Sham equations numerically, in a self-consistent cycle, whereby the electronic wave functions expanded in a plane-wave basis set. Due to the vast number of wave function coefficients required for reliable results in such calculations, the usage of massively parallel computer architectures, such as those provided by the DEISA infrastructure, is required. The envisioned reaction path optimization on the basis of in silico experiments represents a paradigm shift with much broader implications than the improved, much more detailed understanding of the particular reactions studied. Besides fundamental aspects the present project is driven by industrial applications. At the moment, Sumitomo Chemical holds practically all patents relevant to the RuO2 based Deacon Process. It is therefore important for the European chemical industry to find an alternative material for the Sumitomo Process.


Project Title: Deprotonation of organic molecules in solution: study of the reaction mechanism and rate by modern techniques for simulating rare events in combination with ab-initio molecular dynamics.
Project Leader: Simone Meloni – University College Dublin, Dublin, Ireland
Resource Awarded: 2,100,000 on LRZ – HLRB II

Collaborators: Sara Bonella – Università Sapienza, Rome, Italy
Alin Elena – ICHEC, Dublin, Ireland
Deprotonation of organic molecules are key processes involved in the rate limiting step of many chemical and biological processes. Their reaction mechanism depends on several factors such as the nature of the acid functional group (carboxyl, hydroxyl, etc.), the substituents of the molecule, and the type of solvent. However, despite their importance, these processes remain not well understood. In this project we shall study the deprotonation of different classes of organic molecules by ab-initio simulations, elucidating also the role of substitutes of the molecules. Standard atomistic simulations techniques, such as molecular dynamics (MD), cannot be used for simulating these processes by bruteforce because of the time scale involved: deprotonation is a rare event. This problem will be overcame using modern `sampling` techniques, such as Temperature Accelerated Molecular Dynamics, the String Method, and Milestoning, adapting them to the study of deprotonation process and, more generally, chemical reactions. These techniques allow to compute the reaction rate, characterize its mechanism(s), and calculate its free energy along the reaction path (i.e. free energy barrier) of chemical reactions in realistic conditions (i.e. in solution).


Project Title: The optical properties of group IV semiconductor nanocrystals – an ab initio many body perturbation approach
Project Leader: Adam Gali – Budapest University of Technology and Economics, Budapest, Hungary
Resource Awarded: 1,263,800 on CINECA – CNE-SP6

Collaborators: Maurizia Palummo – University of Rome Tor Vergata, Italy
During the last decade, several efforts have been made to exploit the special properties of low dimensional systems, e.g. nanocrystals. Among the countless potential projects we will concentrate on two interesting group IV type semiconductor systems: (i) Recent developments allow the size and shape selected preparation of small diamond nanocrystals (diamondoids). These diamondoids exhibit several properties that would make them ideal for some specific purpose application. (ii) Recently, it has been shown that silicon carbide nanocrystals could be used as efficient environment friendly biomarkers. Since further applications require detailed knowledge of the optical properties of such semiconductor nanocrystals, theoretical ab initio investigations are needed. We will calculate the optical properties of diamond and silicon carbide nanocrystals by the elaborate density functional theory (DFT) and many-body perturbation theory (‘GW+BSE’) technique. While this method has provided superior results over more simple methods (e.g. time-dependent density functional theory), the demanding computational resources usually restricted the application to unrealistically small systems. Harnessing the computational capacity of supercomputers allow the usage of many-body perturbation theory to real problems with the capability to compare results with experimental findings and previous theoretical results


Project Title: Full Configuration Interaction Quantum Monte Carlo study of the Uniform Electron Gas
Project Leader: Ali Alavi – Cambridge University, UK
Resource Awarded: 2,460,000 on CSCS – Rosa

A newly developed Quantum Monte Carlo method which is based on sampling Slater determinants, in which the Fermion sign problem is exactly solved through annihilation processes (as opposed to fixed-node type approximations), is applied to the Uniform Electron Gas (UEG). This method will shed light onto the increasingly multiconfigurational nature of the many-electron wavefunction of the UEG as the density is decreased. Comparison will be made with the Diffusion Monte Carlo method for this system, which has long provided the benchmark energies for the UEG. In addition, the applicability of the new method to condensed-matter problems will be assessed, as well as the suitability of the code on extremely large-scale parallel machines, as afforded by DEISA facilities.


Project Title: First-principles determination of the meltinbehaviour of hydrogen at high pressures
Project Leader: Eduardo Hernández – Centro Superior de Investigaciones Científicas (CSIC), ICMM, Spain
Resource Awarded: 360,000 on CINECA – CNE-SP6

Collaborators: Aitor Bergara – Universidad del País Vasco, Vitoria, Spain
Marcin S. Kaczmarski – University of Osnabrueck, Osnabrueck, Germany
Yanming Ma – National Lab of Superhard Materials, Changchun, China
Chris J. Pickard – University College London, London, UK
This project aims to predict the melting behaviour of hydrogen up to pressures of 4 Mbar (400 GPa) from first principles coexistence simulations. In so doing we will be providing detailed information of the thermodynamics of hydrogen at extreme pressures, information that is crucial to further our understanding of matter at such conditions, and which will be essential for the development of geological models of the interior of the large gaseous planets in the solar system. Currently the melting line of hydrogen has been explored up to pressures of 1.5 Mbar, so its properties at higher pressures remains unexplored. In particular, theoretical predictions have suggested that, at higher pressures hydrogen could become a low- temperature metallic liquid or even a high-temperature superconductor, but it has been thus far impossible to confirm these predictions experimentally. With this project, we will provide new theoretical data that will complement the rather incomplete picture that we have at present about the behaviour of hydrogen at extreme conditions. Hopefully this will allow us to answer questions such as Does hydrogen become a liquid metal at low temperatures? Does it instead follow a continuous sequence of solid phases?, questions that are some of the most fundamental in condensed matter physics and materials science.


Project Title: Water splitting catalysts for artificial photosynthesis
Project Leader: Simone Piccinin – National Research Council (CNR), INFM DEMOCRITOS National Simulation Center, Italy
Resource Awarded: 1,560,000 on CSC – Louhi XT

Collaborators: Alessandro laio – Scuola Internazionale Superiore di Studi Avanzati, Italy
Artificial photosynthesis is one of the most promising methods for the direct conversion of solar energy into renewable chemical fuels. The process involves splitting water by first creating spatially separated electron-hole pairs, which then drive the redox semi-reactions leading to the evolution of molecular hydrogen and oxygen. This project aims at providing an atomistic understanding of the mechanism of water oxidation, the bottleneck of the overall process. To this end, we will use state-of-the-art first-principles numerical modeling based on density functional theory. In particular, we will focus on inorganic ruthenium-containing polyoxometalate homogeneous catalysts that have been recently synthesized and that display unprecedented reactivity and stability in solution. Given the complexity of this four-electron process, to date it has not been possible to determine experimentally a reliable mechanistic model for this reaction. Here we will employ large scale metadynamics simulations to explore the free energy surface as a function of a few key collective variables, to determine the most likely reaction path, the associated activation energies and the key properties of the catalyst affecting its performance. The necessity to employ hybrid functionals to describe the oxidation and reduction of the metal centers makes the project extremely demanding from a computational point of view, requiring the exceptional DEISA infrastructure to be carried out. The activity will be linked to the EU FP7-PEOPLE-IRG-2008 grant awarded on the same topic.


Project Title: How strong are materials?
Project Leader: Mikko Alava – Aalto University, Department of Applied Physics, Finland
Resource Awarded: 1,600,000 on PDC – Ekman

Collaborators: Stefano Zapperi – National Research Council (CNR), Italy
The design and choice of materials that need strength depends on understanding how this property depends on the size of the object or specimen, and how it varies from sample to sample. The research on these issues combines various viewpoints from engineering mechanics, the statistics of extremes, and modern statistical physics. The statistical distribution obeyed by strength depends on the microscopic properties of the material, and on the large-scale features like how the load is applied. In the HoSAM project, we utilize the simulation of simple yet comprehensive fracture models to study the sample size-dependent scaling of material strength. Necessary ingredients include the variation of the microscopic material characteristics – the disorder that gives rise to the stochastic nature of the problem – and the study of the effect of defects or micro-cracks. By the application of the Deisa resources, we may extend such simulations over a substantial range of sample sizes and to acquire reliable data of the properties of the probability distributions describing strength, in all such cases.


Project Title: Superconductivity from Magnetism
Project Leader: Alessandra Continenza – CNR – Università dell Aquila, Dipartimento di Fisica, L Aquila, Italy
Resource Awarded: 1,650,000 on CINECA – CNE-SP6

Collaborators: Christophe Bersier – Max-Planck-Institut für Mikrostrukturphysik, Halle, Germany
Cesare Franchini – University of Vienna, Austria
Sandro Massidda – Università degli Studi di Cagliari, Italy
The recent and unexpected discovery of iron-based superconductors brought a new revolution in condensed matter physics posing once more on the spot the interplay between magnetism and superconductivity. High-temperature superconductivity (above 50 K) was observed in many FeAs-based compounds and this renewed the search for novel superconductors suitable for large-scale applications. However, as usual in these contexts, many unanswered fundamental questions on the normal and superconducting phases of these materials hamper a full comprehension and the possibility to envisage a well determined route towards the discovery of new promising compounds.
A remarkable step forward in predicting superconductivity from first-principles has been achieved by the SuperConducting Density Functional Theory (SCDFT), able to describe on a fully ab-initio ground the properties of the superconducting phase of so called conventionalsuperconductors. The present project proposes a novel theoretical and computational approach to describe unconventional pairing mechanisms, as the mechanism of superconductivity in many superconductors is presently still unknown.
Our objective is twofold:
1) calculate the effective electronic pairing interaction from first-principles methods, taking into account all the details of the structural and electronic properties of real materials, and extend SCDFT so to include possible electron-only pairing mechanisms, such as charge- and spin-fluctuations
2) use this new computational facility to uncover the key role that structural, electronic and magnetic properties play in setting the critical temperature and to understand how these could be manipulated (e.g. through doping, substitutions, defects..) to design new materials for technological applications.
The theoretical/computational tool implemented will allow to predict superconductivity accounting for both electronic and phononic pairing mechanisms, both calculated on the basis of material specific properties, thus providing a complete description of superconductivity (critical temperature, gap symmetry, and many other experimental accessible quantities) based on fully ab-initio calculations.


Plasma & Particle Physics (10)


Project Title: Global electromagnetic gyrokinetic simulation in 3D equilibria
Project Leader: Dr. Ralf Kleiber – Max Planck Institute of Plasma Physics (IPP), Greifswald, Germany
Resource Awarded: 1,500,000 on CINECA – CNE-SP6

Collaborators: Dr. E. Sánchez Gonzáles – CIEMAT para Fusión, Spain
Prof. Laurent Villard – CRPP-EPFL, Lausanne, Switzerland
Modern stellarator experiments, as e.g. Wendelstein 7-X, need accompanying simulations of plasma microinstabilities and related turbulence. Especially important for the comparison of experiment and theory are full torus simulations for stellarator configurations. Gyrokinetics as a first principle based theory is well suited to describe the relevant physics. An established and flexible method for solving the gyrokinetic system of equations is the simulation via the particle-in-cell (PIC) Monte-Carlo method. For this purpose the EUTERPE code has been established which solves the (linear/nonlinear) gyrokinetic equation globally in arbitrary stellarator geometry including kinetic electrons, and electromagnetic perturbations. The full kinetic treatment of the electrons allows the investigation of trapped electron effects (e.g. TEM) and the inclusion of electromagnetic effects establishes the connection to magnetohydrodynamics (MHD). These developments will make EUTERPE the first code worldwide that is able to simulate global gyrokinetic electromagnetic instabilities in three dimensions. By using a third species the destabilisation of MHD modes (e.g. TAE, HAE) can also be studied.


Project Title: Full scale simulations of Fast Ignition for Fusion Energy
Project Leader: Luís O. Silva – Instituto Superior Técnico, Portugal
Resource Awarded: 1,155,000 on FZJ – JUGENE

Collaborators: Warren Mori – University of California, Los Angeles, USA
Energy is the ultimate driver for economic growth and social development. Fusion energy is regarded as a possible long-term energy solution for humanity that is environment-friendly and safe. Fast ignition is one of the most promising and exciting inertial confinement fusion schemes to improve the viability of inertial fusion energy as a practical energy source. Up to now, experiments have been limited to laser energies still far from ideal conditions for ignition and simulations, which are extremely complex due to the different temporal and spatial scales involved, have been limited to reduced scales/simplified models. Novel laser systems with unprecedented energies are now coming online, with the National Ignition Facility and, in the near future, the ESFRI roadmap project HiPER (High Power laser Energy Research), reaching the conditions required for ignition. In this proposal, and using massively parallel simulations, we aim to perform, for the first time with realistic target properties (e.g. density, temperature, dimensions) and the correct simulation dimensionality, self-consistent fast ignition simulations including all the relevant microphysics/particle dynamics, with the particle-in-cell code OSIRIS, with the goal of identifying possible paths to demonstrate fast ignition as an efficient scheme for inertial fusion energy.


Project Title: High Precision Lattice QCD II
Project Leader: Prof.Dr. Zoltán Fodor – Universität Wuppertal, Fachbereich C – Physik, Germany
Resource Awarded: 3,100,000 on FZJ – JuRoPA

Collaborators: Dr. Sandor Katz – Eötvös Loránd University, Institute for Physics, Budapest, Hungary
Dr. Laurent Lellouch – Centre de Physique Théorique, Marseille, France
With the advent of Petascale supercomputers, high precision Lattice QCD (LQCD) calculations finally have become possible, more than 30 years after Wilson`s seminal papers introducing the approach. This is a time of opportunity, where controlled and precise calculations for many observables are feasible for the first time. The key ingredient is reducing extrapolation ranges by simulating directly at the physical pion mass value for a number of different lattice spacings, which is, however, feasible only with our new improved simulation algorithms.
Since our pioneering calculation of the light hadron mass spectrum with fully controlled systematic errors (Science 322, 1224, Nov. 2008), we have now reached the “physical point` (i.e. tuned the pion mass down to its physical value at 135 MeV) for two lattice spacings. Here, we propose to extend our set of gauge field ensembles by one additional ensemble with fine lattice spacing and physical pion masses and at least one ensemble at an even finer lattice spacing. The former data point will help us control chiral extrapolations at this lattice spacing or make these superfluous altogether. The latter will help us to better control the continuum limit. This, together, will allow us to greatly improve the precision and the predictive power of our continuum extrapolated physics results.
Our main physics aim is to compute highly precise estimates of the u-, d- and s- quark masses. These fundamental parameters of Standard Model of Elementary Particle Physics are known only to very low accuracy, a situation that we hope to improve on dramatically with our simulation. Furthermore, we hope to compute precision estimates of hadron masses, weak matrix elements, hadron structure functions and other quantities of phenomenological importance.


Project Title: Perturbation theory for Lattice QCD
Project Leader: Dr. Alistair Hart – University of Edinburgh, UK
Resource Awarded: 768,000 on PDC – Ekman

Collaborators: Prof. Ronald Horgan – Cambridge University, UK
Dr. Georg von Hippel – DESY-Zeuthen, Germany
Lattice QCD calculations are crucial to understanding and interpreting the results from particle collider experiments such as at the upcoming LHC in CERN. The simulations are numerically very expensive and require high-capability HPC resources. Theoretical calculations can be carried out to improve the accuracy and efficiency of the simulations. These supporting, �improvement� calculations can be carried out using capacity HPC resources.

In this project we continue to use DEISA resources to carry out such improvement calculations to enable lattice QCD simulations to uncover signals of new, `Beyond the Standard Model` physics in experimental data. We principally focus on improvements for simulations describing the physically interesting decays of heavy B-mesons. In doing so, we will develop optimised numerical integration tools for use more widely in scientific calculations on a variety of supercomputer architectures.


Project Title: Hadron Structure from Lattice QCD
Project Leader: Prof. Constantia Alexandrou – University of Cyprus, Department of Physics, Nicosia, Cyprus
Resource Awarded: 1,500,000 on FZJ – JuRoPA

Collaborators: Dr. Eric B. Gregory – University of Cyprus, Department of Physics, Nicosia, Cyprus
Dr. Giannis Koutsou – Forschungszentrum Jülich, Germany
Dr. Yiannis Proestos – The Cyprus Institute, Nicosia, Cyprus
Antonios Tsapalis – National Technical University of Athens / Hellenic Naval Academy, Athens, Greece
We utilize Lattice QCD, the only non-perturbative scheme which allows for the exploration of Hadron Structure starting from the fundamental Lagrangian of the theory. Simulating QCD is a challenging and demanding computational task, ideally suited for high-capacity HPC platforms such as those provided by DEISA. The primary computational cost of Lattice QCD calculations outlined in this proposal is the inversion of the large, sparse Dirac matrix, which is a function of the quark mass. Generally Lattice QCD calculations are performed at heavier masses than at the physical quark masses. Computational cost and physical accuracy greatly increase with decreased quark mass (as measured by the mass of the resulting pions).
Utilization of DEISA computing resources becomes crucial, especially when the simulations involve the use of fine dynamical Domain Wall Fermion (DWF) lattices (32^3 x 64 x16) at pion mass as low as 297 MeV. DWF is a discretization scheme which improves the chiral properties of the fermions at increased computational cost (5-dimensional lattice instead of four).
In particular, this project makes use of the resources provided by DEISA in order to employ lattice QCD techniques and perform hadron matrix calculations, which explore the electromagnetic and weak transition form factors for the Nucleon-to-Delta process, the Delta electromagnetic and axial form factors and the Omega baryon electromagnetic form factors. Computation of the highly sensitive suppressed form factors is expected to shed light on the deformation of baryons.
The extraction of the quadrupole form factors in Nucleon-to-Delta transition and as well as in the Delta-Delta connected to deformation requires very high statistics making the problem computationally demanding. To improve the accuracy for the calculation of subdominant form factors we employ the coherent-sink technique which quadruples the statistics per sequential inversion.
Finally the study involves also an exploratory computation where the gauge noise dominated disconnected three-point functions are taken into account. The inclusion of such quantities is important, for instance in the case of the Delta and Omega electromagnetic form factors as well as in studies of decays of unstable particles such as the Delta. To explore methods for the computation of isoscalar form factors, keeping the computational cost as low as possible, the measurements of such quantities, which involve the computation of all-to-all propagators, are performed using two dynamical flavors of Wilson fermions on 32^3x 64.


Project Title: Isoscalar meson spectroscopy from lattice quantum chromodynamics
Project Leader: Dr Michael Peardon – Trinity College Dublin, Dublin, Ireland
Resource Awarded: 2,400,000 on CSCS – Rosa

Collaborators: Dr Robert Edwards – Thomas Jefferson National Accelerator Facility, Newport News, USA
The isoscalar mesons are bound states of quarks and anti-quarks with particularly simple up and down quark flavour structure. Because of this simplicity, these mesons can have the same quantum numbers as other states predicted by QCD made up solely of gluons; the glueballs. Since they share common quantum numbers, QCD allows them to mix and understanding experimental signatures for these states is therefore very challenging since it is unclear whether particles observed in experiments are glueballs or simpler quark-anti-quark admixtures.

This project will attempt the first spectroscopy determination of isoscalar mesons using lattice QCD combined with a powerful new method developed by the Hadron Spectrum Collaboration called `distillation`. The computation will make use of the collaboration`s large ensemble of gauge field configurations generated on an anisotropic lattice with the full quark-field dynamics included in the importance sampling measure. The method will be used to determine the lowest- lying states in this spectrum with unprecedented precision and examine whether the observed experimental states can be wholly explained as quark-anti-quark bound states.


Project Title: Particle simulation of tokamak plasma edge
Project Leader: Jukka Heikkinen – VTT (Euratom-Tekes association), Espoo, Finland
Resource Awarded: 1,760,000 on FZJ – JuRoPA

Collaborators: Timo Kiviniemi – Aalto University (Euraton-Tekes association), Espoo, Finland
Francisco Ogando – Spanish National University for Distance Education (UNED), Spain
Understanding the plasma turbulence is of major importance for success of ITER, which is in turn the most important step in the development of fusion energy. ELMFIRE is a full particle distribution gyrokinetic simulation code developed to study the dynamics of turbulence and its influence on plasma global behaviour. ELMFIRE can be used, and has already been used, to understand the physics underlying the formation and development of plasma turbulence and its undesirable effects on plasma confinement.

Of particular importance in the study of plasma confinement is the transition between L and H states, referring to Low and High confinement states. Understanding the way to induce an L-H transition in a plasma would lead to an immediate increase of plasma confinement applicable to different plasma devices. The DECI resources obtained for this year have supported ELMFIRE simulations with more memory-efficient code version (extending the calculations into the SOL region) for longer (and heavier) simulations to collisional time scale. Further understanding of pedestal transport and its control has been acquired. The diagnostics of turbulent structures has been improved by correlation studies done for Textor parameters. Confinement transitions have been seen in Tuman-3 configuration by letting the plasma to be heated in the simulations. These findings and learning are proposed to be used for searching for the confinement transition in the Asdex Upgrade edge with ELMFIRE and supported by its sister code ASCOT.


Project Title: Semi-Leptonic Decays in Lattice QCD
Project Leader: Dr. Dirk Pleiter – DESY-Zeuthen, Germany
Resource Awarded: 4,100,000 on EPCC – HECToR QC2 and EPCC – HECToR XT6

Collaborators: Dr. James Zanotti – University of Edinburgh, School of Physics, UK
Using the latest state-of-the-art 2+1 flavour lattice QCD simulations, at light quark masses down to the physical masses, we determine the form factors for the semi-leptonic decays of the K meson and strange octet baryons. Combining these form factors with experimental data we are able to accurately determine the CKM matrix element V_us, an essential parameter in constraining the Standard Model of particle physics.


Project Title: Scaling in supersymmetric Yang-Mills theory
Project Leader: Prof.Dr. Gernot Münster – Westfälische Wilhelms-Universität, Germany
Resource Awarded: 3,040,000 on FZJ – JUGENE, FZJ – JuRoPA and IDRIS – BABEL

Collaborators: Prof.Dr. István Montvay – DESY-Zeuthen, Germany
Dr. Enno E. Scholz – Universität Regensburg, Regensburg, Germany
In recent years supersymmetric theories have aroused increasing interest in elementary particle physics. The supersymmetric extension of the Standard Model with N=1 supercharge is considered to be an interesting candidate for a quantum field theory with phenomenological relevance in the near future. Supersymmetry (SUSY) is an essential ingredient also for other models beyond the Standard Model. In recent years we have investigated the N=1 supersymmetric Yang-Mills theory by means of Monte Carlo calculations on a space-time lattice in order to study the spectrum of low-lying particles, the supersymmetric Ward identities and relevant physical observables. If supersymmetry is unbroken on a non-perturbative level, the particle states are expected to form degenerate supermultiplets in the continuum limit. Our previous results show that part of the low-lying particle states appear to become degenerate near the massless gluino limit, but complete supermultiplets cannot yet be identified. In order to find out whether this is an artifact of the lattice discretisation, we intend to perform numerical simulations on lattices with different lattice spacings and to study the corresponding scaling behaviour of the mass spectrum.


Project Title: Simulations of QCD on ultrafine grids
Project Leader: Prof Christine Davies – University of Glasgow, Dept of Physics & Astronomy, Glasgow, UK
Resource Awarded: 2,013,000 on CSC – Louhi XT, EPCC – HECToR XT6 and IDRIS – BABEL

Collaborators: Dr Eduardo Follana – Universidad de Zaragoza, Departamento de Fisica Teorica, Zaragoza, Spain
Dr Craig McNeile – Universität Wuppertal, Theor. Teilchenphysik, Germany
Fully realistic numerical calculations of the effects of the strong force are now possible using the techniques of lattice QCD. This method enables us to simulate theinteractions between subatomic particles called quarks and gluons at distance scales of a tenth of a femtometre and, from this, build up an accurate picture of the quark bound states called hadrons. Hadrons can be studied experimentally whereas freequarks are never seen, so the linkage between the two must be made by theory. In this way lattice QCD is becoming established as a precision tool for strong interactionphysics and for testing the Standard Model against experiment. Bottom quarks, being much heavier than the light up, down and strange quarks are inprinciple sensitive to much shorter distance scales if they are to be treated fully relativistically. Very fine lattices offer a huge numerical challenge but tests against experiment for hadrons made of these quarks are particularly important, becausethey may offer a window into Beyond the Standard Model physics. Here we propose to perform lattice QCD calculations on ultrafine lattices with a spacing of three hundredths of a femtometre and including the effect of sea up, down and strange quarks. This will allow analysis of fully relativistic bottom quarks for the first time and comparison both to experiment and to effective theories for heavy quarks that have been used up to now. It will open the way for more accurate calculations for hadrons made of bottom quarks in the future.


Share: Share on LinkedInTweet about this on TwitterShare on FacebookShare on Google+Email this to someone