Projects

Results of the 14th Call for Proposals for Project Access.

Projects from the following research areas:

Biochemistry, Bioinformatics and Life sciences (8)

iHART – Characterization of genetic risk variants in ASD families using a reference-free approach

Project Title: iHART – Characterization of genetic risk variants in ASD families using a reference-free approach
Project Leader: Daniel Geschwind, USA
Resource Awarded: 5.5 million core hours on MareNostrum

Details

Team Members :
Barcelona Supercomputing Center – ES
Stanford University – USA
University of California Los Angeles (UCLA) – USA

Abstract
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental disorder that affects 1% of the population. In recent years, the analysis of next generation sequencing data has advanced our understanding of the genetics of the disease, which is essential in order to develop effective treatments. However, given its large heterogeneity, most of the genes involved in ASD remain still unknown and the development of large-scale studies is strongly required. In order to overcome this challenge, the Hartwell Foundation has launched one of the largest worldwide initiative to characterize the genetic basis of ASD: the Autism Research and Technology Initiative (iHART), which has already generated open-source whole genome sequencing (WGS) data for more than 4,000 ASD individuals and their family members and will generate WGS data for additional ASD families by early next year (this time focusing on children of self-declared African American ancestry). In order to get a more accurate detection of de novo and inherited genetic ASD risk variants in these families, we will apply a novel strategy by using a reference-free approach, SMuFin, to directly compare the genome between all affected and unaffected children and their corresponding parents. Considering both, the use of a large-scale whole genome sequencing data set and the effectiveness of the method proposed for detecting genetic variants in ASD families, this PRACE project aims to characterize disease-associated genetic variants, identify novel risk genes and gain comprehension of the biological pathways affected in ASD, including populations that have been under-represented in previous genetic research studies.

INFLUM2(S31N) – Mechanism of drug binding to the S31N mutated M2 proton channel of influenza virus

Project Title: INFLUM2(S31N) – Mechanism of drug binding to the S31N mutated M2 proton channel of influenza virus
Project Leader: F. Javier Luque, ES
Resource Awarded: 15 million core hours on MareNostrum

Details

Team Members :
University of Barcelona – ES
University of Florida – USA

Abstract
The influenza A M2 channel is an extremely selective proton channel expressed on the viral membrane and regulated by pH changes in the interior of endosomes. The M2 channel is a homotetrameric protein that contains an integral transmembrane four-helix channel with 97 residues per subunit, each of which comprises an intracellular C-terminal domain (residues 45-97), a transmembrane domain (residues 25-44), and an extracellular N-terminal domain (residues 1-24). The M2 channel is involved in the viral pathogenicity, as it functions as a proton channel and mediates the transfer of protons to the interior of the viral capside. Because of its crucial role in viral life cycle, it is an attractive target against influenza A virus. Amantadine (Amt) and rimantadine were introduced in the 1960s and target the M2 proton channel. However, the Centers for Disease Control and Prevention has advised against its continued use due to the high prevalence of virus expressing drug-resistance phenotypes. This dramatically impacted the therapeutic resources available to treat flu infections, which are estimated to cause around 40000 deaths every year in the US, and nowadays our therapeutic arsenal against influenza infections is only limited to inhibition of neuraminidase. Amt-resistant influenza strains bear mutations in the M2 channel protein. Although several mutations appear to be viable in vitro, a recent analysis of 31251 M2 protein sequences revealed that the vast majority of resistant viruses bear the S31N mutation. However, this mutation has proved to be very challenging for the design of novel inhibitors. In this project, we plan to use standard molecular dynamics simulations and enhanced sampling techniques to determine the mechanism of ligand binding/unbinding to the S31N variant of the M2 channel. The studies will be focused on two major, but mutually dependent goals: i) to unveil the mechanism of drug inhibition in light of the results reported recently for the mechanism of Amt binding to the wild type M2 channel and its V27A variant, and ii) to elucidate the influence of the ligand on the protonation state of the His37 tetrad.. To this end, the project will combine atomistic molecular dynamics simulations, exploratory conventional metadynamics simulations, subsequently used as starting point for multiple-walker well-tempered metadynamics calculations, which will be supplemented with constant pH replica exchange simulations. We hope that this information will provide valuable guidelines to assist the design of novel size-expanded and size-contracted amantadine-like compounds with improved binding toward the mutated forms of the M2 channel.

“DNA VAULT” – simulation of a DNA origami nanostructure able to control single molecule enzymatic activity

Project Title: “DNA VAULT” – simulation of a DNA origami nanostructure able to control single molecule enzymatic activity
Project Leader: Matia Falconi, IT
Resource Awarded: 15 million core hours on Marconi – KNL

Details

Team Members :
Interdisciplinary Nanoscience Center – INANO-MBG, iNANO-huset – DK
University of Rome “Tor Vergata” – IT

Abstract
Aim of this project is to characterize a DNA origami structure that will help to control the activity of enzymes. This aspect is of great interest in several application-oriented fields, such as biotechnology, nanomedicine and synthetic biology. Combining the latest design principles in DNA nanotechnology we want to simulate a dynamic DNA origami container, named “DNA Vault” (DV), that is able to control single-enzyme activities through compartmentalization. Inspired by the compartmentalization mechanisms observed in living cells, the Andersen group designed and built a dynamic 3D DNA origami structure able to house and fully enclose an enzyme inside its inner cavity, shielding it from its substrate that is free in solution. Furthermore, the DNA nanostructure should respond to specific external stimuli that make it open to expose its protein cargo to the external environment, thus triggering the enzymatic reaction. This structure has been called “DNA Vault” because it was constructed to have solid-walls and a multi-lock system to protect the precious biological cargo. The DV features reversible opening and closing to allow enzyme encapsulation under native conditions. Different strategies for loading of a range of cargo molecules have also been tested, showing that the enclosed cargos are actively shielded from interactions with external molecules. Finally, the activity of both encapsulated and free-diffusing enzymes have been successfully modulated by inducing structural reconfiguration in the DNA nanostructure. For this proof-of-principle study, proteases and endonucleases have been employed and is expected that the generality of this approach will allow the control of a much wider range of enzymes. The purpose of this project is to use molecular dynamics simulations to identify the structure/function relationship of the DV in order to improve its ability to incorporate and shield the enzymes contained within and to refine its controlled opening/closing mechanism. The whole closed DV origami, filled with an enzyme or empty, will be simulated through classical MD altering the ionic strength or modifying its nucleotide sequence composition, in order to better understand to which extent the internal cavity is shielded from the external environment. MD simulations of the DV will give us deeper insights about the stability and flexibility of the overall DNA origami structure, leading to the identification of possible weak-points along the nanocontainer that could be corrected through design optimization. Moreover, simulation of the same systems including enzyme bovine alpha-chymotrypsin, anchored and closed in the DV cavity, will permit us to observe any changes in which the protein can undergo, being confined in the narrow internal DV space. The results obtained by simulating the DV system will provide us the basis for the modification and rational optimization of succeeding nanostructures.

A computational nose: decoding odor perception by all-atom molecular simulations of human olfactory receptors

Project Title: A computational nose: decoding odor perception by all-atom molecular simulations of human olfactory receptors
Project Leader: Jerome Golebiowski, FR
Resource Awarded: 15 million core hours on MareNostrum

Details

Team Members :
University of Nice Sophia Antipolis – FR

Abstract
Our sense of smell strongly influences our emotion, memory and behavior. But what happens from an odorant molecule to an odor? Odorants activate olfactory receptors (ORs) in the nose, and trigger a complex signal transduction to the brain where the signal is perceived as an odor. The human genome expresses roughly 400 types of ORs. Each OR can recognize multiple odorants, while one odorant can activate multiple types of ORs. An odor is the result of combined activation signals of multiple ORs. The variability and combinatorial activation of ORs endow us with a spectacular discriminatory power of smell sensations. Subtle changes in the molecular structure of an odorant can drastically alter its odor. Therefore, a fundamental question in odor perception is how odorants bind and activate their ORs at molecular level. ORs belong to the G protein-coupled receptor (GPCR) superfamily and comprise half of the human GPCRs. GPCRs are the largest group of cell membrane proteins and the most heavily investigated drug targets in the pharmaceutical industry. In the last few years X-ray crystallography and molecular modeling have significantly advanced our understanding in GPCR molecular structure and activation. Built on this, molecular modeling also has become a powerful tool for studying OR-odorant interactions and OR activation. In this project, we start with molecular modeling of mutated GPCRs that are highly active. These mutants provide ideal models for studying GPCR activation with reduced computing costs and complication. We will obtain the common structural and dynamic features associated with GPCR activation. These will then be projected on the study of human ORs and their activation by odorants. Our results will be assessed systematically by in vitro and ex vivo data provided by our collaborators. The study will provide insights into the molecular mechanism of GPCR activation in general, as well as OR-odorant recognition and OR-specific activation features in particular. In a longer term, the project will lead to a computational platform for design and screening of novel odorants, odor blockers and odor enhancers. These molecules will be evaluated in vitro and ex vivo, and on human panels. The project outcome will serve the fragrance industry and benefit the GPCR community and pharmaceutical research . The ultimate goal in the near future is to develop a ‘computational nose’ that can predict the odor of an odorant using its molecular structure as input.

New approaches in radiotherapy: hadron minibeam radiation therapy

Project Title: New approaches in radiotherapy: hadron minibeam radiation therapy
Project Leader: Yolanda Prezado, FR
Resource Awarded: 16 million core hours on MareNostrum

Details

Team Members :
CNRS – FR

Abstract
Radiotherapy (RT) is one of the most frequently used methods for cancer treatment (above 50% of patients will receive RT). Despite remarkable advancements, the dose tolerances of normal tissues continue being the main limitation in RT. Finding novel approaches that allow increasing normal tissue resistance is of utmost importance. This would make it possible to escalate tumour dose, resulting in an improvement in cure rate. With this aim, we have proposed a new approach, called proton minibeam radiation therapy (PROTONMBRT), which combines the prominent advantages of protons for RT and the remarkable tissue preservation provided by the use of submillimetric field sizes and a spatial fractionation of the dose, as in minibeam radiation therapy (MBRT). The main objectives of this project are to optimise the proton minibeam generation and to develop calculation tools allowing to proceed to clinical trials that can open the door to an efficient treatment of very radioresistant tumours, like gliomas. In addition, it can specially benefit paediatric oncology (brain and central nervous system). To be able to carry out all this research an intense work in simulation is needed. The small field sizes used require important calculation resources. Only large cluster, like the ones in PRACE would allow us to perform these calculations.

LACEHIP, LArge scale CEllular model of the HIPpocampus

Project Title: LACEHIP, LArge scale CEllular model of the HIPpocampus
Project Leader: Michele Migliore, IT
Resource Awarded: 21 million core hours on Juqueen

Details

Team Members :
EPFL – CH
National Research Council – IT
Yale University – USA

Abstract
The proposed research focuses on the development the first detailed and realistic large scale 3D model of the CA1 region of the hippocampus. The hippocampus is a small brain region located deep in the brain, in the medial temporal lobe, underneath the cortical surface. Its structure is divided into two halves which lie in the left and right sides of the brain. The organ is curved with a shape that resembles a seahorse, explaining its name. It is well known that the processes related to higher brain function, such as memory, learning, and spatial navigation involve this region, which is one of the most studied both experimentally and theoretically. This project was prompted by a critical issue in the experimental investigation of the mechanisms underlying learning and memory. A major problem in interpreting the experimental findings in vivo is that they are usually obtained in single cells or in small (more or less randomly) selected sets of cells. However, a clear understanding of fundamental processes, such as the spatio-temporal organization of the hippocampal network, requires simultaneous recording from a relevant subset of cells activated by specific sensory and cortical inputs. The functional effects of network-wide processes in relation to input patterns, therefore remain relatively unknown and extremely difficult to explore experimentally. Addressing this critical step requires a large scale realistic model maintaining the natural 3D layout of the real system, in which the neurons’ activity can be observed in exactly the same format as the in vivo recordings, with the fundamental advantage of being able to track network, cellular, and synaptic activity at any point of the network, and directly compare the results with experimental data. In this project we will focus on the mechanisms, occurring at the single cell level, that may contribute to the emergence of higher brain functions observed at the behavioral level. From this point of view, the population of pyramidal neurons in the CA1 region of the hippocampus are in a crucial position to participate in the underlying processes, because they represent the main output stage of the hippocampal circuitry. Their morphological and electrophysiological properties suggest that they are exquisitely tailored to modulate the synaptic integration processes occurring in the dendrites and responsible for short-and long-term memory. Starting from this neuronal population, we will implement a structure that can be easily extended in the future with additional network elements or mechanisms. This will allow to obtain specific insight into the problem of understanding how higher brain functions can emerge from the hippocampus and its microcircuits. We expect that this approach will significantly advance of the state of the art in the field, and will help to predict and explain experimental and behavioral data.

CBNR – CereBellar Network Reconstruction

Project Title: CBNR – CereBellar Network Reconstruction
Project Leader: Egidio DAngelo, IT
Resource Awarded: 32 million core hours on Juqueen

Details

Team Members :
Ecole Polytechnique Federale de Lausanne – EPFL – CH
National Research Council – IT
University of Pavia – IT

Abstract
The cerebellum is part of the Central Nervous System (CNS) and plays a critical role in motor learning and control. Recently, the cerebellum involvement in higher functions such as speech, cognition, fear responses and emotions has been also recognized. When the cerebellum does not function properly, either for neuronal loss or neuronal functional abnormalities, the neuropathological consequence is a disease called Ataxia, in which motor control at different levels (including balance, gait, eye movements and speech) as well as cognitive control are compromised. A cerebellar involvement has also been recently observed in widespread neurodegenerative diseases, the Alzheimer and Parkinson disease, which have a huge impact on worldwide population. Until the last decade, the only way to investigate the cerebellum was trough animal models and non-invasive techniques in Humans. Now, that supercomputers allow high performance computation, the cerebellar circuit can also be reconstructed and simulated through realistic network models formed by hundred thousands neurons. The first step is the reconstruction of the biophysical and biochemical properties of each cerebellar neuron to a high degree of detail (realistic modelling), based on available experimental data. This approach allows not just to understand the fundamental physiological properties of the neurons in an unprecedented manner but also to propagate these properties throughout neuronal networks. The realistic modeling approach requires more data for construction and validation as well as more computational power than classical simplified methods, but it eventually delivers high quality predictions about the biological functions of the network, that can thereafter be tested experimentally. The second step is the assembly of these neurons into local microcircuits and large-scale networks able to reproduce, with a high degree of detail, the synaptic connectivity. The final step is the simulation of the network using various representative tasks in order to determine its behaviors. For example, this approach can be used to address functional patterns involved in motor and cognitive control under normal or pathological conditions. This modeling strategy is embraced by the Human Brain Project (HBP), which promotes the definition, construction and use of tools needed for the entire process of reconstruction, simulation and validation of brain neurons and networks. This strategy requires, to be enforced, the support of high-performance supercomputers such as Marconi at CINECA.

FLEXO: The relationship between electrostatic potential and curvature in biomembranes, and its impact on membrane transport

Project Title: FLEXO: The relationship between electrostatic potential and curvature in biomembranes, and its impact on membrane transport
Project Leader: Himanshu Khandelia, DK
Resource Awarded: 18 million core hours on Piz Daint

Details

Team Members :
University of Southern Denmark – DK

Abstract
The goal of this project is to quantify the relationship between transmembrane electrical potential and curvature in cellular membranes, and demonstrate how this, previously ignored “flexo-electric” effect influences key membrane processes such as the activity and conformational transitions of ion-channels. The project breaks new ground in proposing the notion that electrophysiological measurements routinely carried out on ion channels embedded in membranes do not account for the large contribution to the membrane potential which arises from flexoelectricity, and neither do the interpretation of experiments performed on model membrane systems like vesicles and drug delivery vehicles. Cellular membranes operate under a resting electrical potential, and a large number of cellular processes depend on, or are affected by the interplay between curvature and potential. Lipid bilayer membranes are lyotropic liquid crystals, and an electrical potential difference is created across the lipid membrane when it bends. However, this phenomenon, known as the flexo-electric (FE) effect, that describes the relationship between membrane curvature and electric field, has not been investigated in sufficient detail experimentally or theoretically. We will combine computer simulations and experiments to measure the impact of electric field on the curvature of lipid membranes, and vice versa. We will also investigate the relationship between electric field, membrane curvature and the conformational transitions of membrane proteins in general, and tension-gated ion channels and other voltage-sensitive proteins in particular. The main idea is to explore the breadth of the impact of “flexo-electricity” in biological systems of interest, something never done before, and not widely recognized within the realms of modern membrane biology.

top

Chemical Sciences and Materials (21)

CHANCE – CHemical Networks of RNA nucleotides in realistic prebiotiC Environments

Project Title: CHANCE – CHemical Networks of RNA nucleotides in realistic prebiotiC Environments
Project Leader: Antonino Marco Saitta, FR
Resource Awarded: 5 million core hours on MareNostrum

Details

Team Members :
Université Pierre et Marie Curie – FR

Abstract
One of the most relevant questions in the origins of life research is the physical-chemistry behind the ribonucleic acid molecules (RNA). Several experimental assays have put in evidence the great versatility of the RNA molecules since they participate in several biochemical pathways such as gene expression regulation via splicing and translation, the catalysis of chemical reactions like an enzyme, as well as the capability of self-replicating and storage of information which are essential factors for the development and evolution of living forms. In this project we aim to model the reactivity of the RNA monomers, named ribonucleotides (or nucleotides), under two different and very relevant prebiotic scenarios: (i) bulk hydrothermal conditions, and (ii) in presence of a catalytic mineral surface. This proposal is the computational counterpart of a ERC Advanced Grant demanded by the PI, currently under evaluation. Our modeling will be based on ab initio molecular dynamics (AIMD) simulations coupled with enhanced-sampling techniques (Metadynamics, Umbrella Sampling) in explicit water molecules and including a mineral surface. From this approach we will be able to reconstruct the free-energy landscapes, the energy barriers associated to each pathway and more important, we will be able to unveil the chemical mechanism at atomistic details. Moreover, this study will shed light in the understanding of the catalytic role played by the mineral and the reactivity in the hydrothermal vents in the primitive Earth. Additionally we will investigate the breaking down of other nucleotide-related biological/prebiotic metabolites of crucial importance like ATP and ADP to quantitatively estimate thermodynamic and kinetic properties, and study the different features that connect the prebiotic scenario with current biological systems.

STREET-OF-NY — Study of optoelecTRonic propErties of titanium dioxide nanosystEms in The frame OF the maNy body perturbation theorY

Project Title: STREET-OF-NY — Study of optoelecTRonic propErties of titanium dioxide nanosystEms in The frame OF the maNy body perturbation theorY.
Project Leader: Ivan Marri, IT
Resource Awarded: 38 million core hours on Marconi – KNL

Details

Team Members :
CNR – IT
Istituto Nanoscienze – IT
Università Campus Bio-Medico – IT
Università degli studi di Modena e Reggio Emilia – IT

Abstract
The scope of this project is a rigorous determination of electronic and optical properties of realistic TiO2 nanostructures of different size, morphology, crystal phase (rutile and anatase) and passivation. In particular we will: (i) calculate the electronic gap and quantum confinements effects as a function of nanostructures size, shape, passivation and crystal phase, (ii) line up band energies (in particular bands edges) with respect to the vacuum level and calculate the work function, ionization potential and electron affinity for all the considered systems, (iii) for some specific cases, we will study the spatial extension of the excitons and the existence of preferred orientations along which excitons can delocalize or be confined. Effects induced on optical properties by crystal phase, nanostructure morphology and faces orientation, passivation and size will be investigated in detail. Our analysis will be addressed by using solid and rigorous theoretical approaches encompassing the following three steps: (1) ground state calculations at DFT level in order to extract KS single particle wavefunctions and eigenvalues; (2) MBPT according to Hedin’s approach to Green’s functions calculations (namely G0W0). This step is mandatory to correct the ground KS eigenvalues that usually give rise to the electronic and optical theoretical underestimation with respect to experimental data; (3) MBPT level given by BSE for the calculation of excitons. In particular we will consider two kind of systems, TiO2 nanocrystals (NCs) and TiO2 nanowires (NWs) for a total of 16 structures. Following the suggestion of Polleux (J. Polleux et al. Chem. Eur. J. 11, 3541 (2005) ), we will consider truncated bi-piramidal anatase TiO2 NCs formed by 29 and 78 TiO2 units, that contain 87 and 234 atoms, respectively. Starting from these systems we will construct bare NWs that are periodic along the [001] direction and are cut by the (100) and (010) planes. We will consider NWs formed by 9, 16 and 25 TiO2 units in the unit cells in both the rutile and anatase crystal phase and we will verify the trends wrt size of the calculated electronic properties. In the present project NCs and NWs will be passivated in two different ways, thus simulating both extreme acidic and moderate PH synthesis conditions. The realization of this ambitious project requires an intense use of high performance supercomputers; PRACE resources would therefore be extremely beneficial to the realization of the project.

TeraX – Carbon-nanotube eXcitons in the THz frequency range from first-principles many-body perturbation theory

Project Title: TeraX – Carbon-nanotube eXcitons in the THz frequency range from first-principles many-body perturbation theory
Project Leader: Daniele Varsano, IT
Resource Awarded: 58 million core hours on Marconi – KNL

Details

Team Members :
Italian Research Council – IT
Universita di Modena e Reggio Emilia – IT

Abstract
In Carbon nanotubes (CNTs) low-energy electrons are quasi-one-dimensional massless particles. Due to the reduced dimensionality, they exhibit poor dielectric screening and enhanced electron-electron interactions. This results in striking many-body effects, which were described as Luttinger-liquid behavior, Wigner crystallization, gigantic exciton binding energies. While excitations of semiconductor nanotubes were successfully predicted from first principles by many-body perturbation theory (MBPT, in the GW and Bethe-Salpeter schemes), in excellent agreement with linear and nonlinear optical spectroscopies, the ab-initio understanding of transport experiments on metallic tubes is not complete. Accurate low-temperature spectroscopy on ultra-clean and suspended single CNTs, which have become possible only very recently, yield unexpected results when performed on nominally metallic CNTs: narrow energy gaps are found, around 0.1 eV (THz range), and no metallic behavior was ever observed. This TeraX project aims at a coherent understanding of optics and transport by investigating from first principles the excitonic and quasiparticle properties of narrow-gap CNTs. A key issue explicitly addressed by TeraX is the role of dielectric screening in narrow-gap tubes. This is a controversial and yet fundamental problem, as predicted values of the long-wavelength static dielectric function, projected on the relevant lowest conduction / highest valence band, range between infinity and one for vanishing gap. Relevant for recent experiments is also the origin of dark exciton brightening induced by a magnetic field parallel to the tube axis, which widens the gap. Of special interest to TeraX are the lowest interband excitations of narrow-gap CNTs, often neglected because optically inactive in idealized models of nominally metallic tubes. Taking advantage of the new HPC capabilities in Europe and of our very recent efforts in optimizing DFT and MBPT codes, TeraX will produce a systematic assessment of how dielectric screening, transport gap, and excitonic spectrum depend on the tube curvature, chirality, and intensity of the magnetic field. In addition to elucidating fundamental many-body physics, TeraX is relevant for potential applications of CNTs, like generation of coherent THz radiation or next-generation field-effect transistors with widely tunable transport gaps.

MUSIC – MUltiscale SImulations of photoactive molecules in optical Cavities

Project Title: MUSIC – MUltiscale SImulations of photoactive molecules in optical Cavities
Project Leader: Gerrit Groenhof, FI
Resource Awarded: 28.1 million core hours on Curie

Details

Team Members :

Abstract
We want to understand what happens when photoactive molecules interact with confined light inside optical cavities or in the vicinity of surface plasmons. If this interaction is sufficiently strong, new hybrid light-matter states are formed, the so-called polaritons (or plexcitons), that are coherent superpositions (in the quantum mechanical sense) of the molecules and the cavity photon or plasmon. Access to these polaritons may provide a totally new paradigm for controlling chemical reactions. To design cavities or plasmonic systems for achieving such control, a theoretical model is essential, with which the effect of the coupling on the molecular dynamics can be accurately predicted. However, such model is missing and our understanding is mostly based on phenomenological theories that are not easily inverted to design new systems. To overcome the current limitations, we have developed a new method for simulating the dynamics of photoactive molecules coupled with optical cavities and surface plasmons. Running on HPC hardware, our code can predict the effect of strong light-matter coupling on the dynamics of a large number of molecules. We will use our code not only to provide atomistic insights into recent experiments, but also do design new cavities and plasmonic systems for controlling two prototypical photochemical reactions in proteins: excited state proton transfer and photo-isomerization.

F-BIOC: Exploring the flexibility of a biomimetic cubane in solution

Project Title: F-BIOC: Exploring the flexibility of a biomimetic cubane in solution
Project Leader: Sandra Luber, CH 
Resource Awarded: 21 million core hours on Marconi – Broadwell

Details

Team Members :
University of Zurich – CH
MPI for solid state research – DE

Abstract
Solar-light driven water splitting is a very promising solution for the increasing worldwide energy demand. The paradigm for highly efficient water oxidation, which is currently the main bottleneck in artificial water splitting, is nature’s oxygen-evolving complex. We plan to investigate the behaviour of a Co-based cubane, which is currently one of the closest mimics of nature’s oxygen-evolving complex, in aqueous solution under realistic conditions using high-performance ab initio molecular dynamics. In addition, pioneering highly accurate complete active space self-consistent field calculations will be carried out, which will provide highly sought-after insight into the complex electronic structure of the cubane and very valuable knowledge about parameters influencing its water oxidation activity. This will significantly promote understanding of biomimetic water oxidation catalysts as well as methods for computational catalysis and materials science.

PHOTO2 – The catalytic cycle of the water splitting complex Photosystem II

Project Title: PHOTO2 – The catalytic cycle of the water splitting complex Photosystem II
Project Leader: Leonardo Guidoni, IT
Resource Awarded: 16.1 million core hours on Marconi – Broadwell

Details

Team Members :
CNR – IT
Sapienza Università di Roma – IT
Università degli studi dell’Aquila – IT

Abstract
One of the most energetic process catalyzed by biomolecules it is water oxydation. This reaction occurs in photosynthetic organisms, where an integral membrane complex named Photosystem II converts sunlight energy in oxidising potential, accumulating it in a Mn-Ca cluster. The conversion of two water molecules into molecular oxygen is achieved by yusing four oxidizing equivalents, stored one-by-one in a 5 steps catalytic cycle named Kok-Joliot’s cycle (S0-S4). Since more than forty years of investigations, the available data are still not enough to provide a comprehensive understanding of the molecular mechanism of the reaction. In recent years computational approaches are having an important role in the clarification of the catalytic process, as well as in the interpretation of several experimental results. With this project we will extend the mechanistic study so far performed by us on the S2-S3 transition to the crucial step of the O-O bond formation and the O2 evolution. The work will use quantum-classical atomistic modelling of the protein complex Photosystem II. We expect that our data will provide a comprensive interpretation of the catalytic pathway as well as od the available infrared and EPR spectroscopic data.

DDTLES- Large Eddy simulation of the Deflagration to Detonation transition

Project Title: DDTLES- Large Eddy simulation of the Deflagration to Detonation transition
Project Leader: Olivier Vermorel, FR
Resource Awarded: 10 million core hours on Curie

Details

Team Members :

Abstract
Mining, process and Energy industries suffer billions of dollars of worldwide losses every year due to gas explosions. In addition to the direct damages and the secondary costs of such disasters, explosion accidents are often tragic and lead to a high number of severe injuries and fatalities. Sound knowledge of explosion physics is of vital importance for the prediction of explosion consequences and for providing guidelines for preventive measures implementation. Gas explosions are a multi-physics and multi-scale problem involving chemical reactions, species and heat transport, turbulence, thermodynamic instabilities, acoustics, etc. Upon ignition, a combustion wave propagates away from the ignition source. Generally speaking, there are two types of self-propagating combustion waves: deflagrations and detonations. Detonations are known to be far more destructive than deflagrations because they lead to a higher overpressure peak due to the coupling between shock waves and reaction zone. Ignition in an explosive mixture usually does not directly initiate a detonation. Instead a flame acceleration (FA) process is required to reach fast deflagration regime as a precursor for potential onset of detonation. Although FA and Detonation propagation are quite well understood, the exact conditions that must be satisfied before the occurrence of the deflagration to detonation transition (DDT) are still to be clarified. Experimental studies have shown a wide variety of situations that can lead to the onset of detonation. However, experiments do not give access to the details of the DDT process. CFD is then the ideal tool to assess the underlying mechanisms of DDT. This project aims to perform simulations of Deflagration to Detonation Transition (DDT) in a closed channel with obstacles. The target configuration is a 5.4m long enclosed and obstructed chamber filled with premixed H2/air mixture at atmospheric conditions. Obstacles enhance FA process by increasing the flame surface, and therefore the flame speed. Depending on the operating conditions, different explosion scenarios are observed: 1) For very lean and very rich mixtures only FA was observed. The composition is driving the maximum deflagration speed and the maximum overpressure recorded in the channel; 2) Inbetween experiments show the formation of a detonation front propagating toward the end plate. The relevance of such configurations for the industry is evident: the obstacles mimic the omnipresence of congested rooms and chains of connected rooms in industrial facilities, and the dimensions of the channel is close to realistic configurations. The objective is twofold. On one hand, this proposal is a key to understand the underlying physics of DDT by investigating the influence of chemistry modelling on DDT simulations scenarios and the effect of turbulent chemistry modelling on the results accuracy. On the other hand it aims at proving the feasibility of a 3D simulation of DDT in a large-scale chamber with LES.

Two-dimensional inorganic materials under electron beam: insights from advanced first-principles calculations

Project Title: Two-dimensional inorganic materials under electron beam: insights from advanced first-principles calculations
Project Leader: Arkady  Krasheninnikov, FI
Resource Awarded: 16 million core hours on Hazel Hen

Details

Team Members :
Aalto University – FI
CSC – IT
Center for Science – FI

Abstract
Understanding the interaction of energetic electrons with matter is the key to tailoring materials structure and properties by, e.g., controllable introduction of defects and to the assessment of irradiation damage in the materials in radiation-harsh environments such as cosmic space, or in a laboratory – the column of a transmission electron microscope. This is particularly relevant to two-dimensional (2D) materials, which have recently been in the focus of research due to their unique properties and potential applications. However, a growing body of experimental facts indicate that many concepts of energetic particle-solid interaction are not applicable to these systems, as the conventional approaches based on averaging over many scattering events do not work due to their very geometry, e.g., in graphene–a membrane just one atom thick–or require substantial modifications. In the proposed research project, irradiation effects in 2D inorganic materials will be studied with the main focus on transition metal dichalcogenides. New computational atomistic non-adiabatic techniques will be developed, which will enable studying radiation effects from first-principles with account for electronic excitations. This novel approach will be implemented in open-access computer software, and be used to gain fundamental microscopic knowledge of the interaction of energetic electrons with 2D systems. Among other issues which still are not understood, our results will explain in particular the development of irradiation damage in inorganic 2D materials under electron irradiation. Close collaboration with several leading experimental groups will enable an immediate verification and implementation of the theoretical concepts in state-of-the-art experiments opening thus new avenues for irradiation-mediated post-synthesis tailoring of the properties of 2D materials.

UNWRAP – UNderstanding Water induced degRAdation of hybrid Perovskites

Project Title: UNWRAP – UNderstanding Water induced degRAdation of hybrid Perovskites
Project Leader: Alessandro Mattoni, IT
Resource Awarded: 53 million core hours on Marconi – KNL

Details

Team Members :
Consiglio Nazionale delle Ricerche – IT
Università di Cagliari – IT
Università di Perugia – IT
University of Tokyo – JP
Colorado School of Mines – USA

Abstract
UNWRAP aims at addressing the degradation effect of humidity and water wetting on hybrid organic/inorganic halide perovskites for solar cells applications, and devising, by a comprehensive understanding of atomic scale phenomena, novel strategies to improve the material resistance to water exposure so pushing forward a reliable technology based on hybrid perovskites. The project activity will be based on a combination of classical and ab initio molecular dynamics as well as rare events string methods, DFT and beyond-DFT ab initio electronic structure calculations.

CHANCE – CHemical Networks of RNA nucleotides in realistic prebiotiC Environments

Project Title: CHANCE – CHemical Networks of RNA nucleotides in realistic prebiotiC Environments
Project Leader: Antonio Marco Saitta, FR
Resource Awarded: 5 million core hours on MareNostrum

Details

Team Members :
Université Pierre et Marie Curie – FR

Abstract
One of the most relevant questions in the origins of life research is the physical-chemistry behind the ribonucleic acid molecules (RNA). Several experimental assays have put in evidence the great versatility of the RNA molecules since they participate in several biochemical pathways such as gene expression regulation via splicing and translation, the catalysis of chemical reactions like an enzyme, as well as the capability of self-replicating and storage of information which are essential factors for the development and evolution of living forms. In this project we aim to model the reactivity of the RNA monomers, named ribonucleotides (or nucleotides), under two different and very relevant prebiotic scenarios: (i) bulk hydrothermal conditions, and (ii) in presence of a catalytic mineral surface. This proposal is the computational counterpart of a ERC Advanced Grant demanded by the PI, currently under evaluation. Our modeling will be based on ab initio molecular dynamics (AIMD) simulations coupled with enhanced-sampling techniques (Metadynamics, Umbrella Sampling) in explicit water molecules and including a mineral surface. From this approach we will be able to reconstruct the free-energy landscapes, the energy barriers associated to each pathway and more important, we will be able to unveil the chemical mechanism at atomistic details. Moreover, this study will shed light in the understanding of the catalytic role played by the mineral and the reactivity in the hydrothermal vents in the primitive Earth. Additionally we will investigate the breaking down of other nucleotide-related biological/prebiotic metabolites of crucial importance like ATP and ADP to quantitatively estimate thermodynamic and kinetic properties, and study the different features that connect the prebiotic scenario with current biological systems.

NORMA – Novel synthesis Routes for carbon based 1D nanoMaterials

Project Title: NORMA – Novel synthesis Routes for carbon based 1D nanoMaterials
Project Leader: Carlo Antonio Pignedoli, CH
Resource Awarded: 41.8 million core hours on Marconi – KNL

Details

Team Members :
University of Zurich – CH

Abstract
In the search of new low dimensionality nanomaterials, bottom-up fabrication approaches emerge from a need for atomic precision that cannot be satisfied by the traditional top-down ones. Graphene nanoribbons (GNRs) recently attracted considerable attention as a possible solution to introduce a sizeable band gap into the electronic structure of the semimetal graphene, needed to grant sufficient on-off ratios in room-temperature digital logic applications. In 2010 a bottom-up approach for the synthesis of GNRs was developed in collaboration between the nanotech@surfaces laboratory at Empa, and the department for synthetic chemistry at the Max-Planck Institute for Polymer Research. The starting point is a molecular precursor that is designed specifically to yield a particular GNR. The molecular precursor is then deposited onto a noble metal substrate under UHV conditions. Upon annealing at a characteristic temperature T1 the molecules undergo dehalogenation and the radical intermediates start diffusing across the surface. When the radicals meet, they self-assemble into flexible polymer chains via aryl-aryl coupling, similar to the classical Ullmann reaction. Finally, annealing at a temperature T2 > T1 activates the cyclodehydrogenation reaction, which transforms the polymers into planar GNRs. Combining experimental approaches and HPC methods we were successful in investigating the mechanism of the synthesis process and the electronic properties of fabricated nanomaterials. Now, on one side we need to investigate the electronic properties of 1D heterojunctions that could be fabricated with bottom-up approaches. Following our previous results, we plan to characterize the electronic properties of heterojunctions in the search of candidates for photovoltaic applications. This objective, that will rely on GW calculations performed with the CP2K code on realistic heterojunction models, is embedded in a wider project where we are screening at the DFT level of theory heterojunction candidates (staggering of the gap) for photovoltaic applications. On the other side we will investigate possible routes for on surface synthesis of acene based nanostructures. Pentacene and larger acenes, are easily oxidized in air and are not soluble in the most common organic solvents. Despite their electronic and optical properties are appealing for organic semiconductor applications (pentacene is a routinely used p-type semiconductor) the solubility limits of larger acenes and their instability make them unsuitable for commercial applications. The introduction of solubilizing groups, removable after deposition, is a possible approach to provide acenes with solution processability and high device performance. The commonly available thermo-convertible precursors are not suitable for metal supported bottom up fabrication of nanostructures since thermal deposition would decompose the precursors. However, the availability of molecular precursors that can be converted by photolysis and are more thermally stable, offers the possibility to follow surface synthesis routes to fabricate acene based nanostructures. The questions we want to reply are: i) Which molecular precursors can be deposited on a metallic substrate such as Au(111)? ii) Which molecular precursors do not inhibit on surface polymerization due to steric hindrance? iii) Can we thermally activate polymer decomposition to obtain predesigned acene based nanostructures?

STIDS – STrongly Interacting Disordered Systems

Project Title: STIDS – STrongly Interacting Disordered Systems
Project Leader: Fabien Alet, FR
Resource Awarded: 20 million core hours on Hazel Hen

Details

Team Members :
Technische Universität München – DE

Abstract
We propose to study the dynamical and thermalization properties in complex isolated quantum systems in the presence of disorder, strong correlations and driving. These systems exhibit a rich phenomenology of thermal phases with diffusive or slow dynamics as well as localization phenomena and the breakdown of thermodynamic descriptions, often referred to as many-body localization. To understand these new phenomena that question our fundamental understanding of quantum statistical mechanics, we will employ state of the art technology on Tier-0 systems in order to study dynamical and statistical properties of lattice models using shift-invert exact diagonalization and Krylov space exact time evolution approaches. These numerical techniques are ideally suited for this problem since they do not suffer from the large quantum entanglement which appears in these systems. Reaching very large system sizes and times is crucial to understand the asymptotic transport properties, therefore calling for calculations on Tier-0 systems. Our numerical simulations will not only answer some of the fundamental questions in the field, but will also make connection with very recent cold-atom experiments designed to probe this physics.

MORPHO – MOdelling Radiation damage: characterization of elementary PHysical prOcesses

Project Title: MORPHO – MOdelling Radiation damage: characterization of elementary PHysical prOcesses
Project Leader: Christophe Domain, FR
Resource Awarded: 31.9 million core hours on Marconi – KNL

Details

Team Members :
CEA – FR
EDF R&D – FR
Université Lille 1 – FR
KTH Royal Institute of Technology – SE

Abstract
Under extreme conditions (e.g. irradiation), the mechanical properties of structural materials evolve and the natural ageing of materials can be altered and/or accelerated significantly. The evolution of the microstructure is governed by the formation of defects and the evolution of defect clusters and their interaction with the solute atoms in the material. Microstructure Modelling predictions performed at the mesoscale are strongly dependent on these atomistic phenomena. A multi-scale modeling scheme is developed to simulate the ageing of the materials in order to better predict their behavior of structural materials under irradiation in nuclear power plants. The objective of this project is to contribute to the multi-scale modeling of ageing of engineering materials used for some of the most important components of nuclear power plants. The objective of this project is more precisely to better characterize these the two ageing mechanisms mentioned above using state of the art atomic calculations (Density Functional Theory – DFT method) and requiring very large simulations on HPC machines. The project will focus on the threshold displacement energy (TDE) (i.e. the minimal energy required to create a stable Frenkel pair defect: a vacancy and a self-interstitial) determination in nickel and zirconium as function of crystallographic directions, and on the diffusion properties of complex defect – solute clusters (e.g solute and defect migration energies in loop – solute complexes in Fe and Zr). In particular in Fe, the C15 Laves phase based interstitial clusters are more stable than interstitial loops up to nanometer size, their a priori different properties as well as transformation mechanisms between C15 and loops certainly play an important role in the prediction of the microstructure than need to be quantified with the state of the art atomic simulation methods proposed. The VASP code, one of the state of the art DFT codes for accurate calculations in metals with good scalability on HPC machines, will be used. For TDE calculations ab initio molecular dynamics on around 600-atom supercells will be performed. For mobility properties of defect – solute clusters, migration energies of defects and solutes on 1000- to 1500-atom supercells will be performed. The recent published works (on TDE and C15 – loops for example) demonstrate the feasibility and the maturity of the methods used and the necessity of large HPC machines. These new calculations will allow an extended characterization of defect formation properties and defect-solutes complexes, and will be used in microstructure multiscale modelling to improve the quality of the microstructure prediction. More precisely, TDE will be used for the assessment of empirical potentials for defect productions, and solute-defect cluster properties will be used as input parameters for mobility and stability of clusters in kinetic Monte Carlo and rate theory modelling of the microstructure. This work will contribute to increase the precision and robustness of the computational models used in the multi-scale modelling of the structural material with the challenge of bridging the gap between different length and time scales.

HOPS – High-performance Organic Photovoltaics from many-body perturbation theory Simulations

Project Title: HOPS – High-performance Organic Photovoltaics from many-body perturbation theory Simulations
Project Leader: Gian-Marco Rignanese, BE
Resource Awarded: 20 million core hours on Marconi – KNL

Details

Team Members :
CEA – FR
CNRS – FR

Abstract
Organic photovoltaic (OPV) cells constitute a promising solution for energy harvesting. They are made of earth-abundant materials, and they can be produced in large volume at low cost. Furthermore, they are light-weight and flexible. All these advantages make them ideal for large scale deployment. However, they have quite low efficiencies compared to inorganic (mainly silicon) solar cells which are dominating the market. In the last decade, considerable research efforts have therefore been dedicated to improving their efficiencies both theoretically and experimentally. It is believed that the effeciency strongly depends on the charge transfer (CT) effects taking place at the donor-acceptor (D-A) interface inducing the charge separation of the excitons created upon photon absorption. Though, the actual mechanism is not fully clear. The relative position of the lowest intramolecular donor excitation (S1 state) and the CT excitations at the D-A interface is believed to play a major role. Understanding the structural and electronic factors that control the excitation from the photoinduced intramolecular excitons to a bound CT state on the D-A interface is thus critical to improve quantum efficiencies in solar cells. To this end, first-principles calculations (which do not require any empirical parameters) constitute a choice tools. Tens of thousands of molecules have been already tested within the Harvard Clean Energy Project (CEP), leading to a ranking of potential candidates. To obtain such a large amount of data, the CEP relies on the density-functional theory (DFT) calculations corrected through an empirical model to account for the well-known underestimation of the band-gap of DFT. In contrast, many-body perturbation theory (MBPT) constitutes a choice tool for computing the electronic and optical properties of materials. The so-called GW/BSE formalism was shown to correctly reproduce experimental electronic and optical band gap in many extented systems and, more recently, also for a large set of gas phase molecules and D-A complexes. Using the FIESTA code which implements the GW/BSE formalisme, we have already investigated a series of interfaces. These were made of small organic molecules taken at the top of the CEP ranking as a donor and of fullerene as the acceptor. However, the considered D-A complexes did not include several effects such as the screening effects by the surrounding medium, the possible electronic delocalization over several molecules or the influence of the acceptor aggregation size. The present project aims at going one step further towards more realistic systems by studying the electronic and optical properties of larger interfaces. We intend to start from a simple D-A dimer and add one molecule at a time to build up more complex systems. The study of these large interfaces could lead to a better representation of a real interface and to a better comprehension of the mechanisms at play. As an alternative approach, we also would like to use the GW/BSE formalism with different polarizable models very recently implemented in the FIESTA code.

IrrTitZir – Irradiation in Titanium and Zirconium

Project Title: IrrTitZir – Irradiation in Titanium and Zirconium
Project Leader: Emmanuel Clouet, FR
Resource Awarded: 9.9 million core hours on Marconi – KNL

Details

Team Members :
CEA – FR
CNRS – FR

Abstract
This proposal aims at using ab initio calculations to fully characterize the migration of the self-interstitial in Zr and Ti, calculating both the activation energies of the different migration events but also the corresponding attempt frequencies. This will allow us to make a fully ab initio prediction of the self-interstitial diffusion coefficient. Zirconium and titanium are two transition metals belonging to the same column of the periodic table with a hexagonal close-packed (hcp) crystal structure. Under irradiation, point defects, both vacancies and self-interstitials, are created and then diffuse until they become trapped by the different sinks of the system or they cluster. The macroscopic consequences of point defect clustering are a strong hardening and also irradiation growth, both of utmost technological importance. This formation of vacancy and interstitial clusters is controlled by the point-defects diffusion. In particular, it has been proposed that the formation of the dislocation loops responsible for irradiation growth arises from the diffusion anisotropy of the self-interstitial. But no experimental characterisation of interstitial diffusion exists either in Zr or Ti. Recent ab initio calculations have shown that the self-interstitial can adopt different configurations nearly degenerated in energy and the minimum energy pathways between these configurations have already been calculated, at least in Zr. But a full modelling of diffusion necessitates not only the knowledge of activation energies for these different transitions, but also the corresponding attempt frequencies, i.e. the exponential prefactors in the law describing thermal activation of the transition rates. These attempt frequencies are a manifestation of the vibrations at each stable and saddle point configurations. There is no reason to assume that all transitions have the same attempt frequency. We will therefore use transition state theory in the harmonic approximation to calculate these attempt frequencies. We will perform ab initio calculations of the phonon frequencies of the self interstitial for its different stable configurations and for the saddle points of the transitions between these configurations, from which the attempt frequencies can be simply derived. We will thus obtain a fully ab initio modelling of the self-interstitial migration that will be used to calculate the diffusion coefficient, including the variation of its anisotropy with temperature. This kinetic model will be validated thanks to a confrontation to internal friction experiments realized in irradiated Zr and Ti.

Exploration of the full energy landscape of the PI3Ka E545K mutant by enhanced sampling simulations

Project Title: Exploration of the full energy landscape of the PI3Ka E545K mutant by enhanced sampling simulations
Project Leader: Zoe Cournia, GR
Resource Awarded: 15.3 million core hours on MareNostrum

Details

Team Members :
University of Ioannina – GR
University College London – UK

Abstract
The kinase PI3Ka is involved in fundamental cellular processes such as cell proliferation and differentiation. PI3Ka is frequently mutated in human malignancies. One of the most common mutations is located in exon 9 (E545K), where a glutamic acid is replaced by lysine. The E545K mutation results in an amino acid of opposite charge, where the glutamic acid (negative charge) is replaced by lysine (positive charge). It has been recently proposed that in this oncogenic charge-reversal mutation, the interactions of the protein catalytic subunit with the protein regulatory subunit are abrogated, resulting in loss of regulation and constitutive PI3Ka activation, which can lead to oncogenesis. To test the mechanism of protein overactivation, Parallel Tempering Metadynamics MD simulations will be used here to examine the full energy landscape of the WT and mutant proteins. Understanding how the E545K mutation leads to the increased PI3Ka activity will help us design new candidate drugs for cancer patients who carry this mutation. How? The dynamics and structural evolution of this E545K oncogenic protein, as described by our simulations, might reveal possible binding pockets, which will be then exploited in order to design small molecules that will target only the oncogenic mutant protein. Therefore, our simulation results will be used to identify putative allosteric pockets on the cancerous protein and perform computer-aided drug design for the identification of selective E545K inhibitors. These inhibitors will be further validated by in vitro assays with the aim to discover novel candidate cancer drugs.

Direct Numerical Simulation of Bubbly Flows with Interfacial Heat and Mass Transfer

Project Title: Direct Numerical Simulation of Bubbly Flows with Interfacial Heat and Mass Transfer
Project Leader: Assensio Oliva, ES
Resource Awarded: 18 million core hours on MareNostrum

Details

Team Members :
Technical University of Catalonia – ES
Termo Fluids S.L. – ES

Abstract
Bubbly flow with inter-phase heat and mass transfer is a complex phenomenon, which is difficult to understand, predict and model. These flows are significant in both nature and industry, for instance steam generators in thermal power plants, rocket engines, unit operations in chemical engineering such as distillation, absorption, extraction and heterogeneous catalysis. Despite its industrial importance, the current understanding of such flows and their predictive models are far from satisfactory, therefore, the main motivation of present research project is to contribute in the understanding of these type of flows. Due to the rapid development of computational technology, Direct Numerical Simulation (DNS) of gas-liquid multiphase flows by means of interface capturing methods, has advanced to the point that these are able to provide valuable information on the complexities of bubbly flows with interfacial transport phenomena. Thus, in this project different interface capturing methods such as conservative level-set method and coupled volume-of-fluid/level-set method, will be used to perform DNS of bubbly flows where the interplay of thermocapillary, gravity, and inter-phase heat and mass transfer are important. On the other hand, a drawback of interface capturing methods when these are applied on simulation of bubble swarms is the numerical and potentially unphysical coalescence of the fluid particles, therefore, with the aim to circumvent the aforementioned issue, a multiple marker level-set methodology has been developed, which has the ability to avoid the numerical merging of the fluid interfaces, allowing for long time simulations of bubbly flows, including heat transfer effects and bubble collisions. In this context, present research aims to perform detailed DNS of gravity-driven bubbly flows with heat transfer in a vertical cylindrical channel, DNS of the combined effect of gravity and thermocapillary stress on the motion of single bubbles and bubble swarms, and DNS of multi-mode film boiling on plane and cylindrical surfaces. DNS of bubbly flows with heat and mass transfer are still relatively an unexplored field, therefore, it is expected that outcomes from this study may have important implications in the understanding of these type of flows, as well as in the generation of DNS data for the construction of closure models for engineering simulations of the averaged flow field.

MIXER: How stratification, rotation and confinement impact on the turbulent mixing of passive and active particulate matter

Project Title: MIXER: How stratification, rotation and confinement impact on the turbulent mixing of passive and active particulate matter
Project Leader: Alessandra Sabina Lanotte, IT
Resource Awarded: 25 million core hours on MareNostrum

Details

Team Members :
University of Rome “Tor Vergata” – IT
Eindhoven University of Technology – NL

Abstract
The development of small-scale turbulence in geophysical incompressible flows is governed by the competition of various mechanisms, acting with different roles and strength depending on the situation at hand. These factors are for example the vertical shear of the background flow, the background rotation, the buoyancy forces due to the density stratification. A relevant question concerns the transfer of fluctuations among scales, which can take place via instabilities, turbulent motion and waves. This transfer of fluctuations is ultimately responsible for mixing, a key process in the atmosphere and ocean dynamics. ? Here, we aim at studying the role of confinement, stratification and rotation, to provide a comprehensive vision of the roles of the different mechanisms in promoting or hampering mixing of Lagrangian passive and active particles in these complex flow configurations.

Homo- and heterodimerization mechanism of chemokines receptors CCR5 and CXCR4 investigated by Coarse-Grained Metadynamics simulations

Project Title: Homo- and heterodimerization mechanism of chemokines receptors CCR5 and CXCR4 investigated by Coarse-Grained Metadynamics simulations
Project Leader: Vittorio Limongelli, CH
Resource Awarded: 25 million core hours on MareNostrum

Details

Team Members :
University of Lugano USI – CH

Abstract
The development of small-scale turbulence in geophysical incompressible flows is governed by the competition of various mechanisms, acting with different roles and strength depending on the situation at hand. These factors are for example the vertical shear of the background flow, the background rotation, the buoyancy forces due to the density stratification. A relevant question concerns the transfer of fluctuations among scales, which can take place via instabilities, turbulent motion and waves. This transfer of fluctuations is ultimately responsible for mixing, a key process in the atmosphere and ocean dynamics. ? Here, we aim at studying the role of confinement, stratification and rotation, to provide a comprehensive vision of the roles of the different mechanisms in promoting or hampering mixing of Lagrangian passive and active particles in these complex flow configurations.

NANOASSEMBLY-Targeting the polymer assembling in supramolecular nanoparticles for in cell imaging devices

Project Title: NANOASSEMBLY-Targeting the polymer assembling in supramolecular nanoparticles for in cell imaging devices
Project Leader: Adriana Pietropaolo, IT
Resource Awarded: 15.2 million core hours on SuperMUC

Details

Team Members :
University of Miami – USA

Abstract
The goal of the NANOASSEMBLY project is to use a computational approach to facilitate the design and development of polymer nanoparticles with bespoke shapes and photoresponsive properties, for their use in real time cell imaging devices. The class of polymer nanoparticles under investigation is based on amphiphilic polymers composed of a poly(methacrylate) having in the side chains hydrophobic decyl and hydrophilic poly(ethylene glycol) groups. Anthracene and boron dipyrromethene (BODIPY) chromophores are also covalently attached to the macromolecular backbone, in order to incorporate complementary donor and acceptor groups capable of transferring energy within the nanoparticle assemblies upon light stimuli. The amphiphilic character of the class of polymers under examination makes the self-assembling process spontaneous in an aqueous environment. Moreover, hydrophobic molecules which otherwise would be insoluble in water solution can be transported across hydrophilic environments through the nanoparticle scaffold. In addition, the presence of photoactive groups in the polymer main chain permits the opportunity to visualize the nanoparticles and monitor their stability and dynamics with optical methods. For all these reasons, polymer nanoparticles are promising vehicles for the delivery of drugs through the blood stream to target locations in living organisms, opening fascinating prospects in nanomedicine, especially for therapeutic and imaging applications. Specifically, we will predict the structural and dynamic properties of the supramolecular nanoparticles through the reconstruction of their free-energy landscape with a particular focus on the size of the free-energy activation barriers. We will investigate the shapes of regular polygons ranging from a three-chain triangle to a twenty-chain icosagon, disclosing the occurrence of free-energy traps that can perturb the distribution of the nanoparticle shapes in solution. We will use the Parallel Tempering Metadynamics (PTMetaD) approach, an extremely efficient parallel method for the calculation of the free energy differences as a function of one or more collective variables. Furthermore, we will predict how the donor-acceptor distance along the poly(methacrylate) backbone will influence the photophysical properties of the nanoparticle, in terms of their UV absorbance and fluorescence emission. The high performance of this simulation protocol is demonstrated by the previous research we performed with HPC resources (from previous PRACE and national calls). The present calculations will therefore predict the structural and free-energy properties of the supramolecular nanoparticles, tailoring their design with new strategies for a bespoke development of supramolecular probes for in cell imaging.

Mapping the structures and properties of all bulk forming binary systems: a high-throughput study

Project Title: Mapping the structures and properties of all bulk forming binary systems: a high-throughput study
Project Leader: Nikola Marzari, CH
Resource Awarded: 85.2 million core hours on Piz Daint

Details

Team Members :
EPFL – CH

Abstract
Technological development is driven by novel materials and by their properties. To date, however, even for simple materials many properties are still unknown. The goal of this project is to use first-principles calculations, using verified and validated approaches, to determine reference knowledge of fundamental physico-chemical properties, and disseminate the results as an open-access reference database. The unique take of our approach is that this will be done in collaboration with the Pauling File, the most extensive collection of reference experimental data on inorganic crystals, providing thus both extensive data for validation of the results, but in particular allowing to launch a massive search for all current missing structures and stoichiometries for binary compounds. The target list will be based on the Pauling File information, as well as on binary prototype structures and appropriate element substitutions. This effort will require of the order of half a million first-principles calculations, using density-functional theory (DFT) complemented with advanced functionals for the difficult case of transition-metal complexes. Using the code-agnostic automation and workflow environment of AiiDA, we will determine equilibrium structures, electronic and elastic properties as well as vibrational properties. From these, we will derive quasiharmonic free energies to construct finite-temperature convex hulls and phase diagrams, as well as mechanical properties at finite temperature and pressure. When necessary (transition-metal and rare-earth compounds), standard exchange-correlation approximations will be augmented by our recently developed self-consistent Hubbard U+V approach, where all parameters are determined in a highly-automated fashion using density-functional perturbation theory. In addition to providing a reference database for binary compounds this project will also enable verification and validation of the accuracy of DFT codes, functionals and methodologies for the calculations of material properties. In particular, this will be the first large scale project to calculate in high-throughput mode material properties using Quantum-ESPRESSO, (the largest US efforts, such as the Materials Project, OQMD, and AFLOWlib are all based on VASP), providing much needed cross-verification. Furthermore, mechanical properties and vibrational free energy contributions will allow direct comparison to experimental phase diagrams, formation enthalpies and free energies as available through the Pauling File. All results, including the curated data, raw data, and AiiDA directed acyclic graphs will be available via the open-access Materials Cloud portal (materialscloud.org). Last, calculations will use the GPU-enabled SIRIUS domain-specific library (https://github.com/electronic-structure/SIRIUS), currently under active development and already operational, and will thus profit from new hybrid hardware that will become available at CSCS.

ShSwim – Hydrodynamics of flagellated microswimmers near superhydrophobic surfaces

Project Title: ShSwim – Hydrodynamics of flagellated microswimmers near superhydrophobic surfaces
Project Leader: Mauro Chinappi, IT
Resource Awarded: 7 million core hours on SuperMUC

Details

Team Members :
Sapienza, University of Rome – IT
Università di Roma Tor Vergata – IT

Abstract
Evolution had equipped microorganisms with a variety of strategies of locomotion such as gliding, and pseudopodia extension. A quite common strategy is the rotation or the beating of one or more flagella. For instance, Escherichia coli and Pseudomonas aeruginosa, two widely diffused bacteria species that, in specific cases, can have a pathogenic action, employ one ore more rotating flagella to swim across the surrounding liquid. The flagellum rotation exerts an action on the fluid that allows the bacteria to move along a straight line when they are far from interfaces. However, in several relevant conditions, bacteria move in confined geometries, i.e. very close to interfaces. Examples are given by biomedical devices (such as catheters) or thin liquid films (e.g. the layer of liquid that wets the lung epithelium). In addition, bacteria typically accumulate close to the liquid interfaces forming microcolonies and biofilm precursors that can have a dramatic impact of their resistance to antibacterial and chemical disinfectant. Understanding the swimming of the flagellated microorganism hence, not only is an interesting problem in fundamental fluid-dynamics, but can have also an impact on the development of new strategies to control the bacteria motion and, consequently, to reduce the propensity to form biofilms. Moreover, a clear understanding of microswimmer fluid dynamics will also be beneficial for the design of artificial microswimmers that are though to be an enabling technology for several applications, such as targeted drug delivery and water purification. In this project we aim at exploring, by means of hydrodynamics simulations, the possibility to passively control the motion of flagellated microswimmers using properly patterned super-hydrophobic surfaces. Super-hydrophobic surfaces, the most celebrated example being the lotus leaves, are characterized by a rough structure where the air get trapped in the surface asperities. Hence, the liquid is in contact with a complex boundary where the usual solid-liquid (no-slip) boundary is intercalated by air-liquid (perfect-slip) patches. This non-homogeneous boundary condition induces interesting consequences on liquid motion, such as the presence of an effective drag reduction. Our group already studied the motion of flagellated bacteria at homogeneous (solid-liquid and air-liquid) interfaces and the motion of passive tracers at super-hydrophobic surfaces. Based on these results, we expect that the super-hydrophobic structure can potentially control the motion of flagellated microswimmers hindering the formation of the typical circular trajectories that are thought to be a crucial ingredient for the boundary accumulation and consequently, the first stage for microcolonies formation. In this project, we will employ a boundary element solver, named BeSwimmer, we developed in the past years. We already applied BeSwimmer to analyze E.coli motion at both solid-liquid and air-liquid interfaces and we are currently studying the E.coli motion under confinement. In both the cited cases, the homogeneity of the planar surfaces drastically reduces the computational needs, however, the extension to important case of super-hydrophobic surface, call for a computational effort one-two orders of magnitude larger and can be afforded only via Tier-0 resources.

top

Earth System Sciences (4)

Convection-resolving Climate on GPUs (gpuCLIMATE)

Project Title: Convection-resolving Climate on GPUs (gpuCLIMATE)
Project Leader: Christoph Schär, CH
Resource Awarded: 89 million core hours on Piz Daint

Details

Team Members :
ETH Zurich – CH

Abstract
The climate system is intimately coupled with the water cycle, and many uncertainties around climate change depend upon the representation of water vapor, cloud and precipitation processes. The currently ongoing development of high-resolution atmospheric models opens exciting prospects in this area. In particular, a further increase of the horizontal mesh size below a few kilometers will make it feasible to explicitly represent the dynamics of deep convective and thunderstorm clouds, without the help of semi-empirical parameterizations. This development allows reducing some of the key uncertainties in the current generation of climate models, and enables a more adequate representation of extreme events such as heavy precipitation events and thunderstorms. In the current project, we are developing a European-scale climate modeling capability at a horizontal resolution of about 2 km. This resolution is about 10 to 100 times higher than in conventional climate models. From a computer science perspective, this goal poses major challenges. First, the need for increasing compute power requires the use of emerging hardware architectures that includes heterogeneous many-core architectures consisting of both “traditional” central processing units (CPUs) and accelerators (e.g., GPUs). Second, with increasing computational resolution, the model output becomes unbearably voluminous, which requires new approaches to perform the analysis online rather than storing the model output. Using previous support from the Swiss National Science Foundation (project crCLIM) and a major computational grant at the Swiss Center for Scientific Computing (CSCS Lugano), this development is well underway and our group has conducted the first decade-long European-scale climate simulation at km-scale resolution, using a GPU version of our regional climate model COSMO. The underlying project crCLIM is highly interdisciplinary and combines the expertise of climate and computational scientists (see http://www.c2sm.ethz.ch/research/crCLIM.html). The main objectives of the current project are: (i) to increase our understanding of the European climate and its variability, (ii) to provide continental-scale climate-change simulations and thereby to assess future changes of the hydrological cycle and of the associated changes in extreme events, and (iii) to further develop computational strategies for conducting decade-long high-resolution convection-resolving climate simulations on emerging heterogeneous supercomputing platforms. Ultimately, crCLIM will lead to a substantial reduction of some of the key uncertainties in the current generation of climate models, yield an improved representation of the water cycle including the drivers of extreme events (heavy precipitation events, floods, droughts, etc.), and enable more sophisticated climate change scenarios. This, in turn, will provide better guidance for impact assessment and climate change adaptation measures.

High-Resolution Ensemble Sea Ice Reanalyses (HiRes-SIR)

Project Title: High-Resolution Ensemble Sea Ice Reanalyses (HiRes-SIR)
Project Leader: Virginie Guemas, ES
Resource Awarded: 15.6 million core hours on MareNostrum

Details

Team Members :
Barcelona Supercomputing Center – ES

Abstract
The dramatic decline in Arctic sea ice has been an emblematic sign of ongoing global climate change. Profound reductions in sea ice areal coverage and thickness, among others, have already had devastating impacts on local ecosystems, indigenous populations and possibly lower-latitude climate. These rapid changes are also unlocking economic and industrial opportunities. Thinner, younger ice facilitates operations of icebreakers in the High North. Increased marine accessibility promotes polar shipping as an economically viable alternative to existing commercial routes. Ecotourism, resources extraction and industrial fishing are other examples of activities that can take place in an open Arctic Ocean. Further reductions in the sea ice cover are expected in the near future and predicting the first Arctic-free summer is one of the current challenges of the scientific community. A better understanding of the interactions between the long-term externally forced climate trend and the internal variability, which is essential to accurately estimate the amplitude of upcoming sea ice losses, requires long and continuous monitoring of polar climate variables. Unfortunately, before 1973, Arctic sea-ice data are limited to monthly estimates of sea ice extent. The situation is worse in the Antarctic where sea ice data is limited to estimates of extent climatologies over two distinct periods: 1929-1939 and 1947-1962. Filling the critical gaps in sea ice observations is essential for monitoring the long-term evolution of the sea ice system but also to provide a complete description of the sea ice state to initialize seasonal-to-decadal climate predictions. A complete and coherent description of the sea-ice state can only obtained through a physical extrapolation of the sparse observations, relying on the equations that describe the sea-ice dynamics and thermodynamics, via data assimilation techniques. Up until now, this approach has only been applied at a typical resolution of about 1deg. However, resolving mesoscale ocean eddies would allow for more realistic representation of the ice drift and deformation and, consequently, of the Arctic open water percentage. Here, we propose to generate a 25-member ensemble of sea ice reanalyses at 0.25deg, a resolution never achieved up until now in a data assimilation context. Furthermore, the large ensemble size will provide a robust evaluation of the related uncertainty. This reanalysis will use the extended Ensemble Kalman Filter to assimilate sea ice data and will cover the 1978-present period, thus becoming the longest sea ice reanalysis available. This new high-resolution dataset will provide the best estimates of past sea ice covers and offer new opportunities to further our understanding of the various physical processes at play in the polar regions. It will also provide the necessary data with which to initialize the new generation of high-resolution climate models in a seasonal-to-decadal forecasting context.

RegCM atlas for CORDEX phase II

Project Title: RegCM atlas for CORDEX phase II
Project Leader: Erika Coppola, IT
Resource Awarded: 20 million core hours on Marconi – KNL

Details

Team Members :
ICTP – IT

Abstract
The Coordinated Regional Climate Downscaling Experiment (CORDEX) (Giorgi et al. 2009) is moving towards the implementation of its phase 2 (CORDEX2) that is composed of two initiatives. The first is the CORDEX-CORE, by which a minimum set of regional climate models (RCMs) will downscale to a spatial resolution spanning from 12 to 25 km, a common set of Global Climate Model (GCM) scenario simulations over most CORDEX domains. This initiative is also expected to provide a strong contribution to the next reports of the Intergovernmental Panel on Climate Change (IPCC) to be finalized in 2020-21. The second initiative consists of the so called “Flagship Pilot Studies”, or FPSs, i.e. more targeted projects addressing specific scientific challenges and focusing on different sub-regions. One challenge is the development and use of convection permitting (CP) models, which can be run at non-hydrostatic resolutions of 1-5 km. The ICTP is one of the leading institutions in both these initiatives, and plans to participate in them with its RCM system RegCM4 (Giorgi et al. 2012). Specifically, the ICTP group plans to complete scenario projections for three CORDEX domains: the African, South American and pan European ones. In addition, the ICTP plans to participate in a European FPS entitled “Convective phenomena at high resolution over Europe and the Mediterranean” and to use a recently developed non-hydrostatic version of the RegCM4 model for simulations at convective permitting resolutions (3 km or less) over domain covering the greater Alpine region. This FPS is a frontier research project in which coordinated multimodel experiments will allow an in depth assessment of the value of CP models in simulating convective scale processes in mountainous terrain and their response to global warming. The ICTP is one of the co-leaders of the project. Regarding the CORDEX-CORE initiative, for the Africa and South America domains two scenario simulations (RCP 8.5 and 2.6) are foreseen driven by a minimum of 3 GCMs for 140 years at a horizontal grid spacing of 15-25 km. For the pan European domain a set of 7 scenario simulations are foreseen at a horizontal grid spacing of 12 km (same as the EURO-CORDEX standard domain). For the FPS convective permitting simulations a minimum of 40 years of simulation are foreseen, including 10 year time slices for present day, near-term, mid 21st-century and end-of 21st-century. One time slice will cover the near present period from 2000 to 2010 driven by ERA-interim fields to validate the model, while the other time slices will be driven by a GCM. All these RegCM4 simulations will contribute to front-line, innovative RCM ensemble downscaling experiments and will provide a critical contribution to the next generation assessments of impacts and vulnerabilities worldwide. The ICTP group has a long standing experience in producing RegCM-based regional climate projections and it has been very active in CORDEX phase I where it has produced a set of 33 50km scenario simulations for 5 CORDEX domains (Africa, South America, Central America, South Asia, Euro-Mediterranean).

eDUST – High-resolution regional dust reanalysis based on ensemble data assimilation techniques

Project Title: eDUST – High-resolution regional dust reanalysis based on ensemble data assimilation techniques
Project Leader: Sara Basart, ES
Resource Awarded: 21 million core hours on MareNostrum

Details

Team Members :
Barcelona Supercomptuting Center – ES

Abstract
Over the past decade, there has been a growing recognition of the crucial role of sand and dust storms (SDS) on weather, climate and ecosystems, along with their important adverse impacts upon life, health, property, economy. Reacting to the concerns on SDS by its most affected member states, the World Meteorological Organization (WMO) endorsed the launch of the SDS Warning Advisory and Assessment System (SDS-WAS), and more recently of the first Regional Specialized Meteorological Center for Northern Africa, Middle East and Europe with activity specialization on Atmospheric Sand and Dust Forecast. The SDS-WAS mission is to enhance the delivery of timely and quality SDS forecasts, observations, information and knowledge to users through an international partnership of research and operational communities. Understanding, managing and mitigating SDS risks requires fundamental and cross-disciplinary knowledge underpinned by state-of-the-art scientific research, the availability of reliable information on SDS past trends and current conditions, the provision of skilful forecasts and projections tailored to a diversity of users, and the capacity to use the information effectively. At present, all these requirements are confronted by major challenges. These challenges include the lack of reliable dust information in many countries affected by SDS and the very limited integration of dust information and forecasts into practice and policy. A major obstacle to reconstructing comprehensive dust information of the past is the scarcity of historical and routine in-situ dust observations, particularly in the countries most affected by SDS. Model simulations can be used to “fill in the blanks” and overcome the sparse coverage, low temporal resolution and partial information provided by measurements. By objectively combining model simulations with satellite observations, eDUST will produce an advanced decadal high-resolution dust reanalysis for Northern Africa, Middle East and Europe. The proposed dust regional reanalysis will be built on three pillars: a state-of-art dust model and data assimilation system, quality observations and understanding of their respective uncertainties, and flow-dependent uncertainties reflected by the ensemble simulations. So far current reanalysis have been thought for the global domain (missing dust processes associated to finer spatiotemporal scales) and are based on the assimilation of total aerosol optical properties (lacking observational constraints on the model individual aerosol components). The novelty of eDUST will be the generation of a dataset at an unprecedented high-resolution spanning the satellite era of quantitative aerosol observations and the assimilation of satellite products over source regions with specific observational constraints for dust. eDUST aims at making a significant step forward in the way SDS affect society. The reanalysis dataset will become a valuable resource for numerous users to drive or diagnose their models and applications. The wider community will benefit from derived studies on the impact of dust on weather, climate, atmospheric chemistry and ecosystems. The high-resolution dust reanalysis will describe with accuracy the dust variability and trends, and provide extensive information for the socio-economic evaluation of major events, and their short (direct) and long-term (induced) impacts on society. It will also allow the assessment of the efficiency of counter-measures for particular sectors.

top

Engineering (5)

CRITICal – ChaRacterIzing Two-phase Ignition in Combustors

Project Title: CRITICal – ChaRacterIzing Two-phase Ignition in Combustors
Project Leader: Ronan Vicquelin, FR
Resource Awarded: 33.2  million core hours on Curie

Details

Team Members :
CentraleSupélec – FR
CERFACS – FR

Abstract
The objective of the CRITICal project is to carry out large eddy simulation of ignition in realistic combustors accounting for liquid fuel injection. Ignition constitutes indeed a critical phase in aerospace propulsion. It must be completed in a safe and reliable way, even under unfavorable conditions at high altitude, where rapid relight in case of accidental extinction of the combustor is required for engine certification. New combustion technologies currently developed to reduce pollutant emissions, make the ignition process even more critical. Particular attention is then paid to gas turbine ignition characteristics in the earliest stages of design. This relies on the best possible knowledge of the ignition process. Numerical simulation, and in particular the Large Eddy Simulation approach, is a powerful tool to understand, predict and control such an unsteady phenomenon. The ignition phenomenon is particularly complex in the presence of a turbulent flow and all the more so when the injected fuel is in liquid form. This project aims specifically at characterizing two-phase ignition in combustors equipped with a single or multiple injectors in order to cover several phases of the ignition process in gas turbines. This characterization is carried out in terms of first assessing the accuracy of state-of-the-art simulations in such a complex environment and, secondly, understand the impact of liquid fuel injection on gas turbines’ ignitability.  On the one hand, repeated ignition events are computed on a two-phase single burner configuration to study the small flame kernel survival and the successful ignition of a first burner. On the other hand, light-round simulations in a multiple-injectors annular combustor are carried out with different modelling approaches in order to investigate the effects of fuel droplets on the flame propagation while it ignites all the combustor’s burners sequentially. The project brings together two well known teams, EM2C and CERFACS, who have collaborated over many years and are currently successfully completing another PRACE project.

HETS/Heat (and Mass) Transfer in Turbulent Suspensions

Project Title: HETS/Heat (and Mass) Transfer in Turbulent Suspensions
Project Leader: Luca Brandt, SE
Resource Awarded: 18.3 million core hours on Marconi – KNL

Details

Team Members :
Università degli Studi di Padova – IT
TU Delft University of Technology – NL
SINTEF (APPLIED RESEARCH, TECHNOLOGY AND INNOVATION) – NO
KTH Royal Institute of Technology – SE

Abstract
The aim of the project is to perform interface-resolved numerical simulations of heat transfer in turbulent channel flow of droplet suspensions, with and without evaporation. Multiphase flows with heat transfer and phase change are encountered in many applications such as spray dryers, scrubbers and liquid spray combustion. Different transport processes are active in these flows and these interact with each other in a complicated manner, thus creating a challenge in modelling and simplifying the problem. In particular the addition of solid particles or droplet can affect the mixing significantly, an effect which cannot be modelled by available formulations in the literature. Therefore, we aim to use Direct Numerical Simulations (DNS) in order to investigate the details of the heat transfer problem in the presence of rigid/deformable particles/droplets including phase change in a second stage of the project. The work is part of the ERC grant ERC-2013-CoG-616186, TRITOS to Prof Brandt concerning particle suspensions, with a significant extension to include the thermodynamics of phase change. In particular we will start by investigating the effect of the presence of particles on the heat transfer in a turbulent flow, taking the phase change into account in the next step of this study. To tackle this highly relevant and unexplored process, we assume particles that shrink/swell due to evaporation/condensation while maintaining the spherical symmetry. It should be noted that this assumption is justified when the droplets are sufficiently small and the surface tension force larger than inertia and viscous forces (small Capillary and Weber numbers). One of the goals of this study is to validate a novel numerical tool to attack this problem for a wider range of parameters. The Immersed Boundary Method (IBM) will be used in this study to resolve the interface between gas and liquid for velocity, temperature and vapour mass fraction in the domain beside a Lagrangian approach to transport the mass centre of the particles. An indicator function will be used to express the dependency of the fluid material properties on the phase. The code has been parallelized, validated and shown to have an almost linear scaling with the number of processors. The work proposed will be performed in close collaboration between the groups of Prof Brandt at KTH Mechanics, Stockholm, Prof Breugem in TU/Delft, the Netherlands, and Dr. Gruber at SINTEF, Norway, the largest independent research organisation in Scandinavia.

CYCLIC – Large Eddy Simulations of micro-vortex generators for Shock Wave/Turbulent Boundary Layer Interaction

Project Title: CYCLIC – Large Eddy Simulations of micro-vortex generators for Shock Wave/Turbulent Boundary Layer Interaction
Project Leader: Julien Bodar, FR
Resource Awarded: 18 million core hours on Marconi – KNL

Details

Team Members :
ISAE-SUPAERO – FR

Abstract
The shock wave/turbulent boundary layer interaction (SWTBLI) occurs in a wide range of supersonic internal and external flows and has been widely investigated over the last 70 years. The case of an incident oblique shock wave impinging on a flat plate turbulent boundary layer (TBL) is a canonical situation oering a simple framework for understanding the SWTBLI but also frequently encountered in high-speed flows of practical interest. When triggered by a strong SWTBLI at large upstream Mach numbers, a TBL separation bubble appears with a low frequency streamwise motion of the reflected shock position just upstream of the interaction. This leads to fluctuating pressure and thermal loads as undesirable consequences for the application such as engine’s air intake. This unsteady motion of the region including the reflected shock and the separation bubble, exhibits a wide range of frequencies. No consensus about the origin of this low-frequency oscillation has emerged yet, despite the numerous experimental studies that have been conducted to understand and control this unsteadiness. Besides, flow control methods have been proposed to alleviate the SWTBLI-induced impact on performances. Quite recently, microramp vortex generators (MVGs) of size typically smaller than the TBL thickness have drawn a particular interest as they provide the eciency of vortex generators (VGs), in reducing the separation, while minimizing the byproduct induced drag. MVGs are designed to alter the properties of the incoming TBL by introducing vortices in the near-wall region. This research project focus on geometrical characteristics of the MVGs in order to characterise and understand their effectiveness on SWTBLI. This is a first step toward their optimisation.

Direct numerical simulation of partially premixed combustion in internal combustion engine relevant conditions

Project Title: Direct numerical simulation of partially premixed combustion in internal combustion engine relevant conditions
Project Leader: Xue-Song Bai, SE
Resource Awarded: 42.4 million core hours on Juqueen

Details

Team Members :
Lund University – SE

Abstract
In the past decade, the European and world engine industry and research community have spent a great effort in developing clean combustion engines using the concept of fuel-lean mixture and low temperature combustion which offers great potential in reducing NOx (due to low temperature), soot and unburned hydrocarbon (due to excessive air), and meanwhile achieving high engine efficiency. One example is the well-known homogeneous charge compression ignition (HCCI) combustion engine, which operates with excessive air in the cylinder, and produces simultaneously low soot and NOx. However, HCCI combustion is found to be very sensitive to the flow and mixture conditions prior to the onset of auto-ignition. As a result, HCCI engine is rather difficult to control. At high load (with high temperature and high pressure) engine knock may occur with pressure waves in the cylinder interacting with the reaction fronts, leading to excessive noise and even damage on the cylinder and piston surface. At low load (with lower temperature and pressure) high level emissions of CO and unburned hydrocarbons may occur, which lowers the fuel efficiency and pollutes the environment. Recently, it has been demonstrated experimentally that with partially premixed charge compression ignition, also known as partially premixed combustion (PPC), smoother combustion can be achieved by managing the local fuel/air ratio (thereby the ignition delay time) in an overall lean charge. There are several technical barriers in applying the PPC concept to practical engines running with overall fuel-lean mixture, low temperature combustion. For example, it is not known what the optimized partially premixed charge is for a desirable ignition, while at the same time maintaining low emissions. The main difficulty lies in the non-linear behavior of the dominating phenomena and the interaction among them (e.g. chemistry and turbulence). To develop an applicable PPC technology for IC-engine industry, improved understanding of the multiple scale physical and chemical process is necessary. Further, there is a need to develop computational models for simulating the process for the design where a large number of control parameters are to be investigated. The goals of this project are to achieve improved understanding of the physical and chemical processes in overall fuel-lean PPC processes, and to generate reliable database for validating simulation models for analysis of the class of combustion problems. This shall lead to development of new strategies to achieve controllable low temperature combustion IC-engines, while maintaining high efficiency and low levels of emissions (soot, NOx, CO and unburned hydrocarbons). Direct numerical simulation (DNS) approach that employs detailed chemistry and transport properties will be used to study the mechanisms responsible for the onset of auto-ignition, and the structures and dynamics of the reaction front propagation in PPC conditions.

R2Wall Resolved LES to support wall-model development

Project Title: R2Wall Resolved LES to support wall-model development
Project Leader: Koen Hillewaert, BE
Resource Awarded: 30 million core hours on Juqueen

Details

Team Members :
Cenaero – BE

Abstract
The objective consists in performing high resolution Large-Eddy Simulations (LES) and potentially up to Direct Numerical Simulation (DNS) of the NACA 4412 airfoil at maximum lift angle and at a Reynolds number of 1.6 millions. The aim is to provide open-access detailed and high resolution reference data set for the development of lower fidelity CFD methodologies, and in particular (non-equilibrium) wall models for LES. These data will comprise blade pressure and friction distribution, boundary layer profiles of velocity components and temporal correlations as well as spectra, cross boundary layer spatial correlations of velocity and wall stress components. The grid and span independence for each of these quantities will be assessed.Finally, detailed visualisation will be undertaken for a better understanding. All of the post-processed statistics, animations, meshes and flow snapshots will be publicly available after the project. This work inscribes itself in the development roadmap of the solver Argo for LES of wind turbines and jet engines. Currently capable of DNS and LES of transitional flows at moderate Reynolds number (100-500k), Argo is currently being extended to high Reynolds number flows using WMLES. The data set will be used for the calibration of non-equilibrium wall model approaches.

top

Fundamental Constituents of Matter (9)

Non-equilibrium dynamics in strongly interacting and superfluid Fermi systems

Project Title: Non-equilibrium dynamics in strongly interacting and superfluid Fermi systems
Project Leader: Gabriel Wlazlowski, PL
Resource Awarded: 62,9 million core hours on Piz Daint

Details

Team Members :
University of Warsaw – PL
Warsaw University of Technology – PL
Los Alamos National Laboratory – USA
Pacific Northwest National Laboratory – USA

Abstract
You may have heard, that “superfluidity” or “superconductivity” is a fascinating phenomenon. It is a persistent current, where a fluid flows without any loss of energy (or no viscosity). It has attracted scientists to apply it to our daily life via, for instance, efficient electricity transmission or magnetic trains. It has been already 80 years since this phenomenon has been discovered, yet its physics has not been fully explored even now. Especially performing a fully microscopic simulation is not an easy task, since tens to hundreds of thousands, or even more of time-dependent nonlinear coupled 3D partial differential equations have to be solved, and presently only the leadership-class computers like PRACE tier-0 machines can handle this type of calculations. In this project we apply our state-of-the-art computational code, which has been fully developed and tested, to research unexplored territories of great fundamental interests. We investigate superfluid Fermi systems where pairs of two kinds of particles play a fundamental role for the superfluid properties. Here a simple but intriguing question arises: What happens if the numbers of two kinds of particles are imbalanced? As it might be experienced in your life – when you participate in a folk dance party where the numbers of men and women are imbalanced, someone would be frustrated as he/she cannot form a pair. We explore similar but far more complicated problems, i.e. nonlinear dynamics of spin-polarized fermionic superfluid systems, where we will observe non-linear dynamics of a mixture of superfluid and non-superfluid particles. How it behaves is an open problem. In addition, we study one of the smallest superfluid systems in the world: the atomic nuclei – a tiny entity which resides in the heart of each element. Actually the majority of atomic nuclei are known to manifest superfluid properties. We have also developed a computational code which describes superfluid dynamics in nuclear systems. We study two extreme situations as well: fusion – the merger of two nuclei, and fission – the dissociation of a heavy nucleus. In fact, only our group is able to perform fully microscopic simulations for superfluid dynamics in nuclear systems. We also expect that a lot of new phenomena will be discovered in this project, as our recent findings on significant effects of pairing field dynamics in induced fission and nuclear reactions will be further explored. Our scientific goals as well as technical requirements set this project among pioneering studies. We address these scientific problems, which have either not been investigated(pairing dynamics during nuclear collisions), or have represented the long standing goal and challenge for the microscopic description (induced fission). Last but not least, we explore a completely unknown territory (dynamics of topological excitations in polarized ultracold atomic gases). Therefore we expect to make important contributions and possibly breakthrough to a number of topics of great interest in both the field of ultracold atomic gases and low-energy nuclear physics. Thus, our project is cross-disciplinary and has a potential to provide breakthroughs in various fields of science.

Relativistic effects at non-standard temperature and pressure

Project Title: Relativistic effects at non-standard temperature and pressure
Project Leader: Krista Steenbergen, NZ
Resource Awarded: 18.1 million core hours on Marconi – KNL

Details

Team Members :
Universite Grenoble Alpes – FR
Massey University (Albany campus) – NZ

Abstract
Mercury’s thermodynamic properties have long been of interest to experimental and theoretical researchers, as the only elemental metal that exists in the liquid state under standard conditions (room temperature, standard pressure). Recent simulations have shown that mercury is a liquid at room temperature due to relativistic effects. The interplay of relativistic effects and pressure have never been theoretically explored. We will complete a full set of simulations melting bulk mercury at a series of increasing pressures. Comparing the results from relativistic and non-relativistic models, we will explore how relativistic effects alter or dictate the material properties of this interesting metal under pressure. We will additionally compare the non-relativistic model to Zinc melting simulations, in order to determine whether relativity can account for mercury’s unique properties among the group-12 metals.

GYSELA-near-critical

Project Title: GYSELA-near-critical
Project Leader: Guilhem Dif-Pradalier, FR
Resource Awarded: 41.1 million core hours on Curie

Details

Team Members :
CEA – FR

Abstract
The ongoing effort to master fusion energy is a major challenge for the coming years and many key problems are still at a stage of fundamental research. Amongst these, a better understanding of turbulent transport is a matter of steady progress and of strong international competition. As plasma confinement devices grow hotter and bigger their operation regime becomes less collisional (a kinetic approach is required) and closer to turbulence marginal stability (where aspects of self-organisation became dominant). Recent results have also pointed at an interplay between distant regions of the plasma, typically between core and edge plasma regions. This proposal builds upon years of development of the 5D gyrokinetic GYSELA code and confidence-building in its fundmental physics predictions. We plan to investigate near-marginal turbulence self-organisation, unravel its transport characteristics at plasma parameters relevant to current large and future plasma confinement devices like ITER, whilst self-consistently modeling the magnetised plasma from core to edge.

The pseudo-critical line in the QCD phase diagram

Project Title: The pseudo-critical line in the QCD phase diagram
Project Leader: Olaf Kaczmarek, DE
Resource Awarded: 66.6 million core hours on Piz Daint

Details

Team Members :
University of Bielefeld – DE
Brookhaven National Laboratory – USA

Abstract
The critical point in the QCD phase diagram, if it exists, is likely to be connected to the 2+1-flavor chiral phase transition at zero chemical potential. In fact, in the chiral limit the critical end point will be a tri-critical point at which the line of second order chiral transitions turns into a line of first order transitions. The goal of this project is to determine the chiral transition line at moderate, non-zero values of the chemical potential to explore the QCD phase diagram.

ZONALGENE – Non-linear analysis of zonal flow generation in magnetically confined plasmas

Project Title: ZONALGENE – Non-linear analysis of zonal flow generation in magnetically confined plasmas
Project Leader: Jeronimo Garcia, FR
Resource Awarded: 19 million core hours on MareNostrum

Details

Team Members :
Max Planck Institute for Plasma Physics – DE
Barcelona Supercomputing Center (BSC) – ES

Abstract
The onset and generation of turbulence is an open question in physics yet. The prediction and understading of the mechanisms leading to the spotaneous formation of turbulence is a long and outstanding problem in highly non-linear systems far away from equilibrium, such a magnetically confined plasmas. Turbulence generated in small scales compared to the system (Microturbulence) is one of the primary physical mechanisms which limit energy confinement in magnetically confined plasmas. The investigation of turbulence control and suppression is hence a critical goal for the optimization and design of future tokamak fusion reactors. The dependence of microturbulence on the different plasma parameters, e.g. temperature, density and magnetic field, is still not fully understood. This is in particular the case of the formation and damping of the so called zonal flows (large-scale flows similar to the atmospheric and oceanic flows, where zonal means latitudinal) and their dependence on important plasma parameters such as the plasma pressure (and pressure gradient), the plasma main ion mass or the rotation arising from external torque. These large scale flows can lead to strong deviations from the expected transport scaling for turbulence which assumes that the heat flux is mainly due to the local plasma fluctuations and mostly neglect mesoscascle physics, like zonal flows. The objective of this project is to quantitatively investigate the interdependence of zonal flows and other plasma parameters in particular those for which there is less experimental background, like high pressure gradients, multi ion species (including highly energetic ones in the range of the alpha particles generated by fusion reactions) and different isotope mixtures. We will perform non-linear simulations with the GENE code which solves the gyrokinetic equation (a reduction from 6D to 5D dimensions of the Vlasov equation performed by averaging the equation in the gyromotion). Realistic plasma conditions will be used, i.e. real plasmas from present day magnetic devices will be analysed by assuming electromagnetic fluctuations, real geometry and multi ions composition.

KARMA – Kinetic models of magnetic reconnection in pARtially ionised plasMAs

Project Title: KARMA – Kinetic models of magnetic reconnection in pARtially ionised plasMAs
Project Leader: Giovanni Lapenta, BE
Resource Awarded: 8 million core hours on SuperMUC

Details

Team Members :
KU Leuven – BE

Abstract
Plasmas are partially, not fully, ionised in the lower solar atmosphere, in planetary ionospheres, in the warm neutral interstellar medium, in interstellar clouds, in accretion disks and in experiments, e.g. the Magnetic Reconnection Experiment. Magnetic reconnection is a key process to convert energy stored in magnetic fields into bulk kinetic energy and particle heating in astrophysical, space and laboratory plasmas. It is the basic process behind magnetic field reconfiguration and particle accelerations in environments as diverse as planetary magnetospheres, stellar atmospheres, Giant Radio Galaxies and Active Galactic Nuclei. Magnetic reconnection in fully ionised plasma has been extensively studied in the past, both with fluid and kinetic approaches. Reconnection in partially ionised plasmas is, in comparison, still relatively unexplored. We intend to perform pioneering simulations of kinetic reconnection in partially ionised plasmas. Our aim is highly ambitious because kinetic simulations of partially ionised plasmas require one to account for complex interactions between the plasma particles, ions and electrons, and the neutral background. It is however a necessary step to uncover the mysteries behind solar coronal heating, solar flares, coronal mass ejections and other space weather and astrophysical phenomena.? To address collisions with a neutral background, we have augmented our Particle in Cell code, iPic3D, with a dedicated procedure to treat realistically collisions between a neutral gas and ionised species. This development allows iPic3D to model partially ionised astrophysical and laboratory plasma. Kinetic reconnection encompasses two aspects: the triggering mechanism and the energy conversion. The trigger of kinetic reconnection boils down to electron-scale processes that take place in the small region of space where electrons are not magnetised, the electron diffusion region. However, a large part of the energy conversion happens far away from this area, namely at the downstream fronts. The downstream fronts are the regions where the heated and accelerated plasma engulfs the plasma still at rest. In a partially ionised plasma, one can expect deep modifications both of the central electron-scale diffusion region and of the energy conversion processes. At the electron diffusion region, resistivity coming from collisions with the background must give rise to a further non-ideal electric field term, with the related consequences in terms of reconnection rate. Also, the highly structured electron distribution functions identified in collisionless reconnection in fully ionised plasmas must be modified by collisions with the background. The downstream fronts must be loci of enchanted energy exchange between the plasma and the background gas. ?More in general, we can expect that the characteristic signature, simulated and observed, of collisionless reconnection, the Hall electric and magnetic field, will be modified by the spatially inhomogeneously production of ions and electrons via interaction with the background. These three fundamental aspects will therefore be the focus of our investigation. We will study: I) modification of the signature of kinetic reconnection, II) modification of the mechanism triggering kinetic reconnection – electron diffusion region processes and III) modification of the energy exchange mechanism- downstream front processes as a result of interaction with the background.

Combustion Noise and Combustion Dynamics

Project Title: Combustion Noise and Combustion Dynamics
Project Leader: Matthias Meinke, DE
Resource Awarded: 217 million core hours on Juqueen

Details

Team Members :
KU Leuven – BE

Abstract
Research in combustion noise is increasingly gaining interest for mainly two reasons. First, it is expected to become an important contributor to the overall sound emission of future jet aero engines, since the sound emission of the other components, such as jet noise, fan noise, etc., is being reduced through recent technological progress. Second, combustion noise can lead to thermoacoustic instabilities. Especially the lean-premixed regime, which is favorable for a highly efficient combustion process with low emissions, is prone to an unstable feedback loop between the flame’s unsteady heat release and its resulting acoustic emission that can lead to hazardous loads within the combustion chamber. The goal of this project is to gain a deeper understanding of the acoustic source mechanisms, which is necessary for the prediction and avoidance of thermoacoustic instabilities in future combustor systems at lean conditions. This is achieved by a hybrid approach. First, the combustion process is computed with a large-eddy simulation. The resulting flow field determines the acoustic sources, which are used in the second step to compute the acoustic emission. With this approach, the acoustic emission can be analyzed for various setups, i.e., depending on flow paramters, flame parameters and the burner geometry. The numerical data is used to further analyze the interaction between the turbulent flow field, the flame, and the acoustics.

top

Mathematics and Computer Sciences (0)

top

Universe Sciences (12)

MOONSHINE – Magnetospheric Observation Of Numerically Simulated Heavy Ions Near Earth

Project Title: MOONSHINE – Magnetospheric Observation Of Numerically Simulated Heavy Ions Near Earth
Project Leader: Minna Palmroth, FI
Resource Awarded: 75 million core hours on Marconi – KNL

Details

Team Members :

Abstract
Space weather is a term used to characterize potentially harmful effects to human health or to technological systems on ground and in space. It has recently received intensive attention in the US, as the President issued an Executive Order to mitigate effects caused by severe space weather. In Europe, the European Space Agency and European Union are both propagating space weather mitigation, and especially emphasize that the space weather modelling capabilities should be updated. Space weather is caused by the Sun emitting high-energy particles and solar wind, a stream of charged particles carrying the solar electromagnetic field. Vlasiator is a newly developed 6-dimensional Vlasov theory-based simulation. Ions are distribution functions, while electrons are charge-neutralizing fluid, enabling a self-consistent global plasma simulation that can describe multi-temperature plasmas. The novelty is that by modelling ions as distribution functions the outcome is numerically noiseless. The new version of Vlasiator supports multiple ion populations; protons and a second population also modeled using distribution functions. Due to the multi-dimensional approach in ion scales, Vlasiator’s computational challenges are immense. We use advanced high performance computing techniques to allow massively parallel computations. The runs enabled by two previous PRACE Tier-0 accesses have showed that Vlasiator produces physics of space weather with unprecedented quality, revealing a rich playground of collective phenomena that cannot be interpreted by local in situ observations only. This proposal concerns the scientific basis of space weather, in which three phenomena explain the majority environmental effects: 1) magnetic reconnection enabling energy and mass transfer between different magnetic domains, 2) shocks forming due to supersonic relative flow speeds between plasma populations, and 3) particle acceleration associated with both reconnection and shocks. Combining Vlasiator with newest spacecraft data, we aim at breakthroughs in explaining the local physics globally and self-consistently in 1. Role of heavy ions in reconnection in a global system including both dayside and nightside reconnection sites. We define simulations in the noon-midnight meridional plane for a variety of interplanetary magnetic field (IMF) orientations to quantify reconnection rates and instabilities, and compare to our previous results utilising Tier-0 runs without heavy elements. 2. Role of heavy elements in shock formation and acceleration. We outline simulations in the ecliptic and noon-midnight meridional planes to evaluate the interplay between the particle acceleration and shock evolution, and their consequences to magnetosheath processes in the global shock-sheath system. The results are compared to earlier Tier-0 results without heavy elements. These questions are novel, and they have never been targeted with global kinetic scale simulations before. The science objectives form the fundamental core of space plasma physics, having a crucial role also in plasma fusion and astrophysics. The project is extremely timely now that the supercomputing resources are becoming large enough. Our team has a unique opportunity to utilise the novel Vlasiator, and high fidelity new data collections from international spacecraft. The team includes complementary state-of-the-art excellence to carry out the ambitious tasks. This unique combination guarantees numerous breakthroughs in space physics and adjacent fields.

How do globular clusters form?

Project Title: How do globular clusters form?
Project Leader: Florent Renaud, UK 
Resource Awarded: 15.1 million core hours on Marconi – KNL

Details

Team Members :
University of Surrey – UK

 

Abstract
Young galaxies in the early Universe are able to form very massive and dense stellar systems called globular clusters. Despite their ubiquity in most of the galaxies and in our Milky Way in particular, the origin of globular clusters remains shrouded in mystery. Understanding how, when and where they form is however key to pin down the evolution of galaxies. The difficulty of the task lies in the wide range of scales and physical processes involved, from the large scale organisation of dark matter and interstellar gas, down to the inner structures of the cluster-forming clouds. Our project aims at shedding new lights on this topic by resolving the formation of globular cluster in cosmological context. For the first time, we will be able to connect the physics ruling the early life of globular clusters with their observations at present-day, and to understand their co-evolution with their environment, and thus the history of galaxies in general and of the Milky Way in particular.

BNSdisk — Unequal mass binary neutron star mergers – gravitational wave emission and properties of the accretion disk

Project Title: BNSdisk — Unequal mass binary neutron star mergers – gravitational wave emission and properties of the accretion disk
Project Leader: Roberto De Pietri, IT
Resource Awarded: 30 million core hours on Marconi – KNL

Details

Team Members :
Universitat de València – ES
Aristotle University of Thessaloniki – GR
Parma University – IT
Louisiana State University – USA

Abstract
The recent detection of gravitational waves (GWs) from the merger of binary black holes has opened a new window for the investigation of compact objects. Within the coming months, it is expected that the LIGO/Virgo interferometers will also detect GWs produced during the inspiral and merger of binary neutron stars (BNS). These signals will contain information about the behaviour of matter at extreme densities and their analysis will help constraint the equation of state (EOS) of neutron stars. The experimental searches for such signals rely on the filtering of the data with waveform templates from general relativity. Numerical relativity (NR) simulations of BNS are essential to provide the GW signals associated with the merger and post-merger phases. Furthermore, the recent observation of the binary pulsar J0453+1559, with a 0.75 mass ratio, fully justifies the NR investigation of unequal-mass BNS mergers, the central theme of this proposal. This project will simulate unequal-mass BNS mergers leading to black hole – thick accretion disk systems using the Einstein Toolkit, which provides a complete, open-source, production-level infrastructure for state-of-the-art NR simulations. It will consider sequences of BNS mergers with the same total baryonic mass, different mass ratios, and three nuclear EOS (Sly, H4, and a finite-temperature EOS). The goal of the project is to understand the impact of the mass ratio on the GW signal emitted during the inspiral, where tidal forces are most important, as well as the impact of neutrinos, temperature and composition on the physics of the merger and post-merger remnants. The exploration of the large parameter space envisaged in our proposal may break the current empirical relations between the GW peak frequencies and the neutron star compactness, which would require accommodating the GW observations to the neutron star EOS when the source is an unequal-mass BNS system. Moreover, the project will shed light on the long-term dynamics and non-axisymmetric instabilities of the disks and their connection with the generation of gravitational radiation. At the brink of the second observation run of Advanced LIGO and Advanced Virgo (expected to start by late November 2016) and given the increased sensitivity of the observatories for detecting new compact binary coalescence systems, our project is both scientifically sound and extremely timely.

Magneto – “Effect of Magnetar Level Fields in Binary Neutron Star Mergers”

Project Title: Magneto – “Effect of Magnetar Level Fields in Binary Neutron Star Mergers”
Project Leader: Bruno Giacomazzo, IT
Resource Awarded: 33.4 million core hours on Marconi – KNL

Details

Team Members :
University of Trento – IT
University of Trieste – IT

Abstract
The project’s aim is to study the effects of large magnetic field amplifications on the post-merger evolution of binary neutron stars. Such events are powerful sources of gravitational waves that can be detected by ground-based interferometers, such as advanced LIGO and Virgo. Moreover, they are thought to be the cause of short gamma-ray bursts, which are among the most powerful explosions in the universe. Magnetic fields play an important role in the dynamics of these systems, and the focus of this project is on the detailed mechanisms causing large amplification of the magnetic fields during the merger, and the resulting effect on the post-merger dynamics, emission of gravitational waves, and possible generation of short gamma ray bursts.

GWBNS – Gravitational waves from binary neutron star mergers

Project Title: GWBNS – Gravitational waves from binary neutron star mergers
Project Leader: Sebastiano Bernuzzi, IT
Resource Awarded: 33.4 million core hours on Marconi – KNL

Details

Team Members :
Max Planck Institute – DE
INFN – IT
Princeton – USA

Abstract
The gravitational wave signal from coalescing binary neutron stars convey unique information about the stars’ masses and spins as well as their composition, thus placing the strongest constraints on the equation of state of dense matter at supranuclear densities. Such gravitational radiation is observable by the network of ground-based gravitational wave interferometers (LIGO, Virgo, IndIGO and KAGRA) that will start operations within the next five years. A necessary condition for measuring the binary parameters from these observations is the availability of accurate theoretical predictions of the waveforms using general relativity. Waveforms models can be constructed by combining approximate analytical and fully numerical approaches to the relativistic two-body dynamics, including spins and tides. In this project we will perform high performance computing simulations to calculate the gravitational waves from the late stage of the coalescence process. We will employ state-of-art numerical relativity methods to solve Einstein spacetime equations and general relativistic hydrodynamics. The goal is to develop a waveform model suitable for gravitational wave astronomy purposes.

The first luminous objects and reionisation

Project Title: The first luminous objects and reionisation
Project Leader: Joakim Rosdahl, FR
Resource Awarded: 13.7 million core hours on SuperMUC

Details

Team Members :
University of Zürich – CH
Centre de Recherche Astrophysique de Lyon – FR
Universite de Strasbourg – FR
University of Cambridge – UK

Abstract
The Epoch of reionisation (EoR) is a fascinating chapter in the history of the Universe. It began when the first stars formed, bringing an end to the so-called Dark Ages. As their hosting dark matter (DM) haloes grew more massive, intergalactic gas rushed in and these first stars became the first galaxies. They emitted phenomenal amounts of ultraviolet radiation into intergalactic space, which ionised and heated the atoms that make up intergalactic gas, enhancing the pressure of the intergalactic medium to the point where it may have resisted the gravitational pull of the smaller DM haloes, stunting their growth. During the EoR, the large-scale properties of the Universe were thus strongly tied to the small-scale physics of star and galaxy formation. From current observations, we can indirectly infer only limited information about this epoch, when ionised regions grew and percolated to fill the Universe about one billion years after the Big Bang. We don’t know when the EoR started, how long it lasted, what types of galaxies were mainly responsible for making it happen (such as high- versus low-mass), and how this major shift affected the subsequent evolution of galaxies in a now much hotter environment. Soon our view of the EoR will change dramatically, as in 2018 the James Webb Space Telescope (JWST) is deployed into orbit around the Sun, and in 2020 the Square Kilometre Array (SKA) comes online. Both telescopes will perform unprecedented observations of the young and far-away Universe, SKA revealing the large-scale process of reionisation and JWST allowing the first robust measurements of the physical properties (stellar masses, star formation rates, abundances, clustering, …) of a large population of galaxies during the EoR. Yet, while those telescopes will be extremely powerful, most details surrounding the interplaying physics constituting early galaxy evolution and reionisation are still far out of reach observationally. To understand the physics, we need to back the limited information from observations with theory, using cosmological simulations, which combine, in three dimensions, the gravitational forces that led to the formation of galaxies, hydrodynamics of the collapsing gas, star formation, supernova explosions, emission of radiation from stars, and radiation-gas interactions. These simulations will evolve large volumes representing the early Universe, consisting of hundreds of interacting galaxies massive enough for their real counterparts to be observed by the JWST. The radiation-hydrodynamical simulations follow the collapse of matter from an initial almost uniform state and allow populations of stars to form and interact realistically inside thousands of galaxies resolved by millions of resolution elements each. For the first time, our simulations will describe the EoR self-consistently from small ISM scales to large IGM scales, improving our understanding of the significance of small-scale physics inside galaxies on the large-scale early evolution of the Universe and the interactions between those young galaxies. The simulations will aid to clear the picture and understand the underlying physics producing the wealth of data which will come from observations in the coming years.

SIMCODE-1

Project Title: SIMCODE-1
Project Leader: Marko Baldi, IT
Resource Awarded: 6 million core hours on Marconi – KNL, 6 million core hours on MareNostrum

Details

Team Members :
Heidelberg Institute for Theoretical Studies – DE
Alma Mater Studiorum – Bologna University – IT
INAF – IT
SISSA – IT
University of Cambridge – UK

Abstract
The next decade will see the advent of Precision Cosmology, with a large number of observational campaigns promising to unveil the most intriguing mysteries of the Universe, like the origin of the accelerated expansion or the nature of the dark matter, and to constrain the cosmological parameters to percent precision. One of the main goals of such challenge is to test possible extensions of the standard cosmological scenario. In order to exploit at best the high quality of the upcoming data an exquisite accuracy in the prediction of observable quantities for a wide range of non-standard cosmologies is needed, which requires the use of large and complex numerical simulations. Nonetheless, so far simulations have not properly accounted for the possible degeneracies among different and mutually independent extensions of the standard cosmological model. These cosmic degeneracies might lead to gross misinterpretations of the data and to significant biases in the determination of the models’ parameters. The SIMCODE initiative aims to investigate cosmic degeneracies by performing the largest and most comprehensive suite of cosmological simulations that fully account for possible degeneracies among a wide range of non-standard cosmologies, allowing to identify the best observational strategies to disentangle such degeneracies and to correctly assess the true discriminating power of different observables with respect to extended cosmologies. This effort has already started with a previous PRACE allocation in which the observational degeneracies between f(R) Modified Gravity models and a cosmological background of massive neutrinos have been investigated by means of dedicated combined simulations, highlighting for the first time the strong degeneracy between these two independent extensions of the standard cosmological scenario. The present SIMCODE1 project will continue on this track and will perform combined N-body simulations that simultaneously include specific implementations of other independent non-standard models, once again associated with the independent sectors of Dark Energy and Modified Gravity, Dark Matter, and early Universe physics. This is an innovative approach which has not yet been employed by any other research team in the world, and that will allow to correctly assess the true discriminating power of a wide range of present and upcoming observational surveys, such as e.g. DES, HetDEX, VIPERS, LSST, eBOSS, DESI, J-PAS, and ultimately the Euclid satellite mission, which is the main target of the whole SIBEL computational initiative. Hence, the SIBEL2 project is at the same time innovative, feasible, and timely, as it provides the community with an essential guideline for an unbiased interpretation of upcoming cosmological observations. For such challenging numerical program we require about 16 Mio CPU hours on a top-level HPC infrastructure. The computational budget is evenly distributed among simulations of different size, with 3 extremely large simulations and about 64 among Large- and Medium- size runs aimed at sampling the parameter space for the combined models to be performed on Marconi-KNL, while another 10 hydrodynamical simulations to be run on Marenostrum. The budget already includes all the basic post-processing analysis envisaged for the scientific exploitation of the simulations.

Multi-scale simulations of Cosmic Reionization

Project Title: Multi-scale simulations of Cosmic Reionization
Project Leader: Ilian Iliev, UK
Resource Awarded: 52 million core hours on MareNostrum; 34 million core hours on Piz Daint

Details

Team Members :
Leibniz-Institut fuer Astrophysik Potsdam (AIP) – DE
Universidad Autónoma de Madrid – ES
Observatoire Astronomique de Strasbourg – FR
Chosun University – KR
Stockholm University – SE
University of Sussex – UK
The Harvard-Smithsonian Center for Astrophysics – USA
The University of Texas at Austin – USA

Abstract
The first billion years of cosmic evolution are one of the last largely uncharted territories in astrophysics. During this key period the cosmic web of structures we see today first started taking shape and the very first stars and galaxies formed. The radiation from these first galaxies started the process of cosmic reionization, which eventually ionized and heated the entire universe, in which state it remains today. This process had profound effects on the formation of cosmic structures and has left a lasting impression on them. This reionization process is inherently multi-scale. It is generally believed to be driven by stellar radiation from low-mass galaxies, which cluster on large scales and collectively create very large ionized patches whose eventual overlap completes the process. The star formation inside such galaxies is strongly affected by complex radiative and hydrodynamic feedback effects, including ionizing and non-ionizing UV radiation, shock waves, gas cooling and heating, stellar winds and enrichment by heavy elements. Understanding the nature of the first galaxies and how they affect the progress, properties and duration of the cosmic reionization requires detailed modelling of these complex interactions.The aim of this project is to combine a unique set of simulations of cosmic reionization covering the full range of relevant scales, from very small, sub-galactic scales, for studying the detailed physics of radiative feedback, all the way to very large cosmological volumes at which the direct observations will be done. These simulations will be bases on several state-of-the-art numerical tools, including Adaptive Mesh Refinement (AMR) techniques for achieving very large dynamic range in radiative hydrodynamics calculations (RAMSES-RT code), GPU-based acceleration for radiative hydrodynamics (RAMSES-CUDATON),and a massively-parallel, highly numerically efficient radiative transfer method for accurate modelling at large scales (C2-Ray). We will complement the numerical simulations with semi-analytical galaxy formation modelling to explore the large parameter space available, to improve the treatment of reionizing sources in large-scale radiative transfer simulations as well as to derive detailed observational features of the first galaxies in different observational bands. The questions we will address are: 1) how do the radiative feedback from the First Stars hosted in cosmological minihaloes and dwarf galaxies affect the formation of early structures and subsequent star formation?; 2) how much does high-redshift galaxy formation differ from the one at the present day? What are the observational signatures of the first galaxies? 3) how important is the recently pointed out effect of local modulation of the star formation in minihaloes due to differential supersonic drift velocities between baryons and dark matter?; 4) how does the metal enrichment and the transition from Pop III (metal-free) to Pop II stars occur locally and how is this reflected in the metallicity distribution of the observed dwarf galaxies and globular clusters? and 5) How are these feedback effects imprinted on large-scale observational features?

SN-SNR87A – Evolving supernova explosions to supernova remnants through 3D MHD modelling: the case of SN 1987A

Project Title: SN-SNR87A – Evolving supernova explosions to supernova remnants through 3D MHD modelling: the case of SN 1987A
Project Leader: Salvatore Orlando, IT
Resource Awarded: 64 million core hours on Marconi – KNL

Details

Team Members :
Istituto Nazionale di Astrofisica – IT
University of Palermo – IT
Fukuoka University – JP
Kyoto University – JP
National Astronomical Observatory – JP
RIKEN – JP

Abstract
Supernova remnats (SNRs) have a complex morphology and a non-uniform distribution of ejecta reflecting pristine structures of the progenitor supernova (SN) as well as the imprint of the early interaction of the SN blast with the circumstellar medium (CSM). Linking SNR morphology to their SN progenitors is an essential step to open new exploring windows on SN and SNR issues: 1) to probe the physics of SN engines by providing insight into the asymmetries that occur during the SN explosion, and 2) to investigate the final stages of stellar evolution by unveiling the structure of the medium surrounding the progenitor star. A detailed 3D model connecting SN to SNR is presently missing. The aim of this project is to describe the complete 3D evolution of ejecta from the on-set of a core-collapse SN to the development of its remnant with unprecedented spatial resolution and completeness to answer important questions as: how does the final remnant morphology reflect the asymmetries developing in the immediate aftermath of the SN? how does the nucleosynthetic layering of stellar material map in the remnant morphology? Because of its youth and proximity, SN 1987A is an attractive laboratory for studying the SN-SNR connection. Here we propose to describe its evolution from the on-set of the SN to the remnant development. First we model the 3D core-collapse of SN 1987A for 2 days. From then on we follow the system evolution, by using a 3D MHD model describing the interaction of the remnant with the CSM. Our project is based on two large 3D simulations, one for the SN and the other for the SNR. We perform multi-species simulations to follow the evolution of the isotopic composition of ejecta. The geomeric domain is a cartesian box with an effective maximum numbers of grid points 1024^3. We follow the changes in length scale of the system (spanning several orders of magnitude from the SN to the SNR) by gradually extending the computational domain as the forward shock propagates outward and remapping all the MHD variables in the new domain. We follow the SNR evolution for 50 yr. We use the already developed setup with FLASH 3D hydrodynamic code for the SN model and with the PLUTO 3D MHD code for the SNR model. Both codes are already well-tested on our problem in 3D. Also the codes have already been tested on systems similar to CINECA/MARCONI/KNL on a problem with a similar size, showing an excellent scalability up to ~ 32000 processors (CINECA/ISCRA Award N. HP10CWYDMI). The amount of resources requested has been extrapolated from SN simulations performed on the SR16000 cluster at the Kyoto University (Japan) and from SNR simulations performed on the CINECA/MARCONI/broadwell cluster. The high-resolution 3D SN-SNR simulation requires a computing time of ~ 49 Mhours on the MARCONI/KLN system and a number of cores > 10000. Using 16000 cores per step, the simulation can be performed in ~ 128 days of continuous wallclock time and, including the warm-up, a total of ~ 5.5 months and ~ 64 Mhours CPU-time.

SPOTSIM – Spot-forming convection simulations

Project Title: SPOTSIM – Spot-forming convection simulations
Project Leader: Petri Kapyla, FI
Resource Awarded: 20 million core hours on MareNostrum

Details

Team Members :
Max-Planck-Institut for Solar System Research – DE
NORDITA – SE

Abstract
The solar magnetic field is generated by dynamo processes operating in its interior. The large-scale magnetic field is manifested by dark sunspots, where strong magnetic fields inhibit convective heat transport and lower the temperature, at the solar surface. The current paradigm of solar dynamo and spot formation assumes that strong magnetic flux tubes are generated in the tachocline, a layer of strong shear at the interface between the convection zone and the radiative interior, whence they rise without appreciably interacting with the highly turbulent convection on the way to the surface to form sunspots. This paradigm has been recently challenged by numerical simulations which indicate that strong distributed magnetic fields can be generated also in the bulk of the convection zone and that solar-like magnetic cycles can be obtained without tachoclines. However, a mechanism to form magnetic field concentrations is required for sunspots to form in a distributed dynamo. A promising candidate is the Negative Effective Magnetic Pressure Instability (NEMPI) which occurs in a stratified fluid under the influence of large-scale magnetic fields. The instability has been demonstrated to exist in idealised simulations but yet in turbulent convection. Previous convection simulations used to study magnetic structures have relied on either uniform imposed large-scale fields or boundary conditions to feed the system with magnetic flux. Such models produce spots but they leave the question of their formation process untouched. In the present project we use 24 million CPU hours to study the formation of magnetic structures with turbelent convection simulations where a more realistic large-scale magnetic field configuration is either imposed or generated self-consistently by the interaction of highly stratified turbulence and the overall rotation and shear in the system. We use unprecedently high resolution simulations that can enable the first self-consistent models of spot-forming dynamos.

Energy transfer across the Earth’s magnetospheric boundary layers

Project Title: Energy transfer across the Earth’s magnetospheric boundary layers
Project Leader: Takuma Nakamura, AT
Resource Awarded: 12.8 million core hours on MareNostrum

Details

Team Members :
Max-Planck-Institut for Solar System Research – DE
NORDITA – SE

Abstract
Space between planets, stars, and galaxies is filled with plasma (ionized gas) with its density small enough to neglect Coulomb collisions as well as fluid viscosity for many important phenomena. In such a collisionless system, the boundary layer between regions with different plasma properties plays a central role in transferring mass and energy and controlling the dynamics of the system. In the Earth’s magnetosphere, for example, the mass and energy input from the solar wind is transferred and changes its properties through different physical processes at the boundary layer of the magnetosphere, called the magnetopause. These transfer processes through the magnetopause are considered as key research objectives not only in fundamental space plasma physics but also in space weather, since it eventually affects the global dynamics of the magnetosphere related to the auroral substorms and to the generation of energetic particles in the radiation belts. This project will systematically investigate typical patterns of the mass and energy transfer processes at the magnetopause based on state-of-the-art plasma kinetic simulations under realistic conditions obtained from in-situ spacecraft for the first time. Although a number of studies have treated this important topic, quantitative aspects of the realistic transfer process are still poorly understood. This is mainly because the transfer process in the magnetosphere tends to be caused over a broad range of temporal and spatial scales (from the smallest electron kinetic scale to the magnetohydrodynamic (MHD) scale), which cannot be handled only from spacecraft observations. Thus, the scientific focus of this project is to quantify the transfer processes more exactly than previous studies covering all necessary scales by performing large-scale three-dimensional (3D) fully kinetic simulations under realistic magnetopause conditions. This project is timely because the proposed systematic study is feasible only by the combination of the large-scale kinetic simulations in the Tier-0 system and the recently (in 2015) launched high-resolution MMS (Magnetospheric MultiScale mission) spacecraft which allows not only to give realistic conditions to the simulations but also to confirm the adequacy of the simulations resolving the kinetic-scales.

EAGLE-XL

Project Title: EAGLE-XL
Project Leader: Richard Bower, UK
Resource Awarded: 30 million core hours on Piz Daint

Details

Team Members :
Max Planck Institute for Astrophysics – DE
Instituto de Astrofísica de Canarias – ES
Leiden University – NL
Durham University – UK
Jodrell Bank Centre for Astrophysics – UK
Liverpool John Moores University – UK
University of Sussex – UK

Abstract
The Universe has a rich and complex history. Billions of Euro are being invested in new telescopes to survey the sky in more depth and greater detail than has ever before been possible. At the same time, great advances are being made in understanding the processes that drive the formation of galaxies. A generic paradigm is now in place, and the properties of galaxies can be broadly understood as the interaction of gas cooling in dark matter haloes, regulated by feedback from stars and black holes. In order to make further progress, we need to generate a virtual Universe by applying the laws of physics to the initial conditions that emerge from the Big Bang. This is the aim of cosmological hydrodynamic simulations. With our previous allocation of PRACE time (ref-2012071242) our team undertook the EAGLE (Evolution and Assembly of GaLaxies and their Environments) simulation, and succeeded in recreating the most realistic simulations of the Universe to date. The reference paper has had a huge impact on the subject and is the highest cited paper in 2015 in the astrophysics (out of 26000 papers). EAGLE provides a “laboratory” for understanding the physics of galaxy formation, allowing us to experiment, picking out particular objects and quantifying the physical processes that determine (for example) the size and mass of galaxies. Furthermore, while real galaxies are seen only at snapshots in time, the simulations allow entire formation histories to be traced across cosmic time. The success of the simulation is, however, tempered by its modest volume which severely limits its scope and applicability to new observational surveys. In this project we request time to undertake a 100 billion particle simulation of the Universe, increasing the volume of the EAGLE simulation by a factor 15 so that it contains 100,000 realistic galaxies and samples length scales up to 250 Mpc. This will allow us to: (1) identify statistical samples of rare (and bright) objects – this is critical since these are the only objects that can be detected in the distant Universe – and to understand the origin of the diversity of galaxy properties; (2) to create mock catalogues of the Universe, predicting the surveys that will be made with the next generation of telescopes – a crucial step since such surveys can only be interpreted by `forward modelling’ due to biases and selection effects in the observations; (3) trace the gravitational structure of the Universe on large scales galaxies, making it possible to test cosmological models and search for departures from Einstein’s theory of gravity – it is critical that we determine the extent to which galaxy formation physics limits the applicability of such tests, and optimise the observational strategy to minimise this bias.

top

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