Projects from the following research areas:
Biochemistry, Bioinformatics and Life sciences (5)
Targeting Alzheimer’s-associated amyloid-beta using small molecules
Project Title: Targeting Alzheimer’s-associated amyloid-beta using small molecules
Project Leader: Michele Vendruscolo
Resource Awarded: 28 million core hours on MareNostrum
Gabriella Heller University of Cambridge UK , Thomas Löhr University of Cambridge UK
Alzheimer’s disease is the most common cause of dementia, but no cure is currently available to treat it. One of the hallmarks of this disease is the formation of neurotoxic plaques through the aggregation of the intrinsically disordered amyloid-β (Aβ) peptide. Decades of experimental studies have elucidated some of the mechanisms behind this aggregation process, but it is becoming increasingly clear that computational approaches can uniquely provide the atomic-level information required to define the detailed molecular events underlying Aβ aggregation and its inhibition. Recently, by applying novel drug discovery approach based on chemical kinetics, we discovered several small molecules able to suppress key microscopic steps in Aβ aggregation by binding the monomeric form of the protein. The availability of these small molecules provides an unprecedented opportunity to clarify which of their chemical features are responsible for their effects on Aβ aggregation, which in turn will create new opportunities for the rational design of more potent inhibitors. To exploit this opportunity, we propose to use molecular dynamics simulations using state-of-the-art enhanced sampling algorithms and experimental restraints to study how these small molecules modulate the structural properties of Aβ, which, being intrinsically disordered, should be described as a structural ensemble, in which each structure is assigned a distinct probability of occuring. As there are no experimental methods that can achieve the required accuracy, molecular dynamics simulations are currently the most effective way to obtain this atomic-level information of conformational ensembles of intrinsically disordered proteins. The accuracy required to characterise their modulation by small molecules is only currently possible using large computational resources. The availability of these offer us the unprecedented opportunity to simulate up to 100 molecules interacting with monomeric Aβ. Taken together, these simulations will help us elucidate the complex mechanism by which small molecules can interact with disordered proteins towards drug discovery for Alzheimer’s disease.
CryptoPocketSim. Understanding the mechanism of cryptic pocket formation at protein-protein interfaces
Project Title: CryptoPocketSim. Understanding the mechanism of cryptic pocket formation at protein-protein interfaces
Project Leader: Francesco Gervasio
Resource Awarded: 20 million core hours on Piz Daint
Antonija Kuzmanic University College London UK
The long-term aim of this project is to understand the mechanism of formation of “cryptic” pockets and allosteric regulation in relevant drug targets by means of enhanced-sampling simulations. “Cryptic” pockets, that is, sites on protein drug targets that only become apparent when a drug binds, offer an attractive opportunity for the development of drugs for difficult targets where classic drug-design strategies fail. However, due to their hidden nature, they have been in most cases discovered by chance. What is more, the molecular mechanisms by which cryptic pockets are formed, or even how common they are is still unclear. My group has pioneered the use of enhanced sampling simulations to discover cryptic pockets and described a previously unknown one in an anticancer target (FGFR). We also investigated the nature of cryptic sites in a number of pharmacological targets, such as the beta-lactamase TEM1 (a target for antimicrobial resistance). Building on that successful experience, we recently developed an effective Hamiltonian Replica Exchange-based approach (SWISH) and combined it with small molecular probes to sample cryptic pocket opening. Our initial results published in JACS attracted a lot of interest from both academia and industry. Thanks to our PRACE 15 allocation we were able to use a combination of SWISH and Metadynamics on Mare Nostrum to study more complex cases, such as the cryptic pocket forming at the protein-protein interface of the trimeric tumour necrosis factor alpha (TNFa). This revealed an interesting mechanism and allowed us to fine tune our algorithms. Here we propose to complete the study of TNFa and expand our investigation to the tumour necrosis factors super-family, an extremely promising target family for many complex diseases. This project is a continuation and expands on our successful PRACE 15 CryptoPocketSim project and will continue to benefit from a collaboration with a committed industrial partner.
Project Title: IFs/RFs-RC
Project Leader: Maria Ramos
Resource Awarded: 15 million core hours on MareNostrum
Pedro Fernandes University of Oporto PT , Ana Oliveira University of Oporto PT
The bacterial ribosome is an important drug target because of its role in protein synthesis. Stop codon-specific release factors (RFs) bind to the ribosome and catalyse poplypeptide release to terminate translation. However, to date there are no small-molecule inhibitors of bacterial protein synthesis that specifically target RFs. Ana Oliveira and et al. have demonstrated that the function of release factor 2 (RF2) could be directly inhibited in the translation cycle. Starting from in silico characterization of the receptor, virtual screening, they identified two novel inhibitors that killed the bacteria by interacting directly with RF2. Furthermore, we would like to determine if the bacterial Initiation Factor 2 also exhibits druggable properties and characterize the two receptors to develop new drug-like compounds that exhibit antibiotic properties.
MICNA – Mechanisms of ion conduction in sodium channels
Project Title: MICNA – Mechanisms of ion conduction in sodium channels
Project Leader: Carmen Domene
Resource Awarded: 102 million core hours on Piz Daint
Simone Furini University of Siena IT
The transport of ions and water across membrane cells and organelles is a prerequisite for many of life’s processes. Ion channels are proteins that span across the cell membrane allowing the passage of ions from one side of the membrane to the other. Ion channels show ion selectivity, permitting some inorganic ions to pass, but not others, at rates approaching free-diffusion, and they are not continuously open, instead they open and close in response to certain stimuli. Due to the inherent experimental difficulties in obtaining structural data for these systems, it is essential that we maximize our understanding of the structure/function relationships of those membrane proteins for which there is an experimental structure available. As their roles in neurological disorders and diseases in general is prominent, understanding their mechanisms of action and their interactions at atomic level with drugs targets is crucial to develop new treatments.
Realistic cerebellar multi scale network reconstruction
Project Title: Realistic cerebellar multi scale network reconstruction
Project Leader: Egidio D’Angelo
Resource Awarded: 30 million core hours on Joliot Curie – KNL
Stefano Masoli University of Pavia IT , Martina Francesca Rizza University of Pavia IT , Claudia Casellato University of Pavia IT , Stefano Casali University of Pavia IT , Elisa Marenzi University of Pavia IT , Chaitanya Medini University of Pavia IT
The cerebellum is part of the central nervous system (CNS) and performs its physiological function as a massive parallel processor, composed by many parallel modules. It receives a massive quantity of information from the motor and somatosensory brain cortex and physical feedback, from the peripheral nervous system (PNS). Its purpose is to elaborate, as fast as possible, motor and postural information to keep muscle tone, balance and other motor related behaviours, which, if it is disrupted, causes motor dysfunction and ataxias. In the last two decades, its involvement in the higher cognitive functions was proved even more fundamental. This is due to the discovery of specific pathways connecting the cerebellum to cognition related part of the brain, like the amygdala, involved in the fair and emotional responses or the basal ganglia, part of a system dedicated to procedural learning, cognition and emotions. These reciprocal connections uncovered the involvement of the cerebellum in known neurodegenerative diseases, such as frontotemporal dementia, psychosis, Alzheimer and Parkinson diseases. Until the last decade, the only way to investigate the cerebellum was trough the animal model and non-invasive techniques in humans. Since supercomputers have progressed a lot in their computational power, the reconstructed and simulated, of the cerebellum network, through realistic network models formed by hundred thousand neurons. The first step is the reconstruction of the biophysical properties of each neuron to a high degree of detail (realistic modelling), based on available experimental data. This approach allows to understand the fundamental physiological properties of the neurons and 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, 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 multiscale 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. The ability to simulate multi-scale networks, composed of hundred thousand neuron models, can help to shed light on the overall network activity, intrinsic and extrinsic signals coming from other parts of the brain, to better diagnose the supra mentioned diseases.
Chemical Sciences and Materials (10)
ADRENALINE – hAliDe peRovskites sEqueNtiAL deposItioN mEchanism (by ab initio rare events simulations).
Project Title: ADRENALINE – hAliDe peRovskites sEqueNtiAL deposItioN mEchanism (by ab initio rare events simulations)
Project Leader: Simone Meloni
Resource Awarded: 78 million core hours on Piz Daint
Team Members: Lorenzo Gontrani University of Rome Sapienza IT , Diego Di Girolamo Sapienza University of Rome IT , Sara Marchio Sapienza University of Rome (IT) IT
The objective of the ADRENALINE project is to study the atomistic mechanism, energetics and kinetics of the sequential deposition method of (hybrid) lead halide perovskites for 3rd generation solar cells and optoelectronics applications. The project is based on ab initio rare event simulations of the key step of the process, the intercalation of the cation A+ (methylammonium, CH3NH3+, formamidinium, (NH2)2CH+, and Cesium cation, Cs+) and anion X- (I-, Br- and Cl-) and reorganization of the intermediate product into the final one. We intend to address several questions about these stages: • What are the intercalation site of cations and anions? How do they change from cation to cation and from anion/precursor to anion/precursor? • What is the driving force pushing cations and anions solvated in isopropyl alcohol (or other suitable solvents) to intercalate into the precursor structure? • What is the atomistic structure of the intermediate compound? • What is the mechanism and kinetics of formation of the intermediate compound from the intercalated ions? • What is the driving force pushing the intermediate compound to transform into the corresponding perovskite? • What is the mechanism and kinetics of the transformation of the intermediate compound into the corresponding perovskite? • Overall, what is the mechanism and energetics of the entire intercalation and reorganization step of the sequential deposition process? What is the rate limiting step? How do these quantities depend on the type of halide, and the associated precursor? How do they depend on the nature of the cation? The ADRENALINE project will have a significant impact on the fabrication of perovskite devices; the knowledge of the energetics and kinetics of the various stage of the intercalation and reorganization step of the overall formation reaction, the identification of the role of the solvent will allow improving the deposition mechanism, thus contributing to further enhance the efficiency and stability of lead halide perovskites solar cells.
TRICEPS – TRansport ImpliCations of Electron-Phonon Scattering
Project Title: TRICEPS – TRansport ImpliCations of Electron-Phonon Scattering
Project Leader: Feliciano Giustino
Resource Awarded: 25.5 million core hours on MareNostrum
Martin Schlipf University of Oxford UK , Samuel Ponce University of Oxford UK , George Volonakis University of Oxford UK, Lorenzo Paulatto Sorbonne Universite FR , Wenbin Li University of Oxford UK , Weng Hong Sio University of Oxford UK
Improving the performance of optoelectronic devices can reduce the carbon footprint of society in the form of renewable energy source (photovoltaics) and more efficient light sources (LEDs) or transistors. Traditionally, optoelectronic devices are build from ultra pure crystalline semiconductors. The recent discovery of halide perovkistes offers a new route based on solution processing which delivers excellent optoelectronic properties. Despite the success of these materials, there are still fundamental open questions on the origin of their high performance. In TRICEPS, we will unravel the atomistic origin of the transport properties, and enable the engineering of more efficient materials in the future. Fundamentally, optoelectronic devices convert electric charges into light or vice versa. An important criterion limiting the efficiency of these materials is the energy loss that the electric charges experience before they emit light. In our previous PRACE project CATNIP, we revealed the fundamental mechanism responsible for energy losses immediately after photoexcitation. Recent experiments suggest that at longer time scales the heating due to the energy loss cannot be ignored. In TRICEPS, we will therefore extend our methodology to include these heating effects. Furthermore, we will expand the scope of the investigated materials to tackle their poor stability. To this end, we will employ numerical methods based on the first principles of quantum mechanics, which offer a complementary viewpoint to experiments as they do not require empirical input. Within TRICEPS, we will develop an atomistic view on how the vibrating crystal structure dissipates heat to understand on which time scale the energy losses heat the materials. Finally, we will combine the calculation of electric and thermal transport to study the out-of-equilibrium processes where any electric charges lose their energy after initial photoexcitation. Our calculations will open new routes to design novel materials, and will allow us to engage with our collaborators at Oxford to translate computational predictions into real materials. The results of this project will be disseminated in high-profile scientific journals, as well as through our group’s GitHub repository that contains the calculated data and workflows to reproduce our results.
QMC-straintronics – Tuning electronic and optical properties of 2D materials via strain
Project Title: QMC-straintronics – Tuning electronic and optical properties of 2D materials via strain
Project Leader: Jaroslav Fabian
Resource Awarded: 36 million core hours on MareNostrum
Ivan Stich Slovak Academy of Sciences SK , Jan Brndiar Slovak Academy of Sciences SK , Kamil Tokar Slovak Academy of Sciences SK , Robert Turansky Slovak Academy of Sciences SK , Tobias Frank Universität Regensburg DE
The 2D-straintronic project will employ the best available benchmark electronic structure method, the quantum Monte Carlo (QMC), to study the effects of tuning the band gap of free-standing 2-dimensional (2D) materials by applied strain. Based on our expertise in investigating unstrained phosphorene, two distinct single-layer direct-gap semiconductors are proposed for the study of strained systems: phosphorene and MoS2. Numerous experimental studies on 2D MoS2 suggest strainability of up to 10% and gap tunability in the range from 1.8 to 0 eV; large strainability/gap tunability is expected also for phosphorene. The unprecedented tunability opens up a new window on functionalization of 2D straintronic materials, which is important for optical and electronic applications. However, due to the effect of dielectric environments (substrates and capping) the experimentally measured gaps show spreads by about 1 eV as do the gaps calculated by mainstream DFT/GW methods. We propose quantum Monte Carlo (QMC) as the method of choice to provide the least biased values of the band gap of 2D materials strained or in equilibrium. The project outcome will be benchmark values of the band gap under strain. We also argue that QMC in addition to providing band gaps at least an order of magnitude more accurate than the alternative methods, due to the stochastic nature of the algorithm makes an almost perfect use of the largest supercomputer power making QMC an ideal complement to interpret and understand the straintronic experiments.
Supported two-dimensional transition metal dichalcogenides under ion irradiation
Project Title: Supported two-dimensional transition metal dichalcogenides under ion irradiation
Project Leader: Mahdi Ghorbani Asl
Resource Awarded: 37.1 million core hours on Hazel Hen
Arkady Krasheninnikov Helmholtz-Zentrum Dresden-Rossendorf DE , Silvan Kretschmer Helmholtz-Zentrum Dresden-Rossendorf DE , Sadegh Ghaderzadeh Helmholtz-Zentrum Dresden-Rossendorf DE , Thomas Joseph Helmholtz-Zentrum Dresden-Rossendorf DE
Ion irradiation techniques have been extensively used for material modification, post-synthesis engineering and imaging purposes. Although the response of bulk targets to ion irradiation has been studied at length, including simulations, much less is known about the effects of ion bombardment on layered materials. In particular, the details of damage creation in supported 2D materials are not fully understood, while the majority of experiments have been carried out for 2D targets deposited on substrates. Layered transition metal dichalcogenides (TMDs) have shown spectacular physical properties which make them intriguing 2D materials for various nanoelectronic and optoelectronic applications. In this project, we assess the effects of ion irradiation on supported TMDs by using classical molecular dynamics simulations. We are particularly interested in modeling of high-dose radiation damage which requires prolonged timescale simulations. We characterize the types and assess the abundance of point defects in our structures. The evolution of atomic structure and mobility of defects at different annealing temperatures are studied. By understanding the detailed mechanisms of defect production, we investigate the possibility for nanopatterning in TMDs. The outcome of the project will help to understand the fundamental physical mechanisms underlying ion irradiation of low-dimensional materials and finding optimum parameters for a controlled defect production. This information can be useful for designing experimental setups for defect engineering of 2D nanostructures with optimized properties.
MESP – Modelling Ensembles of Semiconducting Polymers
Project Title: MESP – Modelling Ensembles of Semiconducting Polymers
Project Leader: Alessandro Troisi
Resource Awarded: 35 million core hours on Marconi – KNL
Paola Carbone University of Manchester UK , Javier Burgos University of Liverpool UK , Maryam Reisjalali University of Liverpool UK
Promising results on the charge mobility displayed by conjugated polymers have generated a great interest in these materials. They have the potential to replace inorganic semiconductor in many applications (large scale devices, photovoltaics, sensing) where low manufacturing cost, flexibility or biocompatibility are important. However, the relation between conjugated polymer structure and charge mobility is not yet well understood, as good performance has been reported for both crystalline and amorphous polymers. Computational modelling of (semiconducting) polymers cannot follow the pace of experimental investigations, as the typical time to produce and characterise a new polymer in the laboratory is substantially lower than the times currently employed in producing an accurate computational model for any specific polymer. This project aims at drastically reducing the ordinary times employed in modelling a new polymer by investigating a novel hybrid atomistic and coarse-grained methodology that is particularly suitable for this class of materials where a complex conjugated unit (described atomistically) and long solubilizing side chains (to be described at a coarse grain level) are combined. Studying a large ensemble of materials it is also possible to develop hybrid potentials with a greater degree of transferability across materials containing similar units. With such (more approximated) potentials one can study polymers of similar chemical nature just by assembling previously parametrised coarse-grained units.
Overcoming Bottlenecks in Disordered Kesterite Photovoltaics (KESTPV)
Project Title: Overcoming Bottlenecks in Disordered Kesterite Photovoltaics (KESTPV)
Project Leader: Aron Walsh
Resource Awarded: 25 million core hours on Piz Daint
Photovoltaics constitute one of the main technologies to achieve the targets defined by the EU Energy Roadmap 2050. The major technologies, however, are based on several Critical Raw Materials listed by the European Commission. It is essential to overcome this constraint for the consolidation of independent and secure European photovoltaic technology to meet our greenhouse and energy supply commitments. Kesterites are formed from low toxicity metals, which are abundant in the earth’s crust. For applications in solar energy conversion, material properties of quaternary I2-II-IV-VI4 compounds have been investigated, and a Cu2ZnSn(S,Se)4 solar cell with the maximum efficiency of 12.6 % was made relatively recently. The solar cell can be solution processed at low cost, showing the potential to support a terawatt photovoltaic industry. To foster further research and development and to meet the engineering requirements, deeper knowledge of the physical properties of kesterites including their defects and interfaces is necessary. The aim of this proposal is to harness atomistic simulations to understand the large voltage losses that limit the efficiency of kesterite solar cells to less than 15% sunlight to electricity and find a means to overcome them. We will study (I) the non-radiative deep level recombination process limiting the carrier lifetime, (II) the secondary phase emerging under the off-stoichiometry condition and (III) the electronic and optical properties of kesterite alloys as alternative absorber materials. These calculations, involving first-principles electronic structure and electron-phonon coupling analysis, will be carried out in close collaboration with experiment. The supporting experiments are being performed as part of the EU H2020 STARCELL project (http://www.starcell.eu). A successful collaboration with experiment requires access to large computational resources supported by the PRACE Tier 0 system.
DECONVOLVES – Determining watEr COntact aNgle on eVOLving peroVskitEs Surfaces
Project Title: DECONVOLVES – Determining watEr COntact aNgle on eVOLving peroVskitEs Surfaces
Project Leader: Alessandro Mattoni
Resource Awarded: 33 million core hours on Marconi – KNL
Team Members: Giacomo Giorgi Università di Perugia IT , Alessio Filippetti Università di Cagliari IT , Claudia Caddeo Consiglio Nazionale delle Ricerche IT , David Dell’Angelo Consiglio Nazionale delle Ricerche IT
Understanding the interaction of water with the surface of lead-halide perovskites is the key towards the practical exploitation of these materials in solar cells; the experimental results on water contact angle obtained so far, in absence of a true atomic scale knowledge of the microstructure evolution, appear to be contradictory and time- depending, reporting an apparent hydrophobic behavior for a material that is instead expected to be highly hydrophilic. The objective of the DECONVOLVES project is to determine by parallel-replica large-scale molecular dynamics the contact angle of water on degradable hybrid perovskite surfaces. The classical modeling of the evolving water/perovskite system will be combined to ab initio electronic structure calculations and developed in synergy with experiments, with the aim of establishing a grounded protocol for the study of contact angle on evolving substrates.
DFCS – Disentangling Degrees of Freedom by Computing Susceptibilities for Strongly Correlated Systems
Project Title: DFCS – Disentangling Degrees of Freedom by Computing Susceptibilities for Strongly Correlated Systems
Project Leader: Mark van Schilfgaarde
Resource Awarded: 30 million core hours on SuperMUC
Swagata Acharya King’s College London UK , Dimitar Pashov King’s College London UK , François Jamet King’s College London UK , Cedric Weber King’s College London UK
Strongly correlated electronic systems are playground for multiple energy scales. Novel complex phases emerge here as temperature, pressure, chemical composition and doping are tuned. The complexity further intensifies when their bulk, monolayers, and interfaces behave completely disparately. However, identifying which degree of freedom drives what phase is difficult due to such interplay of several degrees of freedom. There are several advanced experimental methods that attempt to probe one or the other degrees of freedom to answer this enigmatic question. In the same spirit, theoretically it is possible to disentangle these degrees of freedom by computing susceptibilities in different channels. But before that it is mandatory that we have a very high level ab-initio theory that gets the single particle description right. With our recently developed three-tier ab-initio QSGW+DMFT+BSE technique we gain insights into the single and two-particle spectral properties of some such very complex and highly interesting correlated electron systems like never before. However, the prediction and analysis of novel symmetry broken phases by computing susceptibilities have been significantly hampered till this point, due to limited computational resources.
SENT_TO_NY – Study of CovEred and functioNalized Tio2 nanostrucTures: the role Of maNy-bodY
Project Title: SENT_TO_NY – Study of CovEred and functioNalized Tio2 nanostrucTures: the role Of maNy-bodY
Project Leader: Ivan Marri
Resource Awarded: 30 million core hours on Marconi – KNL
Andrea Ferretti Istituto Nanoscienze IT , Daniele Varsano Istituto Nanoscienze IT , Giovanni Cantele CNR-SPIN IT , Letizia Chiodo Università Campus Bio-Medico IT , Elisa Molinari Università degli studi di Modena e Reggio Emilia IT
Titanium dioxide (TiO2) is emerged in the last two decades as one of the most important materials for photocatalytic and photovoltaic applications. For these motives TiO2 have been extensively investigated both theoretically and experimentally. In this project we aim at investigating electronic and optical properties of TiO2 nanosystems within the framework of the Many Body Perturbation Theory (MBPT), that is using one of the most powerful theoretical method to gain direct access to electronic and optical properties of solids. Excited states calculations are normally performed for bare and pristine systems, neglecting fundamental effects induced by passivation in the nanomaterial response. In this proposal we will apply GW and BSE@GW techniques to investigate two fundamental aspects of TiO2 nanostructures, that are the role played by molecular functionalization and by the passivation. We will consider TiO2 nanocrystals (NCs) and nanowires (NWs) of different phase and coverage. Moreover we will focus on functionalization, by analysing two prototypical photocatalytic and photovoltaic systems, methanol and catechol adsorbed on TiO2 NCs. We will investigate how passivation modify properties of these systems and we will identify the best configuration for both photocatalytic and photovoltaic applications. Remarkably, due to the complexity of the calculations, these important organic-oxide interfaces have been studied so far only considering TiO2 surface as substrate, despite the most important technological applications are based on engineered TiO2 nanostructures. The project SENT_TO_NY will provide a complete characterization of electronic levels and optical excitations in such hybrid systems, filling an important gap in metal-oxide nanostructures knowledge.
Platinum Reaction Observed by Monte-Carlo for Insight into Sustainable Energy. (PROMISE)
Project Title: Platinum Reaction Observed by Monte-Carlo for Insight into Sustainable Energy. (PROMISE)
Project Leader: Philip Hoggan
Resource Awarded: 41.6 million core hours on Joliot Curie – KNL
Neil Drummond Lancaster University UK , Tapio Rantala Tampere University of Technology FI , Christine Robert-Goumet Institut Pascal FR , Guillaume Monier Institut Pascal FR , Rejesh Sharma Institut Pascal FR
Production runs on this project would efficiently use any of the supercomputers in this PRACE call. This project can only run in production on Tier-0 machines since it employs Quantum Monte Carlo (QMC) methods for a heterogeneous catalyst. The solid is platinum, with an exposed compact Pt(111) face. The reaction investigated is selective production of hydrogen as a source of sustainable energy (clean fuel). The water-gas shift equilibrium is shifted towards the hydrogen product by co-adsorbing water and carbon-monoxide on the Pt(111) face. This reaction can only be investigated theoretically by QMC, because it involves drastic changes in electron correlation and activation barriers are quite low and required to within 1 kcal/mol (0.043 eV). The QMC methodology for this system is now mature and two mechanisms: one limited by water dissociation and the other limited by concerted formation of adsorbed formate species have been the topic of our preliminary publication on this topic (ACS Symp. 1234 (2016).p77) . Now, with vastly improved methodology the proposal is ready for production on a scale only available through PRACE (see some software improvements, below).
Earth System Sciences (2)
eFRAGMENT1 – eFRontiers in dust minerAloGical coMposition and its Effects upoN climaTe, phase 1
Project Title: eFRAGMENT1 – eFRontiers in dust minerAloGical coMposition and its Effects upoN climaTe, phase 1
Project Leader: Carlos Pérez García-Pando
Resource Awarded: 34 million core hours on MareNostrum
Soil dust aerosols are mixtures of different minerals, whose relative abundances, particle size distribution (PSD), shape, surface topography and mixing state influence their effect upon climate. However, Earth System and Chemical Transport Models typically assume that dust aerosols have a globally uniform composition, neglecting the known regional variations in the mineralogy of the sources. The representation of the global dust mineralogy is hindered by our limited knowledge of the global soil mineral content and our incomplete understanding of the emitted dust PSD in terms of its constituent minerals that results from the fragmentation of soil aggregates during wind erosion. The emitted PSD affects the duration of particle transport and thus each mineral’s global distribution, along with its specific effect upon climate. Coincident observations of the emitted dust and soil PSD are scarce and do not characterize the mineralogy. In addition, the existing theoretical paradigms disagree fundamentally on multiple aspects. A recently granted ERC Consolidator Grant called FRAGMENT (FRontiers in dust minerAloGical coMposition and its Effects upoN climaTe) was designed to fill this gap. FRAGMENT will use field campaigns, new theory, remote spectroscopy and modeling to understand and constrain the global mineralogical composition of dust along with its effects upon climate. In this context, eFRAGMENT1 is designed to tackle the modelling and HPC related activities of FRAGMENT during the 1st year of the project. The experiments proposed consist of 1) a dust data assimilation run involving 10-year long simulations of 20-member ensembles and 2) a global dust mineralogical composition run involving 10-year long simulations of 60-member ensembles, both in the period 2007-2016. Such experiments would not be possible without appropriate access to tier-0 computing resources and the associated support offered by PRACE. A total of 33 Mcore-hours are requested in this proposal. The project will be carried out by members of the Earth Sciences Department of the Barcelona Supercomputing Center with the MONARCH model developed in the Department.
SODust – Scattering simulations of Oriented large Dust particles
Project Title: SODust – Scattering simulations of Oriented large Dust particles
Project Leader: Vassilis Amiridis
Resource Awarded: 35 million core hours on MareNostrum
Nikolaos Kontos, RAYMETRICS SA, GR
The Scattering simulations of Oriented large Dust particles (SODust) project aims at quantifying the scattering properties of large, irregularly-shaped desert dust particles, oriented due to triboelectrification processes within the dust cloud. The electrification of desert dust plumes and its effect on climate is the subject of the new ERC Consolidator Grant «D-TECT» and the SODust simulations will be central for realising its scopes. The scattering simulations will be performed with the Discrete Dipole Approximation (DDA) method, considering irregular-shaped particles with different aspect ratios. DDA has been proven capable of accurately reproducing the scattering of irregular-shaped dust particles with radius of only up to ~1.6 μm due to the high computational cost for larger sizes. However, desert dust particle sizes are much larger, while their scattering properties for realistic shapes are completely missing for the size range between 1.6 and 16 μm. The required computational cost for these calculations is quite high, of the order of hundreds of millions CPU core-hours. SODust aims at bridging this existing gap by extending the DDA scattering calculations to larger particles (up to ~5 μm) using the extensive computer resources of the PRACE RI. The SODust unique scattering database is expected to be extended even further in the future and it will be made available to the scientific community to be used by all remote sensing applications considering complex particle shapes. It is indisputable that such simulations will provide unprecedented knowledge, affecting many research fields ranging from Earth System science to Astronomy and Astrophysics.
PROSPeCt – Pore-Resolving Simulation of turbulent flow in Porous Channels
Project Title: PROSPeCt – Pore-Resolving Simulation of turbulent flow in Porous Channels
Project Leader: Sergio Pirozzoli
Resource Awarded: 40 million core hours on Marconi – KNL
Paolo Orlandi Sapienza, University of Rome IT , Davide Modesti University of Melbourne AU , Antonio Memmolo Sapienza, University of Rome IT , Simone Di Giorgio Sapienza, University of Rome IT
We carry out pore-resolving direct numerical simulations (DNS) of turbulent channel flow over a permeable bed, using an incompressible Navier-Stokes solver relying on immersed-boundary representation of complex geometries. Unlike all previous studies, the flow in the porous material is directly resolved accounting for its full geometrical complexity, rather than modeled through an ad-hoc set of equations. Additional physical insight will be gained through higher fidelity coupling of the flow within the channel and within the porous bed, hence avoiding the prescription of fictitious jump conditions for the tangential stresses which are known to be affected by large uncertainty. Numerical simulations will be carried out for several types of porous beds, including: i) foams with random spherical pores; ii) granular media with spherical pores, either self-intersecting or not. Different values of the medium porosity will be considered, as well as different values of the pore size in the range from 0.05 h to 0.2 h (where h is the channel half-height). Numerical simulations with assigned p.d.f. of the pore/void radius will also be carried out to mimic realistic flow cases. Access to PRACE resources will allow to compile a comprehensive database covering 44 geometries, at shear Reynolds number up to Reτ = 1000, thus significantly widening the current DNS envelope for this class of flows, so far limited to the canonical value of 180.
CryoFARE – Cryogenic Flame Acoustic Response
Project Title: CryoFARE – Cryogenic Flame Acoustic Response
Project Leader: Thierry POINSOT
Resource Awarded: 29.2 million core hours on Joliot Curie – SKL
Gabriel Staffellbach CERFACS FR , Charlelie Laurent CERFACS FR , Simon Blanchard CERFACS FR
Transcritical conditions correspond to the thermodynamic state of a fluid subjected to a pressure exceeding its own critical pressure, and a temperature below its critical value. These extreme thermodynamic conditions are found for instance in high performance liquid rocket engines (LREs), where oxidizer and fuel are injected as dense cryogenic fluids through coaxial injectors. These real-gas combustion regimes are expected to become more and more prevalent in future LREs, due to a growing need for reusability and increased operability. Both SPACE X and ARIANEGROUP are working on methane/oxygen engines which will operate in these transcritical conditions. However, developments of innovative LREs are historically known to be plagued by thermoacoustic instabilities, that is highly unsteady fluctuations in the combustion chamber resulting from a two-way coupling between the acoustic field and the flame dynamics. The most dramatic outcome of a thermoacoustic instability is the destruction of the combustion chamber and the loss of the whole space launcher. The current state-of-the-art knowledge concerning transcritical flame dynamics is rather limited, due to the inherent difficulty in reproducing such extreme thermodynamic conditions both experimentally and numerically. The proposed work therefore aims at improving our fundamental understanding of transcritical flame dynamics under acoustic perturbations. A doubly-transcritical LO2/LCH4 flame in the geometry of the academic laboratory test rig Mascotte (operated by ONERA, France) will be computed by means of Large Eddy Simulation (LES), with non-ideal gas effects modelled by the Soave-Redlich-Kwong (SRK) equation of state. In a first stage, the structure of the stable flame will be studied. As the flame root is expected to have first order effects on the overall flame acoustic response, the study will focus on the mechanisms responsible for the flame stabilization at the coaxial injector lip. In particular, flame-wall interactions will be accounted for by using a realistic isothermal boundary condition extracted from a steady-state temperature field in the coaxial injector lip, as well as a detailed kinetic scheme valid for CH4/O2 combustion in high pressure conditions. Then, the stable LO2/LCH4 flame will be subjected to acoustic forcing over a wide range of frequencies, in order to partially build the classical Flame Transfer Function (FTF), which is the backbone for all studies of thermoacoustic instability. Most importantly, the proposed work will endeavour to compare results obtained with existing theoretical models of non-premixed flame FTF, and to extend these models to more realistic real-gas conditions by providing a fundamental understanding of doubly-transcritical flame response to acoustic perturbations. The size of the simulation to perform, the inclusion of complex chemistry and the coupling with heat transfer lead to a very large CPU effort which only a PRACE allocation can allow.
TURB-ROT – Inverse and direct cascades in rotating turbulent flows
Project Title: TURB-ROT – Inverse and direct cascades in rotating turbulent flows
Project Leader: Luca Biferale
Resource Awarded: 60 million core hours on Marconi – KNL
Fabio Bonaccorso University of Rome Tor Vergata IT , Michele Buzzicotti University of Rome Tor Vergata IT , Patricio Clark Di Leoni University of Rome Tor Vergata IT
Turbulent flows under strong rotation exhibit fascinating phenomena which can be appreciated in the might of convective storms, in the awe inspiring swirls in Jupiter’s Big Red Spot, in the beautiful structures seen in experiments and in many other instances. Despite of the huge number of theoretical, numerical and experimental investigations, it is not yet comprehended what are the main dynamical and statistical mechanisms leading to the transition from a quasi-isotropic evolution to a strongly-anisotropic behaviour with elongated structures, as empirically observed when the rotation rate is increased. Rotating turbulence is the result of the intricate interactions between quasi 2D slow motions with 3D fast wave propagations, coupled to strong small-scales turbulent fluctuations. As a result, rotating turbulence develops both the formation of strong columnar cyclonic large-scale condensates and small-scales highly intermittent and non-Gaussian fluctuations, the outcomes of a split-energy cascade both forward and backward. The need to resolve two range of scales, larger and smaller than the injection scale, leads to extremely demanding resources for Direct Numerical Simulations (DNS) approaches. In this project, we aim to study the transition from a forward to a split-energy cascade in rotating turbulence at changing two different control parameters. For fixed aspect ratio, we will perform a refined scanning of the rotation intensity close to the critical one in order to make a definitive statement whether there exists a smooth or a discontinuous jump in the energy flux and clarify the nature of this out-of-equilibrium (phase) transition. Second, for a few selected rotation intensities, we intend to push the vertical size towards the “infinite’ volume limit, in order to understand whether the Wave-Turbulence prediction of a vanishing inverse energy cascade is correct: a theoretical and applied problem open for more than 20 years and never clarified yet. Our results will have impact in problems broader than the case of flows under rotation, as Wave Turbulence is a basic approach for many other configurations, such as internal gravity waves in stratified flows, surface waves, acoustic waves, Alfven waves in conducting flows, to cite just a few of the most important ones.
LiLiPlaTE – Light Limited Plankton Population Dynamics in a Turbulent Environment
Project Title: LiLiPlaTE – Light Limited Plankton Population Dynamics in a Turbulent Environment
Project Leader: Filippo De Lillo
Resource Awarded: 15 million core hours on Marconi – Broadwell
Matteo Borgnino Università di Torino IT , Francesco Toselli Università di Torino IT , Guido Boffetta Università di Torino IT
This project will investigate the population dynamics of phytoplankton in turbulence in conditions when light can be considered as the limiting factor for growth. We will study how the classic picture (due to Sverdrup) of a critical depth of the mixed layer and its extensions and generalizations are modified when the intense fluctuations typical of turbulent flows are taken into consideration. Extensive, high resolution numerical simulations will be performed to clarify how shading, growth and cell buoyancy interact with turbulence to determine the ability of a phytoplankton population to thrive and outperform competing species.
P-TURB – Prandtl number effect on TUrbulent Rayleigh-Bénard convection
Project Title: P-TURB – Prandtl number effect on TUrbulent Rayleigh-Bénard convection
Project Leader: Enrico Stalio
Resource Awarded: 65 million core hours on Marconi – KNL
Sergio Chibbaro University Pierre et Marie Curie FR , Andrea Cimarelli Università Politecnica delle Marche IT , Andrea Fregni Università degli Studi di Modena e Reggio Emilia IT , Paolo Gualtieri Sapienza Università di Roma IT , Francesco Battista Sapienza Università di Roma IT
Thermally driven turbulence is a phenomenon of relevance in various areas of science and technology, as it affects geophysical and astrophysical systems and occurs also in industrial applications. Its importance for atmospheric flows involves both small length and timescales for weather predictions and large scales for climate calculations. Rayleigh-Bénard convection, the buoyancy-driven flow in a fluid layer heated from below and cooled from above involves the main thermally driven turbulence features and is a classical problem in fluid dynamics. It is also a rich multi-physics problem with fundamental outcomes in heat transfer and important fluid mechanics implications where the concomitant action of turbulence, plumes and large scale circulation strongly influences the transport properties of the medium. This does not include only heat transfer characteristics but more generally the transport of substances or particles dispersed in the fluid. The aim of this project is to perform three numerical simulation of the Rayleigh-Bénard convection, representing the same physical phenomenon occurring in three very different fluids of interest for applications: liquid metals, air and water. Numerical simulations proposed are as close to reality as possible, because no turbulence model is involved in the calculations and spatial as well as temporal discretisation grids are so fine as to be able to catch all the spatial and temporal scales actually involved. The three simulations will help reserchers to understand important physical processes related to the statistical properties of turbulent fluctuations and buoyancy-induced turbulence, These include the large-scale dynamics of the plumes and its interaction with small-scale fluctuations, the turbulent cascade, and the instantaneous paths of mechanical and thermal energy in the flow.
Direct Numerical Simulations of Transient Turbulent Autoigniting Jets
Project Title: Direct Numerical Simulations of Transient Turbulent Autoigniting Jets
Project Leader: Christos Frouzakis
Resource Awarded: 64.6 million core hours on Marconi – KNL
Miriam Rabacal Swiss Federal Insitute of Technology Zurich CH , George Giannakopoulos Swiss Federal Insitute of Technology Zurich CH
The transient nature of starting jets has a strong impact on the entrainment of ambient fluid in the near field, enhancing significantly the mixing of the jet with the ambient fluid. The increasing fuel/air interface and the enhanced mixing facilitate the initiation of the chemical reactions that can lead to autoignition, typically at multiple locations, and propagation of flames from the ignition kernels to consume the injected fuel. This subject is important for the development of innovative injection and ignition systems for internal combustion engines (ICEs) for the use of low-carbon alternative fuels towards the process of decarbonisation of the transportation sector, and heat and power production based in ICE technology. The objective of the proposed work is to provide new insights into the transient flow, mixing and turbulence-chemistry interactions in transient turbulent autoigniting jets at elevated pressures. Direct Numerical Simulations will be used to provide the complete description of the flow and chemistry, and generate much needed high resolution datasets for the development of better models for turbulent autoignition. The DNS code developed at the Aerothermochemistry and Combustion Systems Laboratory (LAV) for the direct numerical simulation of low Mach number reactive flows, based on the open source spectral element solver Nek5000, will be used to study two cases of relevance to compression ignition engines with direct gas injection: while in conventional engines combustion occurs during the jet acceleration or constant injection rate phases, the low temperature strategies employed in modern engines result in delayed autoignition and combustion during the deceleration phase. The proposed simulations will enhance the in-depth understanding of the role of the flow dynamics during the different phases of injection on the jet penetration and angle, on the transient rate of mixing and entrainment, and on the processes and local conditions leading to ignition kernel formation and appearance and propagation of the ensuing flames.
TurEmu – The physics of (turbulent) emulsions
Project Title: TurEmu – The physics of (turbulent) emulsions
Project Leader: Federico Toschi
Resource Awarded: 40 million core hours on Marconi – KNL
Gianluca Di Staso Eindhoven University of Technology NL , Abheeti Goyal Eindhoven University of Technology NL , Pinaki Kumar Eindhoven University of Technology NL , Xiao Xue Eindhoven University of Technology NL , Ivan Girotto International Centre for Theoretical Physics IT , Sebastiano Fabio Schifano Università degli Studi di Ferrara IT , Roberto Benzi Università di Roma Tor Vergata IT , Xiaowen Shan Southern University of Science and Technology (SUSTech) CN
Multi-component fluids are extremely common in industrial as well as natural processes and, in particular, multi-component emulsions are an important ingredient of many foods and cosmetics. Emulsions are fascinating systems from the point of view of fundamental science and in partic-ular for fluid dynamic phenomenology. Ordinary substances, such as mayonnaise or ketchup, are good examples of complex fluids that can display a surprisingly rich phenomenology, typi-cal of soft-glassy materials. These complex fluids can behave either as solids (below the yield stress) or as fluids (above the yield stress). Above the yield stress, when these complex fluids flow, their rheological properties can be complex and non-Newtonian; below the yield stress, the statistical physics of localized topological rearrangements (plastic events) give rise to ava-lanches that closely resembles the behaviour of earthquakes. In this project we will employ highly optimized computational codes, based on the multicomponent Lattice Boltzmann model (LBM), to explore the physics of complex fluid emulsions: from their production, via turbulent stirring, to their (statistical) behaviour under flowing as well as resting conditions. The large produced database will allow to explore the process of turbulent emulsification in great details, including individual breakup and coalescence processes between emulsion droplets and the temporal changes in the macroscopic rheological properties of the complex fluid, as a function of the varying internal emulsion structure and of the build-up of yield stress. Once emulsions are formed, we will study their temporal dynamics both below as well as above the yield stress. Below yield our simulations will allow to explore, for the first time ever in 3d, the statistical physics of avalanches, thus paving the way to a deeper understanding of avalanche physics, including the causal-connection between different plastic events, and the relation between our model system and the physics of earthquakes aftershocks. Additionally, our study may allow to develop phenomenological macroscopic models for the behaviour of complex fluids, potentially of great relevance for the simulations of emulsification in industrial processes.
LESRudProp – LES of the Interaction Between a Rudder and the Wake of a Propeller
Project Title: LESRudProp – LES of the Interaction Between a Rudder and the Wake of a Propeller
Project Leader: Riccardo Broglia
Resource Awarded: 31.5 million core hours on Marconi – KNL
Antonio Posa CNR-INSEAN IT
Interaction of vortices with obstacles or bodies, as rudders or wings, is a common scenario in several practical flow problems, especially in naval and aeronautical fields. However, accurate numerical studies on this topic are very limited. This is mainly due to the demanding simulation of coherent structures, as well as boundary layers, at Reynolds numbers typical of engineering problems, requiring huge computational resources and highly accurate numerical tools with optimal conservation properties. For instance, conventional – albeit still challenging – numerical techniques, as those resolving the Reynolds-averaged Navier-Stokes (RANS) equations, demonstrated to be unsuitable to simulate properly coherent structures and to handle separation phenomena, as those tied to boundary layer/vortex interaction, due to their time-averaged approach and important modelling assumption on turbulence. Detached-eddy simulation (DES) is obviously a better alternative, since vortices are directly resolved, with no turbulence modelling, but boundary layers are tackled via a RANS methodology, with some significant drawbacks on the accuracy of the overall approach. Here our target is to simulate via Large Eddy Simulation (LES) the behavior of tip and hub vortices shed by a marine propeller in presence of a downstream rudder (hydrofoil) for three different load conditions (rotational speeds). This will allow us to assess how tip and hub vortices, featuring variable intensity across load conditions, interact with the downstream obstacle and its boundary layer and the way such interaction affects their stability and footprint on turbulence statistics. In a similar way the effect on the boundary layer over the rudder can be analyzed. Although we are going to consider a naval hydrodynamic problem, we expect that the impact of this study will go beyond the particular application, based on the small number of high-fidelity computations in this class of flow problems. Resolution requirements of LES, in both space and time, are very demanding in the present class of flows, since all important energy-carrying structures need to be resolved. Such requirements can be met only on large supercomputers and leadership computing facilities. The solver we are going to adopt in this study, coupling LES and the Immersed-Boundary (IB) method, has been already validated on several practical flow problems, involving also rotating machinery applications, turbines and propellers, demonstrating accuracy, performance and stability. It was tested on many parallel clusters in the framework of several HPC grants, including also Marconi-KNL at CINECA, which is the machine where we are planning to run the simulations of interest of the present proposal. Therefore, both computational tool and setup of the simulations are ready for production runs.
MALEDRAG – Machine Learning Optimisation for Drag Reduction in a Turbulent Boundary Layer
Project Title: MALEDRAG – Machine Learning Optimisation for Drag Reduction in a Turbulent Boundary Layer
Project Leader: Sylvain Laizet
Resource Awarded: 30 million core hours on Hazel Hen
Andrew Wynn Imperial College London UK , Francesco Montomolli Imperial College London UK , Charles Moulinec STFC Daresbury Lab UK , Omar Mahfoze Imperial College London UK , Athanasios Giannenas Imperial College London UK
The need to reduce the skin-friction drag of aerodynamic vehicles is of paramount importance. Just a 3% reduction in the skin-friction of a long-range commercial aircraft would save £1.2m in jet fuel per year per aircraft and prevent the annual release of 3,000 tonnes of carbon dioxide. However, achieving net skin-friction drag reduction of wall-turbulent flows is notoriously difficult, even at low Reynolds numbers. Despite decades of research, frustrating performance penalties have prevented the development of any functional and economical system for full-scale applications. Any practical turbulent skin friction reduction control strategy must be able to rapidly and autonomously optimise the aerodynamic surface with minimal power input, have far-reaching effects downstream of control, and must be reactive to changes in flow speed. Enabling such a strategy would mark a new era in turbulence control and is a major objective of the MALEDRAG project. MALEDRAG will develop a new type of machine learning paradigm, which has the ability to optimise different control inputs rapidly, autonomously, adaptively and simultaneously with only a few simulations. This new capability will be exploited to minimise the turbulent surface friction for a flat plate with minimal power input. The propose control strategy is based on wall blowing, which is a very effective method to reduce shear stress and skin-friction drag (up to 70% drag-reduction have been reported in the literature). However, the energy expenditure can be high, leading to net energy saving as low as 5%. This new approach will open up the opportunity to rapidly optimise all existing types of actuation based control strategies as well as discover new types of actuation techniques and drag reduction mechanisms to generate significant levels of turbulent skin friction reduction. This ambitious goal will be achieved by developing and implementing a novel machine learning paradigm based on Bayesian optimisation, a global optimisation technique that only requires few simulations to be performed, and is yet to be exploited in the field of fluid dynamics.
DNS4ICE – DNS for modelling highly diluted and multi-fuel combustion in internal combustion engines
Project Title: DNS4ICE – DNS for modelling highly diluted and multi-fuel combustion in internal combustion engines
Project Leader: Karine Truffin
Resource Awarded: 16.8 million core hours on JUWELS
Edouard Suillaud IFP Energies nouvelles FR , Maxime Tarot IFP Energies nouvelles FR , Olivier Colin IFP Energies nouvelles FR , Stephane Jay IFP Energies nouvelles FR , Karine Truffin IFP Energies nouvelles FR
Future hybrid vehicles will essentially be powered by spark-ignition engines (SIE), whatever the hybridisation level. Designing and calibrating SIE to achieve optimal performance in the context of hybridization and real driving conditions usage represents a major scientific and technological challenge. These technological solutions could not be targeted without mastering flow, mixing and combustion down to the individual engine cycle, for which novel simulation approaches such as Large-Eddy Simulation offer the possibility to study the behavior of instantaneous engine cycles. This is of critical importance for improving the engine performance and emissions in highly non-stationary situations such as fast transients, cold start, extreme operating conditions or the restarts of powertrains when coupling electrical/thermal engines. One technology favoured today by engine manufacturers for SIE is downsizing, which consists in reducing the displacement and increasing the specific power by using a turbocharger. This technology is however limited in practice due to an increased occurrence of abnormal combustions which lead to using sub-optimal spark timings. A key measure for limiting the occurrence of abnormal combustion is to increase the EGR (Exhaust Gas Recirculation) rate up to 30% but this leads to larger cycle to cycle variability and decreased heat release rates. In order to reach such high EGR rates, complex strategies have to be developed (aerodynamics, spark ignition, injection targeting and timing, chamber geometry etc…) the design and optimization of which increasingly rely on Computational Fluid Dynamics (CFD). Dual Fuel Combustion (DFC) is seen as another promising concept to combine advantages of spark-ignited and compression ignited engines. Its principle is to inject a high cetane number fuel to initiate combustion of the premixed charge (gas/air or gasoline/air) initially admitted. The objective of DNS4ICE is to improve the understanding of the phenomena during DFC and highly diluted combustion and provide appropriate combustion models. The usage of dedicated Direct Numerical Simulations (DNS) on academic configurations under the extreme conditions found in future engines will provide detailed local flame statistics for orienting and supporting the combustion model developments.
ESECELS – Electromagnetic Simulations of Extremely Complex and Electrically Large Structures
Project Title: ESECELS – Electromagnetic Simulations of Extremely Complex and Electrically Large Structures
Project Leader: Franco Moglie
Resource Awarded: 30 million core hours on Marconi – KNL
Valter Mariani Primiani Universita` Politecnica delle Marche IT , Luca Bastianelli Universita` Politecnica delle Marche IT , Salvador Gonzalez Garcia Universidad de Granada ES , Amelia Rubio Bretones Universidad de Granada ES , Rafael Gomez Martin Universidad de Granada ES , Mario Fernandez Pantoja Universidad de Granada ES , Luis Manuel Diaz Angulo Universidad de Granada ES , Miguel Ruiz Cabello Nuñez Universidad de Granada ES , Gabriele Gradoni University of Nottingham UK , Sendy Phang University of Nottingham UK , David Thomas University of Nottingham UK , Stephen Greedy University of Nottingham UK , Kristof Cools University of Nottingham UK , Chris Smartt University of Nottingham UK , Hayan Nasser University of Nottingham UK , John F. Dawson University of York UK , Stuart Porter University of York UK , Ian Flintoft University of York UK , Sam Bourke University of York UK , Damienne Bajon Institut Supérieur de l’Aéronautique et de l’Espace FR
This project brings together researchers in electromagnetic and stochastic computational techniques. Many researchers of the team are involved in the solution of electromagnetic compatibility problems, where incoherent radiators, semi coherent emitters and complex devices are quantified. Usually, the involved geometry is large and may have highly complexity in field pattern. Moreover, chaotic structures are investigated and the results can be obtained only as an ensemble average of simulations by changing geometrical parts or sources. Three dimension simulations of complex sources in complex environments require Tier-0 machines. All the participants of this team have a background in the parallel computation. The group of Ancona developed an FDTD code for the simulation of reverberation chambers; the group of Granada developed “UGRFDTD”, a general purpose (EMC-oriented) state-of-the-art FDTD solver; the group of York developed “Vulture”, an open source FDTD solver for electromagnetic simulations, and the group of Nottingham developed GGITLM a TLM based time domain simulation code. We participated to previous PRACE projects and domestic calls. We will use the same computer code of the previous projects “CSSRC – Complete statistical simulation of reverberation chamber”, approved during the PRACE 7th Regular Call for the year 2013- 2014, “ASOLRC – Advanced simulation of loaded reverberation chambers” approved during the PRACE 9th Regular Call for the year 2014-2015 and “SREDIT – Simulations of Radiated Emissions in Densely Integrated Technologies” approved during the PRACE 13th Regular Call for the year 2016-2017. Our codes are capable to simulate different geometries as set of stirrer angles in the reverberation chamber and complex sources for the propagation of the stochastic noise emissions. The code of Ancona was optimized for the FERMI and Marconi-KNL architectures during all the previous PRACE projects and the code of Granada was optimized for Marconi-KNL architecture during the PRACE 13th Regular project. The Ancona code is mainly divided in three modules: 1) an electromagnetic time domain solver; 2) a fast Fourier transform; 3) a statistical module to obtain the cumulative results. All the modules were previous optimized for high-performance parallel computers using hybrid method (MPI and OpenMP) and they was used successfully in the previous PRACE projects. The availability of a code, that solves the previous three steps in a unique job, makes the simulations very appealing. Moreover, the availability of an optimized simulation code will give the results in short time avoiding long measurement campaigns. The team is a part of the group of the COST Action IC1407, that began in April 2015. In particular, the works of WG1: numerical methods for addressing the propagation of stochastic fields and WG3: equivalent models of noise sources. COST project provides financial support to travel and PRACE project provides computer access, so the two projects are complementary. Not all the participants are mainly working in the EMC topics, they are mathematicians and physicists working on the more general topic of chaotic systems making this project multidisciplinary..
Fundamental Constituents of Matter (4)
PLepNuGam – Radiative leptonic decays of light and heavy pseudoscalar mesons from Lattice QCD
Project Title: PLepNuGam – Radiative leptonic decays of light and heavy pseudoscalar mesons from Lattice QCD
Project Leader: Roberto Frezzotti
Resource Awarded: 45 million core hours on Marconi – KNL
Team Members: Silvano Simula INFN IT , Nazario Tantalo University of Roma “Tor Vergata” IT , Carsten Urbach University of Bonn DE , Ferenc Pittler University of Bonn DE , Bartosz Kostrzewa University of Bonn DE , Marco Garofalo INFN IT , Guido Martinelli University of Rome “La Sapienza” IT , Francesco Sanfilippo INFN IT , Vittorio Lubicz University of Rome III IT , Cecilia Tarantino University of Rome III IT , Davide Giusti University of Rome III IT
The search for inconsistencies of the Standard Model of Particle Physics, which would signal the presence of new physics, requires the determination of several hadronic quantities with a high level of precision of the order of O(1%) or even better. This is the case of the leptonic weak decays of mesons, which allow to investigate one of the fun- damental ingredients of the Standard Model, namely the Cabibbo-Kobayashi-Maskawa matrix describing the weak mixing of quark flavors. In the last few years calculations of hadronic observables from first principles, taking systematic errors under very good control, have been carried out through large-scale QCD simulations on the lattice. In order to make further progress lattice simulations must include both electromagnetic and strong isospin-breaking effects and reach the physical pion mass point . Recently a new approach to deal with the lattice determination of the leptonic weak decay rates has been proposed and successfully applied to the case of kaon and pion decays thanks to the PRACE project 2014112693 “QED corrections to meson decay rates in Lattice QCD”. The aim of the present project is to continue the previous project by extending the lattice calculations of the leptonic weak decay rates from the light to the heavy quark sectors. In doing so the emission of real photons, which is always included in the experimental data, will for the first time be calculated from first principles. Moreover the pure (isosymmetric) QCD part of the computation will be improved thanks to new simulations including ensembles at the physical pion mass point.
PULSAR – Plasma physics of ultra high fields in neutron stars
Project Title: PULSAR – Plasma physics of ultra high fields in neutron stars
Project Leader: Luis Silva
Resource Awarded: 40 million core hours on MareNostrum
Team Members: Marija Vranic Instituto Superior Tecnico PT , Thomas Grismayer Instituto Superior Tecnico PT , Kevin Schoeffler Instituto Superior Tecnico PT , Fabio Cruz Instituto Superior Tecnico PT , Fabrizio Del Gaudio Instituto Superior Tecnico PT
How are electron-positron pair plasmas generated at the polar cap of pulsars? How does the produced pair contribute to feeding the pulsar magnetosphere? What are the expected radiation signatures? These are prominent scientific questions where plasma astrophysics is intimately connected with quantum electrodynamics. This proposal aims to exploit the outstanding computing facilities provided by PRACE to address these compelling challenges by leveraging on the pioneered advances in coupling quantum electrodynamics effects with particle-in-cell simulations. The main scientific route to answer some of these questions consists in understanding the coupling of the pair production region with the global field organization of the pulsar magnetosphere. In particular, it is crucial to assess self-consistently the dependence of the pair multiplicity at the polar cap on the global magnetic field topology and vice versa. This task requires a special computational effort given the separation of the scales involved, namely the macroscopic size of the magnetosphere down to the plasma kinetic and quantum electrodynamics scales. The first goal of this research project is to study the production of pair plasma via cascades that occur in a localized portion of the magnetosphere around the polar cap of pulsars. The second goal is to model the global pulsar magnetosphere, leveraging on the knowledge coming from the first task. The self-consistent study of the interdependence among the plasma dynamics pertinent to such different spatial and temporal scales is difficult and massively parallel kinetic particle-in-cell simulations are critical to establishing this bridge. The unique computational infrastructures provided by PRACE would be critical to explore some of the most exciting fundamental physics questions at the forefront of science identified in this proposal. This project will leverage on the massively parallel, fully relativistic particle-in-cell code OSIRIS, which includes a set of dedicated modules to study the development of pair cascades at the polar cap of pulsars and their global magnetosphere. These are (i) a QED module for photon emission and pair creation, (ii) a merging module for withstanding the exponential increase of macro-particles in the cascades, (iii) a Compton collisional module for the accounting of direct exchange of energy between photons and electron-positron pairs, and (iv) a spherical coordinates module for modeling the global magnetosphere. This research route embraces the spirit and the path outlined by the ERC advanced grant in Pairs, which aims at understanding the dynamics of electron-positron pair plasmas in extreme fields. Many steps were already taken in this direction, and our numerical infrastructure is now ready to be massively employed in studying some of the most puzzling but marvelous questions in plasma astrophysics.
TMDs — Scale dependence of TMDs
Project Title: TMDs — Scale dependence of TMDs
Project Leader: Andreas Schaefer
Resource Awarded: 44 million core hours on SuperMUC
Team Members:Michael Engelhardt New Mexico State University USA , Jeremy Green Deutsches Elektronen-Synchrotron DE , Rajan Gupta Los Alamos National Laboratory USA , Piotr Korcyl Jagiellonian university PL , Andreas Rabenstein Regensburg University DE , Maximillian Schlemmer Regensburg University DE , Christian Zimmermann Regensburg University DE , Alexey Vladimirov Regensburg University DE
This project is part of a major German-US lattice effort to investigate the properties of Transverse Momentum Dependent parton distribution functions (TMDs). The experimental investigation of TMDs is one of the main physics goals of the projected new US accelerator EIC (Electron Ion Collider) the construction of which is pursued jointly by the Brookhaven National Laboratory (BNL) and the Thomas Jefferson National Accelerator Facility. The emphasis on TMDs is motivated by the fact that they link experimental observables to the detailed properties of QCD gauge links and thus to the most characteristic features of QCD. The physics of TMDs is still only partly understood. The open problem which is phenomenologically most relevant is the scale dependence which, in contrast to the case of normal parton distribution functions, has unsuppressed non-perturbative contributions. The aim of the present project is, therefore, to study the scaling of TMDs on the lattice. To do so we want to use the large number of ensembles which were generated by the CLS collaboration, precisely with the aim to study the dependence on the lattice constant a. For these ensembles, open boundary conditions were used, which avoid the problem of diverging topological autocorrelation times for small a. Settling the non-perturbative scaling properties of TMDs would mark a major advance in our understanding of hadron structure.
IONFAST-Coherent interplay description between thermal and energetic ions in magnetically confined plasmas
Project Title: IONFAST-Coherent interplay description between thermal and energetic ions in magnetically confined plasmas
Project Leader: Jeronimo Garcia
Resource Awarded: 26 million core hours on Joliot Curie – SKL
Team Members: Tobias Goerler Max Planck Institute for Plasma Physics DE , David Zarzoso Aix-Marseille Université FR , Samuele Mazzi CEA-Cadarache FR , Maiko Yoshida National Institutes for Quantum and Radiological Science and Technology JP
Microturbulence is one of the primary physical mechanisms which limit energy confinement in magnetically confined plasmas. The investigation of turbulence control and suppression is a critical goal for the optimization and design of future tokamak fusion reactors. The impact of the fast ions generated by means of heating systems has been demonstrated to play a very important role on such plasmas by suppressing heat transport. Those studies were performed solving the Vlasov equation using the gyrokinetic code GENE. However, significant drawbacks in the modelling previously performed could limit such turbulence reduction. In this project, a particular aspect will be addressed: the interaction among electromagnetic fluctuations, thermal ion transport and fast ion transport which is driven by large scale plasma fluctuations of the type of MagnetoHydroDynamics (MHD). This is essential in the magnetically confined plasmas field, as the main goal of such plasmas is to generate enough fusion power and for that purpose the thermal pressure and the fast ion fast ion content must be maximized. Such analysis will help to find a coherent ion distribution function, calculated by means of a Monte-Carlo code, between the external heating and the thermal and fast ion transport obtained by means of kinetic simulations. Usually, such coherence has been neglected in previous studies about fast ion impact on thermal transport including electromagnetic effects.
Mathematics and Computer Sciences (1)
TopBridge – Topology optimization of bridge girders in cable supported bridges
Project Title: TopBridge – Topology optimization of bridge girders in cable supported bridges
Project Leader: Ole Sigmund
Resource Awarded: 15 million core hours on Joliot Curie – SKL
Team Members: Niels Aage Technical University of Denmark DK , Mads Jacob Baandrup COWI A/S DK
Since its introduction in the late 1980’s, the material distribution method known as topology optimization has become an integrated design method in many industries such as automotive, aerospace, architecture, etc. However, until recently the method has been restricted to component design due to the lack of scalable software that allows for full-scale optimization of complete assemblies. This severe bottleneck was circumvented using HPC in a previous PRACE project TopWing, which demonstrated the application of ultra largescale structural optimization to design the support structure of a transatlantic airplane wing and was published in Nature (and awarded with 2016 PRACEDays best industrial presentation). The resulting design provided new insight into for example the layout of ribs and spars that allowed for a reduction in the yearly fuel consumption by 40-200 tonnes per aircraft. In this project, we wish to apply this giga scale optimization methodology to the design of bridge steel box girders in cable supported bridges. Current state-of-the-art design methods, which are now 50+ years old, are based on orthotropic box girders, which in turn are prone to stress hot spots and fatigue issues, and leave almost no room for further optimization. Furthermore, the sheer size of each bridge section means that discretizations in the excess of one billion voxels is required in order to obtain designs with feature sizes in the same order of magnitude allowed by current production methods. This advocates the use of HPC to investigate the bridge girder design problem, and we hypothesize that once a systematic study of the design space has been conducted, we will be able to push next generation bridge design into an era with sustainable and lighter bridge girders and increased lifespans.
Universe Sciences (9)
PlanetsInBed – Embedded planets in wind-driven discs
Project Title: PlanetsInBed – Embedded planets in wind-driven discs
Project Leader: Colin McNally
Resource Awarded: 23.6 million core hours on Joliot Curie – SKL
Team Members: Richard Nelson Queen Mary University of London UK , Oliver Gressel University of Copenhagen DK , Pablo Benitez-Llambay University of Copenhagen DK , Sijme-Jan Paardekooper Queen Mary University of London UK
Driven by the ever and rapidly increasing number of observed, and increasingly well-characterized, exoplanetary systems, understanding how planetary systems form is one of the most active areas in astrophysics. Crucial to this is understanding the protoplanetary disc from which these planets form, and how they interact with it. Forming planets in a disc nec-essarily drives gravitational interactions which cause the semimajor axis to evolve – a process known as migration. Both the formation process, composition, and final positions of planets can be affected by migration. Thus, understanding the mechanisms and dynamics of migra-tion torques is a longstanding central problem in planet formation theory. At the same time, protoplanetary discs have long posed significant theoretical challenges. Although significant accretion is observed onto central stars during the presumed planet-forming era, the mecha-nisms from which this material is driven through the disc have been problematic. The tradi-tional assumption of a weakly-specified turbulence giving rise to a turbulent viscosity and viscous accretion disc has been problematic, due to a lack of a sufficiently vigorous and ge-neric mechanism that can operate in these very low ionization discs. Studies of the magnetic configurations that arise in simulations including the non-ideal MHD effects which are the main stumbling block from viscous models has given rise to a viable alternative, a paradigm of nearly laminar wind-driven accretion discs. However, these discs, having very low turbulence and hence viscosity, give rise to qualitatively new types of migra-tion torques and behaviours. So far, our project team has studied a subset of these in two-dimensional simulations, but the most interesting case, in which a surface accretion flow driven at the top of the disc by the wind, cannot be reduced to a two-dimensional model. Having developed the necessary scientific background, codes, and setups, we will to attack this crucial three dimensional problem with PRACE Tier-0 resources.
TAPES – Turbulence in Astrophysical Plasmas from fluid to Electron Scales
Project Title: TAPES – Turbulence in Astrophysical Plasmas from fluid to Electron Scales
Project Leader: Giovanni Lapenta
Resource Awarded: 35 million core hours on SuperMUC
Team Members: Jorge Amaya KU Leuven BE , Fabio Bacchini KU Leuven BE , Francesco Pucci KU Leuven BE , Diego González KU Leuven BE , Vyacheslav Olshevsky KU Leuven BE , Elisabetta Boella KU Leuven BE , Sergio Servidio University of Calabria IT , Maria Elena Innocenti KU Leuven BE , Francesco Pecora University of Calabria IT
This project is devoted to the study of turbulence in astrophysical plasmas. Turbulence is one of the still not understood problems of classical physics. Similar to what happens for regular fluids on Earth, like air or water, astrophysical plasmas in space are found to behave in a turbulent way, developing vorticose motions over a broad range of scales. Due to its mathematical complexity turbulence has been studied in the last decades by means of numerical simulations. The difficult part of it relies on the need of resolving numerically a broad range of scales that all together constitute the phenomenology of turbulence. In this project we propose to conduct cutting edge numerical simulations of plasma turbulence using a numerical method that overcome several limitations in the description of plasma motions from large to small scales. This method will allow us to address still open questions on this topic with an unprecedented realism of the model and precision of the description. The results of this project will be useful for a better theoretical understanding of astrophysical plasma turbulence and will serve as a guidance for the interpretation of new spacecraft observations of heliospheric plasmas that are already or will be available in the near future.
FROMTON – Particle acceleration in pulsars: FROm the Magnetosphere TO the Nebula
Project Title: FROMTON – Particle acceleration in pulsars: FROm the Magnetosphere TO the Nebula
Project Leader: Benoit Cerutti
Resource Awarded: 26.7 million core hours on Joliot Curie – SKL
Team Members: Guillaume Dubus CNRS & Univ. Grenoble Alpes FR , Alexander Philippov University of California Berkeley USA
Pulsars create and launch an ultra-relativistic, ultra-magnetized wind of electron-positron pairs that terminates into a mildly relativistic, weakly magnetized nebula observed as a bright source of pure non-thermal radiation. Observations coupled with classical magneto-hydrodynamic models of pulsar wind nebulae suggest that significant magnetic dissipation must be occurring somewhere between the neutron star and the nebula. This issue is generic to other relativistic magnetic outflows found in astrophysical objects such as active galactic nuclei, microquasars and gamma-ray bursts. The current sheet forming in the equatorial regions of the pulsar wind is a promising site for magnetic dissipation via relativistic reconnection. The objective of this proposal is to simulate the dynamics of the pulsar wind self-consistently, from the magnetosphere where the wind is created up to unprecedentedly large distances from the star, using full 3D particle-in-cell simulations. Three regimes will be investigated: (i) a freely propagating pulsar wind relevant to isolated pulsars (ii) a truncated wind in the context of tight binary systems, and (iii) a pulsar wind and pulsar wind nebula under a strong inverse Compton cooling relevant to both isolated and binary systems. The numerical requirement for this campaign of simulations is estimated to about 26.73 millions core-hours on TGCC Joliot Curie SKL for one year.
The first luminous objects and reionisation with SPHINX
Project Title: The first luminous objects and reionisation with SPHINX
Project Leader: Joakim Rosdahl
Resource Awarded: 54 million core hours on JUWELS
Team Members: Jeremy Blaizot Centre de Recherche Astrophysique de Lyon FR , Thibault Garel Centre de Recherche Astrophysique de Lyon FR , Leo Michel-Dansac Centre de Recherche Astrophysique de Lyon FR , Taysun Kimm Yonsei University KR , Martin Haehnelt University of Cambridge UK , Harley Katz University of Oxford UK , Sergio Martin Alvarez University of Oxford UK , Marius Ramsoy University of Oxford UK , Jonathan Chardin Universite de Strasbourg FR , Romain Teyssier University of Zürich CH , Lewis Weinberger University of Cambridge UK , Laura Keating University of Toronto CA , Pierre Ocvirk Universite de Strasbourg FR
The Epoch of reionization (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 reionization 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 reionization 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 and thermochemistry of the collapsing gas, star formation, supernova explosions, emission of radiation from stars, radiation-gas interactions, and gas-magnetic field interactions. In a previous PRACE allocation in 2017, we received computing time to start the SPHINX suite of simulations, running cosmological volumes with almost two thousand resolved galaxies and their contributions to reionisation (Rosdahl+2018). We now wish to expand the SPHINX simulations to an eight times larger cosmological volume, resolving up to 15 thousand galaxies and capturing almost an order of magnitude larger galaxy masses. This unprecedented range of resolved galaxies performed with full radiation-hydrodynamics finally enables us to find out whether reionization of the Universe was powered by a plethora of low-mass dwarf galaxies, a few massive galaxies, intermediate ones, or all of the above. The simulations will aid to clear the picture and understand the underlying physics producing the wealth of data from observations in the coming years.
Obelisk – Radiation hydrodynamical simulation of the formation of a proto-cluster
Project Title: Obelisk – Radiation hydrodynamical simulation of the formation of a proto-cluster
Project Leader: Maxime Trebitsch
Resource Awarded: 20 million core hours on Joliot Curie – SKL
Team Members: Karl-Joakim Rosdahl Centre de Recherche Astrophysique de Lyon FR , Harley Katz University of Oxford UK , Taysun Kimm Yonsei University KR , Christophe Pichon CNRS / Institut d’Astrophysique de Paris FR , Yohan Dubois CNRS / Institut d’Astrophysique de Paris FR , Marta Volonteri CNRS / Institut d’Astrophysique de Paris FR , Ricarda Beckmann CNRS / Institut d’Astrophysique de Paris FR , Adrianne Slyz University of Oxford UK , Julien Devriendt University of Oxford UK
Massive galaxies assembled most of their mass in the first 3 Gyr of evolution of the Universe, before the peak of cosmic star formation. Supermassive black holes powering bright quasars are often found at the centre of these massive galaxies. The intense star formation and the accretion onto central supermassive black holes release a tremendous amount of energy in the surrounding gas via various feedback channels, whether radiative or hydrodynamical, strongly shaping the gas flows in and around galaxies, and will affect all neighbouring galaxies. Some of the the most pressing questions in astrophysics today are related to the formation history of these early galaxies. Understanding how these objects assembled their mass, how their star formation history is tied to the growth history of their central black hole, how they contribute to the chemical enrichment of the Universe, or what is their exact role in the reionization history of the Universe. While upcoming facilities such as the JWST or the next generation of Extremely Large Telescopes will revolutionize our views on these questions, there is an need for theoretical models to help understand future observations. To this end, we introduce the Obelisk project: a radiation-hydrodynamical cosmological simulation of a proto-cluster and its environment until the peak of cosmic star formation at z ~ 2. Combining state-of-the-art numerical cosmological codes with a detailed comparison with observations will enable us to understand better the high redshift Universe.
Simulating the Euclid Universe
Project Title: Simulating the Euclid Universe
Project Leader: Romain Teyssier
Resource Awarded: 136 million core hours on Piz Daint
Team Members: Doug Potter Universitat Zurich CH , Joachim Stadel Universitat Zurich CH
The Euclid satellite will be launched by the European Space Agency shortly after 2020 to map the entire Universe with more than a billion galaxies. This will provide an unprecedented measurement of the galaxy and mass distribution with sub-percent accuracy and a unique opportunity to discover new physics. Such an ambitious measurement requires a precise theoretical prediction of the observables. The challenge resides mostly at small scales, where Euclid will provide invaluable new information, complementary to other cosmology probes, but where the theory requires a full nonlinear treatment of collisionless fluid dynamics. Only N-body simulations with very large volumes and particle numbers can provide both large scale statistics and small scale accuracy. These simulations are used for two main purposes: 1- generate mock galaxy catalogues that we can compare directly to observations and assess the quality of the entire data processing pipeline, 2- provide emulators of the observables as a function of a multi-dimensional cosmological parameter space, in order to perform the likelihood analysis of the data. This proposal aims at: 1- producing a mock galaxy catalogue for the Euclid collaboration, including massive neutrinos and 2- producing a matter power spectrum emulator on an extended parameter space that include the massive neutrinos mass. Objective 1 will be met by simulating our fiducial model in a 4 Gpc periodic box with 4.4T particles, and by post-processing the simulation data on the light cone to obtain the final galaxy catalogue. Objective 2 will be met by performing 100 simulations of moderate size on a predefined set of parameters, and by using Uncertainty Quantification techniques to provide an accurate interpolator across parameter space. The products of this project will be made available to the entire Euclid community. For the first time, we will be able to assess the importance of nonlinear scale in determining the most likely cosmological parameters, compared to other cosmological probes.
GEE – Gravitational and electromagnetic emission from binary neutron star mergers: from coalescence to jet formation
Project Title: GEE – Gravitational and electromagnetic emission from binary neutron star mergers: from coalescence to jet formation
Project Leader: Carlos Palenzuela
Resource Awarded: 15.3 million core hours on MareNostrum
The recent direct detection of gravitational waves through interferometric observatories opens a new window to study the coalescence of binary neutron stars, which can be regarded as unique astrophysical laboratories to study gravity, plasma physics and dense matter under very extreme conditions. Concurrent observations of gravitational and electromagnetic waves produced by these sources start a new era of multi-messenger astronomy that will enhance our understanding on the parameters of the system and the physical processes at play, allowing us to test our theories and validate our astrophysical models. The project proposed here is based on the study of the full dynamics of magnetized binary neutron star mergers through extremely accurate numerical simulations, focusing on the physical mechanisms that are most relevant for the formation of detectable electromagnetic signals like short Gamma-Ray Bursts and kilonovae. Our simulations will focus on the different processes and instabilities increasing the strength of the magnetic field during the merger, as well as its conversion from small to large scales through a dynamo mechanism. Our results will allow us to study the effects of these strong magnetic fields on the electromagnetic signatures coming from the binary coalescence after the merger, affected at least by two different processes: (i) magnetic fields transfer angular momentum quite efficiently, accelerating the collapse to a black hole of the hyper-massive neutron star produced during the merger, and (ii) large scale magnetic fields are an essential ingredient for the jet formation. The proposed activities belong to an on-going multidisciplinary program that matches the intense theoretical and observational upcoming activities following the first gravitational wave detections of binary neutron stars and will allow us to maximize the scientific outcome of the upcoming data made soon available by the upgraded and new gravitational wave detectors.
MFDUSTY – Multi-fluid dust dynamics in protoplanetary disks
Project Title: MFDUSTY – Multi-fluid dust dynamics in protoplanetary disks
Project Leader: Turlough Downes
Resource Awarded: 60 million core hours on Marconi – KNL
Team Members: Donna Rodgers-Lee University of Hertfordshire UK , Antonella Natta Dublin Institute for Advanced Studies IE
Virtually all low and intermediate mass stars (such as our Sun) form by accreting material from a surrounding molecular cloud through an accretion disk. It is in these disks, so-called protoplanetary disks, that planets form. Observations indicate that many of these disks are in Keplerian rotation, implying a balance between the centrifugal and gravitational forces. There is a fundamental mystery surrounding the issue of how material can move inward through a protoplanetary disk and onto the young stellar object (YSO). As the material moves inward it must lose angular momentum rapidly – if this did not happen the material could not move in toward the forming star at the observed rates and star formation would be extremely difficult. How, then, does the accreting material lose its angular momentum? One possibility is that an instability such as the magnetorotational instability (MRI) could produce turbulence in the disk which would, itself, create a highly effective viscosity. This viscosity would then act to transfer angular momentum from material at a particular point in the disk to material further out in the disk, thereby enabling accretion. Protoplanetary disks, though, are complex systems comprised of many chemical species and dust grains some of which are charged (and interact with magnetic fields), and some of which are neutral: in short, it is a multifluid system made up of various charged and neutral fluids which move under differing equations of motion. To gain a proper understanding of how turbulence in accretion disks behaves and how it can be generated we must move towards full multifluid modeling. With such modeling we are able to self-consistently investigate possible regions of dust concentrations (of relevance to planet formation) and the creation of disk “gaps” by the action of MRI. This is particularly interesting because current observations of these protoplanetary disks focus on both the gas and the dust in the disks. The dust is not distributed in the same way as the gas, and this project will model the dynamics of the dust self-consistently including, for the first time, the effect of magnetic fields thereby allowing us to make a direct link between disk simulations and dust observations. We will use the the state-of-the-art, massively parallel multi-fluid magnetohydrodynamic code HYDRA, coupled with the power of the MareNostrum4 system, to perform full 3D modeling of the dynamics of these proto-planetary disks including charged dust dynamics. The observational signatures will be derived a posteriori from the simulations using a dust radiative transfer code such as the open-source HYPERION code. HYDRA is well-tested and proven both from the point of view of the published physical results it has generated and in terms of its scalability it has been shown to perform extremely well using 16384 cores on the MareNostrum4 system.
Multi-scale simulations of Cosmic Reionization
Project Title: Multi-scale simulations of Cosmic Reionization
Project Leader: Ilian Iliev
Resource Awarded: 20.8 million core hours on Marconi – Broadwell
Team Members: Hannah Ross University of Sussex UK , Peter Thomas University of Sussex UK , Garrelt Mellema Stockholm University SE , Azizah Hosein University of Sussex UK , Kyungjin Ahn Chosun University KR , Anastasia Fialkov The Harvard-Smithsonian Center for Astrophysics USA , Paul Shapiro The University of Texas at Austin USA , Jenny Sorce Lyon University FR , Gustavo Yepes Universidad Autónoma de Madrid ES , Stefan Gottloeber Leibniz-Institut fuer Astrophysik Potsdam (AIP) DE , Sergey Pillipenko Lebedev Physical Institute of the RAS RU , Taha Dawoodbhoy The University of Texas at Austin USA , Raghunath Ghara Stockholm University SE , Michele Bianco University of Sussex UK , Rajesh Mondal University of Sussex UK , Keri Dixon New York University Abu Dhabi AE
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 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), 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 does 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 that at 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 do alternative dark matter models affect reionization and do they have any unique observational signatures? and 6) How are these feedback effects imprinted on large-scale observational features?