Latest Results Gauss Centre for Supercomputing e.V.

LATEST RESEARCH RESULTS

Find out about the latest simulation projects run on the GCS supercomputers. For a complete overview of research projects, sorted by scientific fields, please choose from the list in the right column.

Computational and Scientific Engineering

Principal Investigator: Romuald Skoda, Lehrstuhl für Hydraulische Strömungsmaschinen, Ruhr-Universität Bochum

HPC Platform used: JUWELS of JSC

Local Project ID: chbo46, chbo48

While for the design point operation of centrifugal pumps an essentially steady flow field is present, the flow field gets increasingly unsteady towards off-design operation. Particular pump types as e.g. single-blade or positive displace pumps show a high unsteadiness even in the design point operation. Simulation results for the highly unsteady and turbulent flow in a centrifugal pump are presented. For statistical turbulence models an a-priori averaged turbulence spectrum is assumed, and limitations of these state-of-the-art models are discussed. Since the computational effort of a scale-resolving Large-Eddy-Simulation is tremendous, the potential of scale-adaptive turbulence models is highlighted.

Materials Sciences and Chemistry

Principal Investigator: Gerd Steinle-Neumann, Bayerisches Geoinstitut, Universität Bayreuth

HPC Platform used: SuperMUC-NG

Local Project ID: pn34wi

Metal hydrides have become of great scientific interest as high-temperature superconducting materials at high pressure, with hydrogen-hydrogen interactions suspected as critical in this behavior. Here, nuclear magnetic resonance experiments and electronic structure calculations are combined to explore the compression behavior of FeH and Cu2H, and results show that within the hydrides a connected hydrogen network forms at significantly larger H-H distances than previously assumed. The network leads to an increased contribution of hydrogen electrons to metallic conduction, and seems to induce a significantly enhanced diffusion of protons.

Computational and Scientific Engineering

Principal Investigator: Geert Brethouwer, Department of Engineering Mechanics, KTH, Stockholm, Sweden

HPC Platform used: JUWELS of JSC

Local Project ID: PRA108

Flows over the curved surface of wings, cars, turbine blades in gas turbines and impeller blades in pumps have curved streamlines. The influence of streamline curvature on flows, drag and also heat transfer in flows is substantial to large. However, engineering models have difficulties in correctly predicting flows over curved surfaces and our knowledge on streamline curvature influences on flows is still limited. In this project, turbulent flows in moderately to strongly curved channels are studied by highly accurate, large-scale numerical simulations fully resolving the turbulent fluid motions. These give important insights into streamline curvature influences on flows, and produce data that form the basis for better engineering models.

Computational and Scientific Engineering

Principal Investigator: Johannes Schemmel, Kirchhoff Institute for Physics, University of Heidelberg (Germany)

HPC Platform used: JUWELS of JSC

Local Project ID: chhd34

Impressive progress has recently been made in machine learning where learning capabilities at (super-)human level can now be produced in non-spiking artificial neural networks. A critical challenge for machine learning is the large number of samples required for training. This project investigated new high-throughput methods across various domains for biologically based spiking neuronal networks. Sub-projects explored tools and learning algorithms to study and enhance learning performance in biological neural networks and to equip variants of data driven models with fast learning capabilities. Applications of these learning techniques in neuromorphic hardware and design for their future application in neurorobotics were also included.

Computational and Scientific Engineering

Principal Investigator: Klaus Hannemann, Spacecraft Department, Institute of Aerodynamics and Flow Technology, German Aerospace Center (DLR)

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr62po

The aerodynamics of generic space launch vehicles, in particular the flow field at the bottom of the vehicle, at transonic conditions are investigated  numerically using hybrid RANS-LES methods. The focus of the project is the investigation of the impact of hot plumes and hot walls on the flow field. It is found that both higher plume velocities and higher wall temperatures shift the reattachment location downstream, leading to a stronger interaction of shear layer and plume. An additional contribution in the pressure spectral content is observed that exhibits a symmetric pressure footprint. The increased wall temperature leads to reduced radial forces on the nozzle structure due to a slower development of turbulent structures.

Elementary Particle Physics

Principal Investigator: Dénes Sexty, Bergische Universität Wuppertal, IAS/JSC Forschungszenturm Jülich

HPC Platform used: JUQUEEN and JUWELS of JSC

Local Project ID: chwu32

At high temperatures the nuclear matter melts into a plasma state. This phase transition is expected to have a “critical point” for systems which have increasingly more protons than antiprotons. The search for this elusive critical point on the QCD phase diagram is one of the greatest challenges in today’s high energy physics, both in theory and in experiment. The calculations of the theory at non-zero densities in supercomputers are hampered by the sign-problem. In this project multiple research tracks were pursued and the methods that deal with the sign-problem and search for signals of the critical point on the phase diagram were developed.

Computational and Scientific Engineering

Principal Investigator: Sylvain Laizet, Imperial College London, United Kingdom

HPC Platform used: Hazel Hen of HLRS

Local Project ID: PRACE4381

The need to reduce the skin-friction drag of aerodynamic vehicles is of paramount importance. Nominally 50% of the total energy consumption of an aircraft or high-speed train is due to skin-friction drag. Reducing skin-friction drag reduces fuel consumption and transport emissions, leading to vast economic savings and wider health and environmental benefits. In this project, wall-normal blowing is combined with a Bayesian Optimisation framework in order to find the optimal parameters to generate net energy savings over a turbulent boundary layer. It is found that wall-normal blowing with amplitudes of less than 1% of the freestream velocity of the boundary layer can generate a drag reduction of up to 80% with up to 5% of energy saving.

Elementary Particle Physics

Principal Investigator: Jeremy Green, Theoretical Physics Department, CERN, Geneva, Switzerland

HPC Platform used: JUQUEEN and JUWELS of JSC

Local Project ID: chmz37

Protons are composite particles: bound states of quarks and gluons, as described by the theory of quantum chromodynamics (QCD). Using lattice QCD, we know in principle how to use supercomputers to compute various properties of the proton such as its radius and magnetic moment, however this is very challenging in practice. A major part of this project was devoted to developing and studying methods for more reliable calculations, in particular for obtaining more accurate results in a finite box and for better isolation of proton states.

Materials Sciences and Chemistry

Principal Investigator: Karsten Reuter, Lehrstuhl für Theoretische Chemie, Technische Universität München

HPC Platform used: JUWELS of JSC

Local Project ID: tmcscat

As most notorious greenhouse gas, CO2 emissions prevail as high as about 364 million tons carbon with the concentration reaching over 400 ppm in the atmosphere. A drastic reduction of CO2 is urgently necessary for sustainable growth and to fight climate change. The electrochemical reduction of CO2 (CO2RR) is a promising approach to utilize renewable electricity to convert CO2 into chemical energy carriers at ambient conditions and in small-scale decentralized operation. Researchers from Technical University of Munich have employed an active-site screening approach and proposed carbon-rich molybdenum carbides as a promising CO2RR catalyst to produce methanol.

Environment and Energy

Principal Investigator: Clemens Simmer, Institute for Geosciences, University of Bonn

HPC Platform used: JUQUEEN and JUWELS of JSC

Local Project ID: chbn29, chbn37

A multi-institutional team of researchers is developing a data assimilation framework for coupled atmosphere-land-surface-groundwater models. These coupled models, which potentially allow a more accurate description of the coupled terrestrial water and energy fluxes, in particular fluxes across compartments, are affected by large uncertainties related to uncertain input parameters, initial conditions and boundary conditions. Data assimilation can alleviate these limitations and this project is focused in particular on the value of coupled data assimilation which means that observations in one compartment (e.g., subsurface) are used to update states, and possibly also parameters, in another compartment (e.g., land surface).

Life Sciences

Principal Investigator: Wolfgang Wenzel, Institute of Nanotechnology, Karlsruhe Institute of Technology (KIT)

HPC Platform used: SuperMUC-NG of LRZ

Local Project ID: pr27wi

GPCRs sit in the cell membrane and transmit signals from the outside of the cell to its interior. Currently, drugs targeting these receptors only work by mimicking ligands, i.e. they activate or inhibit the receptors by changing their conformation. If the GPCR adopts an active conformation, it can bind proteins on the intracellular side of the cellular membrane, which then transmit the signal inside the cell. In this study, we investigated how a protein that stops the GPCR from signaling, interacts with a prototypical GPCR. We discovered that specific lipids can modify how signals are transmitted by modifying the way of interaction between the GPCR and arrestin. In the future this could enable the discovery of a new kind of drugs for GPCRs.

Computational and Scientific Engineering

Principal Investigator: Manuel Keßler, Institute of Aerodynamics and Gasdynamics, University of Stuttgart

HPC Platform used: Hazeln Hen of HLRS

Local Project ID: GCS-CARo

Helicopters and other rotorcraft like future air taxis generate substantial sound, placing a noise burden on the community. Advanced simulation capabilities developed at IAG over the last decades enable the prediction of aeroacoustics together with aerodynamics and performance, and thus allow an accurate and reliable assessment of different concepts long before first flight. Consequently, this technology serves to identify promising radical configurations initially as well as to further optimize designs decided on at later stages of the development process. Conventional helicopters may benefit from these tools as much as breakthrough layouts in the highly dynamic Urban Air Mobility sector.

Computational and Scientific Engineering

Principal Investigator: Panagiotis Stathopoulos, Hermann-Föttinger-Institut, Technische Universität Berlin

HPC Platform used: SuperMUC-NG of LRZ

Local Project ID: pr27bo

Hydrogen-enriched fuels can reduce the CO2 emissions of gas turbines. However, the presence of hydrogen in fuel mixtures can also lead to undesirable phenomena like flashback. Swirling combustors can take advantage of an axial air injection to increase their resistance against flashback. Such an example is the swirl-stabilized presented in experiments at the TU Berlin. The axial momentum ratio between the fuel jets and the air was found to control flashback resistance. This experimental hypothesis motivates the present study where large-eddy simulations of the combustion system are carried out to study the physics behind flashback phenomena in hydrogen gas turbine combustors.

Astrophysics

Principal Investigator: Hubert Klahr, Max-Planck-Institut für Astronomie, Heidelberg (Germany)

HPC Platform used: JUQUEEN and JUWELS of JSC

Local Project ID: chhd19

MPIA scientists have developed a planetesimal formation model based on high-resolution hydro-dynamical simulations performed on JSC HPC systems. The simulations were used to model disk turbulence and its two effects on the dust, the mixing and diffusion of the dust on large scales but also the concentration of dust on small scales. This research helped to better understand the efficiency of these processes and to derive initial mass functions for planetesimals and gas giant planets to predict when and where planetesimals and Jupiter-like planets should form and of which size they will be. This is a fundamental step forward in understanding the formation of our own solar system as well as of the many planetary systems around other stars.

Elementary Particle Physics

Principal Investigator: Karl Jansen, Deutsches Elektronen-Synchrotron (DESY), Zeuthen

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr74yo

In this project the most inner structure of the proton has been deciphered through a large-scale numerical simulation of quantum chromodynamics. This could be achieved by novel algorithms developed by the project team. In particular, the project made a large leap forward to solve the spin puzzle of the proton. While theory predicted a dominant contribution to the spin of the proton from the quarks, in experiments it was found that this contribution is surprisingly small. The research team found out that it is actually the gluon which is contributing a large fraction of the spin. Although still a number of systematic uncertainties have to be fixed, this is a most remarkable result which will lead to eventually resolve the proton spin puzzle.

Computational and Scientific Engineering

Principal Investigator: Paul Zimmermann, French National Institute for computer science and applied mathematics (INRIA), France

HPC Platform used: JUWELS of JSC

Local Project ID: RSA250

Data sent over the internet relies on public key cryptographical systems to remain secure. A project under leadership of Dr. Paul Zimmermann of the French National Institute for computer science and applied mathematics (INRIA), run on HPC system JUWELS of the Jülich Supercomputing Centre, has been carrying out record computations of integer factorisation and the discrete logarithm problem, the results of which are used as a benchmark for setting the length of the keys needed to keep such systems secure.

Materials Sciences and Chemistry

Principal Investigator: Vangelis Daskalakis, Cyprus University of Technology

HPC Platform used: SuperMUC-NG of LRZ

Local Project ID: pn34we

In this project, the biophysics of Photosynthesis are probed employing high-performance computing. Photosynthesis is based on the Sun light and fuels the metabolic pathways of numerous organisms in our biosphere. However, fluctuations in the light intensity or quality are expected due to the diurnal cycle, or the environmental conditions and could be detrimental to plants. Absorption of light and tunnelling of the associated energy towards the reaction centres of the photosynthetic apparatus are finely-tuned within a well-orchestrated photoprotective mechanism. The atomic-scale details of this mechanism is probed by computational biophysics, with applications on the increase of crop yields and artificial photosynthesis.

Materials Sciences and Chemistry

Principal Investigator: Karsten Reuter, Lehrstuhl für Theoretische Chemie, Technische Universität München

HPC Platform used: JUWELS of JSC

Local Project ID: LMcat

It is well-known that the catalytic properties of metals may extend beyond their melting point. Recently, this has been exploited to grow high-quality 2D materials such as graphene. To improve our understanding of the growth mechanism on liquid metal catalysts, researchers at the Technical University of Munich have employed a multi-scale modelling approach. Here, detailed simulations of various building blocks for the final graphene sheet such as simple hydrocarbons and smaller graphene flakes on solid and liquid Cu surfaces have been carried out. The insights from these simulations were then used to propose a mesoscopic model for the dynamics of graphene growth on molten Cu based on capillary and electrostatic interactions.

Materials Sciences and Chemistry

Principal Investigator: Jadran Vrabec, Chair of Thermodynamics and Process Engineering, Technische Universität Berlin

HPC Platform used: Hazel Hen and Hawk of HLRS

Local Project ID: MMHBF2

Computational fluid dynamics (CFD) simulations play an important role in today’s science and technology. Therefore, it is crucial to validate its underlying methods and models. This can be done by experiments or with molecular dynamics (MD) simulations, but in some cases only the latter are applicable. Since MD simulations follow the motion of each molecule individually, they are computationally very demanding, but they rest on an excellent physical basis. In this project, large systems of several hundred million atoms are considered to study the thermo- and hydrodynamic behavior of fluids during shock wave propagation, droplet coalescence and injection. The results are compared to that of macroscopic numerical methods.

Elementary Particle Physics

Principal Investigator: André Sternbeck, Institute for Theoretical Physics, Friedrich-Schiller-University Jena

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr48ji

Super-Yang-Mills theory is a central building block for supersymmetric extensions of the Standard Model. While the weakly coupled sector can be treated within perturbation theory, the strongly coupled sector must be dealt with a non-perturbative approach. Lattice regularizations provide such an approach but they break supersymmetry and hence the mass degeneracy within a supermultiplet. Researchers of Uni Jena study N=1 supersymmetric SU(3) Yang-Mills theory with a lattice Dirac operator with an additional parity mass. They show that a special 45° twist effectively removes the mass splitting at finite lattice spacing–thus improves the continuum extrapolation—and that the DDαAMG algorithm accelerates such lattice calculations considerably.

For a complete list of projects run on GCS systems, go to top of page and select the scientific domain of interest in the right column.