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: Ulrich Rist, Markus Kloker, Christoph Wenzel, Institute of Aerodynamics and Gas Dynamics, University of Stuttgart

HPC Platform used: Hazel Hen and Hawk of HLRS

Local Project ID: GCS-Lamt

This project explores laminar-turbulent transition, turbulence, and flow control in boundary layers at various flow speeds from the subsonic to the hypersonic regime. The physical problems under investigation deal with prediction of laminar-turbulent transition on airfoils for aircraft, prediction of critical roughness heights in laminar boundary layers, turbulent drag reduction, the origins of turbulent superstructures in turbulent flows, the use of roughness patterns for flow control, effusion cooling in laminar and turbulent supersonic boundary-layer flow, DNS of disturbance receptivity on a swept wing at high Reynolds numbers, and plasma actuator design for active control of disturbances in a swept-wing flow.

Astrophysics

Principal Investigator: Daniel Ceverino, Zentrum für Astronomie, Institut für Theoretische Astrophysik, Universität Heidelberg

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr92za

The FirstLight project at LRZ is a large database of numerical models of galaxy formation that mimic a galaxy survey of the high-redshift Universe, before and after the Reionization Epoch. This is the largest sample of zoom simulations of galaxy formation with a spatial resolution better than 10 pc. This database improves our understanding of cosmic dawn. It sheds light on the distribution of gas, stars, metals and dust in the first galaxies. This mock survey makes predictions about the galaxy population that will be first observed with future facilities, such as the James Webb Space Telescope and the next generation of large telescopes.

Computational and Scientific Engineering

Principal Investigator: Christian Bauer, Institute of Aerodynamics and Flow Technology, German Aerospace Center (DLR)

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr62zu

A large amount of the energy needed to push fluids through pipes worldwide is dissipated by viscous turbulence in the vicinity of solid walls. Therefore the study of wall-bounded turbulent flows is not only of theoretical interest but also of practical importance for many engineering applications. In wall-bounded turbulence the energy of the turbulent fluctuations is distributed among different scales. The largest energetic scales are denoted as superstructures or very-large-scale motions (VLSMs). In our project we carry out direct numerical simulations (DNSs) of turbulent pipe flow aiming at the understanding of the energy exchange between VLSMs and the small-scale coherent.

Computational and Scientific Engineering

Principal Investigator: Barbara Wohlmuth, Lehrstuhl für Numerische Mathematik, Technische Universität München

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr74ne

Large scale simulations are particularly valuable and important for a better understanding of coupled multi physics problems describing a large class of physical phenomena. This research project focuses on the development of new numerical methods for efficiently solving coupled non-linear and time-dependent fluid flow problems on a large scale. In particular, two applications are considered. Namely, the Navier–Stokes equations coupled to a transport equation describing diluted polymers and geodynamical model problems which involve non-linearities in the viscosity. The goal is to develop new methods for solving these problems, evaluating their performance and scalability, and to perform simulations based on these new methods.

Computational and Scientific Engineering

Principal Investigator: Andreas Goerttler, Anthony Gardner, Institute for Aerodynamics and Flow Technology, German Aerospace Center (DLR), Göttingen

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr53fi

Using DLR’s finite-volume solver TAU, researchers of the Institute for Aerodynamics and Flow Technology at DLR Göttingen numerically investigated the vortex system of four rotating and pitching DSA-9A blades. The computations were validated against experimental data gathered using particle image velocimetry (PIV) carried out at the rotor test facility in Göttingen. Algorithms deriving the vortex position, swirl velocity, circulation and core radius were implemented. Hover-like conditions with a fixed blade pitch were analyzed giving a good picture of the static vortex system. These results are used to understand the vortex development for the unsteady pitching conditions, which can be described as a superposition of static vortex states.

Elementary Particle Physics

Principal Investigator: Hartmut Ruhl, Karl-Ulrich Bamberg, Ludwig-Maximilians-Universität München, Faculty of Physics, Chair for Computational and Plasma Physics

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr74si

The availabiltiy of ultra-short, high-power lasers has led to greater interest in their potential use for accelerators, as the charge separation in plasmas can induce enormous electromagnetic field strengths on a sub-micrometer scale. With the high performance and extreme scalability of the Plasma Simulation Code (PSC) for fully kinetic simulations, a wide field of applications was researched: From ions for medical purposes (Ion Wave Breaking Acceleration and Mass-Limited Targets) to breakthrough Lepton acceleration by proton-driven wakefields (AWAKE), all the way to radiation generation (attosecond X-ray pulses from Ultra-Thin Foils). Even QED based approaches were covered in this project.

Elementary Particle Physics

Principal Investigator: Harvey Meyer, Johannes Gutenberg University Mainz

HPC Platform used: JUQUEEN and JUWELS of JSC

Local Project ID: chmz36

Although Quantum Chromodynamics (QCD) has long been established as the correct theory of the subatomic strong interaction, obtaining quantitative predictions from it often represents a challenging computational task. In this project, large-scale lattice QCD simulations are used to determine structural properties of protons and neutrons. The lattice approach to QCD amounts to discretizing space-time and applying importance-sampling techniques to the path-integral representation of QCD. One specific observable under scrutiny in this project is the “scalar matrix element” of the proton, which provides a quantitative answer to the question of “How much would the proton mass change if the light quark masses changed by a small amount?”.

Computational and Scientific Engineering

Principal Investigator: Xiangyu Hu, Chair of Aerodynamics and Fluid Mechanics, Technische Universität München

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr53vu

As a Lagrangian method, Smoothed Particle Hydrodynamics (SPH) has been explored and demonstrated for a wide range of applications. Several open-source frameworks exist for the large-scale parallel simulation of particle-based methods in which the resolution of simulation is fixed. Some preliminary work has also been published to tackle the difficulties encountered in extending codes with adaptive-resolution capability. However, the support for fully parallelized adaptive-resolution in distributed systems is generally still limited in the aforementioned codes. This research project focuses on an alternative approach by introducing a new multi-resolution parallel framework employing several algorithms from previous work.

Materials Sciences and Chemistry

Principal Investigator: Fakher Assaad, Lehrstuhl für Theoretische Physik I, Julius-Maximilian-Universität Würzburg

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr94vu

To unravel the complexity of the solid state, researchers from the University of Würzburg have mastered very different and complementary methods. Density functional theory in the local density approximation with added dynamical local interactions using the dynamical mean-field approximation has the merit of being material dependent since one can include the chemical constituents of materials. Spacial and temporal fluctuations are crucial to understand e.g. the Iridates, a topic that is explored with the new pseudo-fermion functional renormalization group. Another aspect of this research are realistic quantum Monte Carlo simulations of free standing graphene aiming to elucidate the role of electronic correlations.

Computational and Scientific Engineering

Principal Investigator: Franco Magagnato, Institute of Fluid Mechanics, Karlsruhe Institute of Technology

HPC Platform used: Hazel Hen and Hawk of HLRS

Local Project ID: Imp_DNS

The heat transfer in the stagnation region of an impinging jet at given jet to distance ratio, Re-number and Temperature ratio also depend on the turbulent inflow characteristics. Using Direct Numerical Simulations, the Nusselt-number distribution as well as the turbulent statistics close to the heated wall have been investigated. At first a calculation has been done comparing the results with published DNS and experiments from Dairay et al. (2015). Since in their paper not all necessary turbulence values were given, the missing values (e.g. turbulent length scale) had to be adjusted in order to fit their results. A good agreement has been found of our calculations with their DNS and experiments.

Astrophysics

Principal Investigator: Luciano Rezzolla, Institute for Theoretical Physics, Goethe University FIAS – Frankfurt Institute for Advanced Studies

HPC Platform used: SuperMUC and SuperMUC-NG

Local Project ID: pr27ju

Two major events are responsible for what is considered the “golden age” of relativistic astrophysics. One is the detection of gravitational waves from merging neutron stars heralding the beginning of the multimessenger age. The other is the effort of the Event Horizon Telescope collaboration culminating in the first image of a black hole. Both events have been aided by simulations that require HPC. With this project, several studies could be conducted well alligned with these type of simulations expanding our knowledge about these important astrophysical events.

Computational and Scientific Engineering

Principal Investigator: Claus-Dieter Munz, Institute of Aerodynamics and Gas Dynamics, University of Stuttgart

HPC Platform used: Hazel Hen and Hawk of HLRS

Local Project ID: hpcmphas

In order to simulate compressible multi-phase flows at extreme ambient conditions, researchers from the Institute of Aerodynamics and Gas Dynamics have developed a multi-phase flow solver based on the discontinuous Galerkin spectral element method in conjunction with an efficient tabulation technique for highly accurate equations of state. The aim of this development is the simulation of phase transition, droplet dynamics and large-scale multi-component phenomena at pressures and temperatures near the critical point. Simulations of liquid fuel injections and shock-drop interactions have been performed on the HPC systems installed at the High-Performance Computing Center Stuttgart (HLRS).

Elementary Particle Physics

Principal Investigator: Hartmut Wittig, Institute for Nuclear Physics and PRISMA Cluster of Excellence, Johannes Gutenberg University of Mainz

HPC Platform used: Hazel Hen and HAWK of HLRS

Local Project ID: GCS-HQCD

The Standard Model of Particle Physics is a highly successful theoretical framework for the treatment of fundamental interactions, but fails to explain phenomena such as dark matter or the abundance of matter over antimatter. Precision observables, such as the anomalous magnetic moment of the muon, aμ, play a central role in the search for “New Physics”. A promising hint is provided by the persistent tension of 3.7 standard deviations between the theoretical estimate for aμ and its experimental determination. In our project we employ the methodology of lattice QCD to compute the hadronic contributions to aμ from first principles. In the long run, our results will supersede the estimates based on data-driven approaches and hadronic models.

Computational and Scientific Engineering

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

HPC Platform used: Hazel Hen and Hawk of HLRS

Local Project ID: DGDES

The aerodynamic flow field around helicopters is challenging to simulate due to complex configurations in relative motion. In an effort to evolve computational fluid dynamics (CFD) technology to new levels of accuracy, reliability, and parallelization efficiency, the helicopter & aeroacoustics group at the IAG of University of Stuttgart employs advanced, high-order Discontinuous Galerkin (DG) methods to help solve difficult rotorcraft-based engineering applications. Complex geometries, curved surfaces, relative motion with elaborate kinematics, and fluid-structure coupling to blade dynamics call for sophisticated techniques within the simulation tool chain to account for all important physical phenomena relevant to the field of study.

Environment and Energy

Principal Investigator: Stefan Emeis , Institute for Meteorology and Climate, Atmospheric Environmental Research, Karlsruhe Institute of Technology

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr27po

To avoid dangerous climate change, we have to reduce the emission of greenhouse gases radically. This requires – among other measures – an increase of renewable sources of energy like solar and wind. In 2019, already a quarter of Germanys electricity demand has been met by wind power. In order to increase this share, one has to develop sites in hilly terrain. High resolution models are required to assess the suitability of candidate sites with respect to turbulence intensity, power production and variability. This project supports the development of the test-site WINSENT, which is located on the Swabian Alp near Stuttgart.

Materials Sciences and Chemistry

Principal Investigator: Jakob Timmermann, Chair of Theoretical Chemistry, Technical University of Munich

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr53qu

The water electrolysis in Proton Exchange Membrane (PEM) cells is fitting plenty of industrial requirements. The main drawback of PEM cells however is the overpotential of the oxygen evolution reaction. In its acidic environment iridium dioxide (IrO2) is currently the only stable catalyst. Yet the low abundance of iridium makes a reduction of its loading inevitable. One approach to decrease the catalyst loading is the use of nanoparticles. For catalyst optimization a general understanding of shape and surface structure of these nanoparticles is required. In this project a protocol has been developed to generate and simulate IrO2 nanoparticles based on energies of slab models and to provide insights regarding stability and structure.

Astrophysics

Principal Investigator: Tim Dietrich, University of Potsdam, Dutch National Institut for Subatomic Physics Amsterdam

HPC Platform used: SuperMUC and SuperMUC-NG of LRZ

Local Project ID: pr48pu

Neutron stars are ultracompact stars in which densities above the nuclear saturation densities are reached and that provide one of the best laboratories to test nuclear physics principles. Within this project, researchers perform 3+1-dimensional numerical-relativity simulations studying the last few orbits before the merger of two of these stars. In fact, a binary neutron star merger is one of the most energetic phenomena in our Universe and is accompanied by a variety of electromagnetic signatures and with characteristic gravitational-wave signatures. With the help of these simulations existing theoretical models can be developed and verified and the growing field of multi-messenger astronomy is supported. 

Computational and Scientific Engineering

Principal Investigator: Harald Klimach, Simulation Techniques and Scientific Computing, University of Siegen

HPC Platform used: SuperMUC of LRZ

Local Project ID: pr62cu

This project looked into various strategies to couple domains of distinct physical and numerical properties to tackle direct aero-acoustic simulations. A turbulent flow around an airfoil and the emitted sound waves in a large area of interest was simulated. Different physical effects can be observed in spatially separated domains and the appropriate equation systems are solved in each one using the best fitting numerical discretization. The main focus of the project was the evaluation of different coupling methods to enable this partitioned simulation on massively parallel systems.

Materials Sciences and Chemistry

Principal Investigator: Johannes Ehrmaier, Department of Theoretical Chemistry, Technical University of Munich

HPC Platform used: SuperMUC of LRZ

Local Project ID: pr53wo

Carbon nitride materials have attracted vast interest in the field of photocatalytic water splitting. However, the underlying mechanism is not fully understood. Herein, results are being reported from large-scale first-principles simulations for the specific electron- and proton-transfer processes in the photochemical oxidation of liquid water with heptazine-based photocatalysts. The results reveal that heptazine possesses energy levels that are suitable for the water oxidation reaction. Moreover, the critical role of the solvent in the overall water-splitting cycle is shown. A simple model is developed to describe the water oxidation mechanism.

Life Sciences

Principal Investigator: Frauke Gräter, Interdisciplinary Center for Scientific Computing, Heidelberg University; and Group for Molecular Biomechanics, Heidelberg Institute for Theoretical Studies

HPC Platform used: JUWELS of JSC

Local Project ID: chhd33

Cells communicate with each other through biochemical as well as mechanical signals. Essential biological processes such as cell division are critically steered by the tension across the cell-cell contacts. Using extensive molecular dynamics simulations, scientists analyzed the underlying molecular principles of mechano-sensing at cell-cell contacts. These simulations can give first insights into how proteins present at the cell-cell contact change their structure and localization and thereby help to sense mechanical stimuli. The findings can help understanding the mechanisms by which tissues, e.g. skin, grow along the direction of pulling forces which were applied by adding virtual springs into the simulation system.