Researcher: Fabien Jaulmes
Project: Computational modelling of fast ion orbits in tokamak plasmas
Allocation: 1 000 000 core hours
Abstract: Nuclear fusion technology might enable us to generate energy without releasing large amounts of greenhouse gases into the atmosphere or leaving behind us long lived radioactive waste. The tokamak concept involves the use of magnetic fields to confine plasma hot enough to sustain fusion within itself. Today, as a part of international project under the title ITER, a new tokamak is built in southern France. If successful, the device would be the first one of its kind to produce net energy.
COMPASS Upgrade will be a large magnetic field (5T) tokamak that will allow the scientific investigation of various physics issues related to the operation of the future ITER. In particular an 80keV Neutral Beam Injection system is planned to heat up the plasma with 4MW of external power. The study and modelling of NBI-born particle behavior is of great relevance and might impact future design of the system and its integration in the overall design. We request here computational time for the modelling of the interaction of the fast particles with the background plasma. Our modelling tool, the EBdyna code with its new collisional features, has been benchmarked against the NUBEAM code on several test-cases. A publication in the Nuclear Fusion journal summarizes the results of our initial modelling effort [1].
Researcher: Jakub Vymola
Project: CPU and GPU scaling of DFT calculations, part III
Allocation: 228 000 core hours
Abstract: GPU-accelerated are being recognized as a very serious contender to speed up scientific calculations. More and more codes are being ported to support GPUs. They have been shown to give huge performance benefits in contrast x86 platform.
In this study we promote scaling tests of Vienna Ab initio Simulation Package (VASP) for number of compounds, i.e. one containing f-electrons (U2C3), metallic one (PtN2 in fluorite structure) and an insulator (pyrite structure of PtN2) for both CPU and GPU nodes demonstrating various sizes of the systems as well as various schemes to speedup calculations using different ratio of parameters (e.g. k-points vs. number of bands).
We intend to compare the speeds of calculations with VASP version 6 on different numbers of compute nodes (with and without acceleration) and on different machines, including Karolina. We aim to find and verify the fastest parallelization configuration of VASP6 for CPU nodes for each of the test cases. After that, we perform the very same tests on GPU nodes and compare the final results with respect to the total amount of resources consumed. We aim to provide solid insight into the use of GPU on IT4Innovations machines and to show how to utilize them efficiently. In general, we are hoping to reduce time and resources usage and to lower the price (core-hours) of research that heavily relies on the use of VASP code.
Researcher: Pavel Hobza
Project: IN SILICO Drug Design
Allocation: 7 599 000 core hours
Abstract: Computer aided drug design is a well-established tool for discovering and optimizing ligands of pharmaceutically relevant targets. An accurate prediction of ligands' activities critically depends on a chalenging task – a reliable description of noncovalent interactions, including quantum effects. Semiempirical quantum mechanical (SQM) methods are able to carry this task on the large protein-ligand systems in a reasonable time. However, the original SQM methods, such as PM6, have rather poor description of noncovalent interactions due to many approximations. We have developed corrected SQM methods (e.g. PM6-D3H4X) which accurately describe noncovalent interactions. We have successfully applied our scoring function (SF) based on these methods to ranking of inhibitors of various kinases, proteases, aldo-keto reductases, influenza polymerase16 and other targets. For virtual screening of large numbers of drug candidates, we have developed an accelerated SQM-based SF called SQM/COSMO.20 We have shown that SQM/COSMO outperforms classical SFs in native pose identification, ligand ranking,23 and virtual screening. Currently, we examine the performance of SQM/COSMO on an extended series of targets. Simultaneously, we also examine its ability to identify hotspots in protein-protein interactions for peptidomimetic compound design (e.g. the insulin receptor-insulin complex).
Researcher: Pablo Nieves
Project: High-throughput screening of novel Rare-Earth free magnets with high magnetostriction
Allocation: 6 000 000 core hours
Abstract: Magnetostriction is a physical phenomenon in which the process of magnetization induces a change in shape or dimension of a magnetic material. Nowadays, in many technical applications such as electric transformers, motor shielding, and magnetic recording, magnetic materials with extremely low magnetostriction are required. By contrast, materials with large magnetostriction are needed for many applications in electromagnetic microdevices as actuators and sensors. Until recently, the crystalline magnetostrictive materials with high magnetostriction have come mainly from Rare-Earth (R) based magnets like cubic RFe2 and RCo2 Laves or hexagonal phases of RCo5 and R2Co7. The problem of R availability has also motivated the exploration of R-free magnetostrictive materials like Galfenol (Fe-Ga), spinel ferrites (CoFe2O4), Nitinol (Ni-Ti alloys), Fe-based Invars, and Ni2MnGa. In this proposal, we aim to explore the magnetostriction of novel R-free systems by using our new developed software MAELAS for high-throughput first-principles calculations.
Researcher: Pavel Balaz
Project:
Allocation: 268000 core hours
Abstract: Skyrmions are particle-like topological defects in magnetic textures which emerge in certain class of magnetic materials under given external conditions. Due to their high stability, skyrmions can be efficiently manipulated by magnetic fields and electric currents. This makes skyrmions hot candidates for information carriers in future computational devices, which can operate at low energy costs and high efficiency. Therefore, research of skyrmions and their ability to move and interact is crucial for development of novel spintronic applications.
In our project, we focus on the study of ordered systems of skyrmions, which are known as skyrmion lattice. Hexagonal skyrmionic lattice can be stabilized in certain range of external magnetic field, and temperature. In order to identify the boundaries between the skyrmion lattice and other phases, we shall develop a new machine learning technique, which will be able to fasten the construction of the phase diagram. Second, we shall draw our attention to study of magnetization dynamics of a single topological defect inside a skyrmion lattice, which might open new possibilities towards utilizing skyrmion lattices in spintronic devices.
Researcher: Jana Pavlu
Project: The effect of oxygen on stability of eta phases in Mn-based systems
Allocation: 3 739 000 core hours
Abstract: Our modern society, based on the utilisation of advanced materials, requires the development of new materials. Nevertheless, their properties are significantly affected by impurities. The proposed project aims to understand how oxygen influences the stability of selected Mn-based phases and their magnetic properties. Here, the eta phases in Mn-Ta and Mn-Nb systems are chosen as the model systems as their formation is closely related to the presence of oxygen. Further, they constitute promising materials for future technological applications as (i) Mn increases the yield strength of Ti-Nb-Ta-Mn alloy foams for biomedical implants [1,2], (ii) niobium-manganese composite electrodes are more suitable in supercapacitors than niobium ones [3]; (iii) Mn, Nb and B additions maximise strength and toughness in martensitic micro-alloyed steels for heavy-duty engine connecting rods [4]; and Ta increases the pitting corrosion in super duplex stainless steels [5]. Unfortunately, very little is known about the mechanisms of the oxygen-stabilisation of the eta phase and the effect of oxygen on the magnetism of neighbouring atoms is experimentally inaccessible. Fortunately, it can be studied by means of theoretical approaches such as computational modelling, as planned in this project.
Researcher: Jan Simkanin
Project: Thermochemically-driven convection and dynamos operating at low Ekman numbers – Part II (follow up to OPEN-21-1)
Allocation: 2 279 000 core hours
Abstract: The Earth's magnetic field is one of the most variable geophysical fields and provides us with effective protection against high-energy, electrically charged particles from the solar wind, solar flares and is a useful tool for navigation, not only for us, humans but also for animals. The Earth's magnetic field is generated by convective motions of an electrically conductive melt in the Earth's outer core and penetrates the surface of the Earth. These generation processes are called Geodynamo for short. On the Earth's surface, a large-scale, dipole-dominated field is observed. However, we have no direct information about the magnetic field in the Earth'score. For this reason, we model magnetohydrodynamic processes in the Earth's core numerically. Numerical modelling of Geodynamo has made enormous progress during the last years thanks to the development of computer technology. Numerical models produce magnetic fields that are close to the observed geomagnetic field. We are also able to reproduce its temporal changes, whether short-term or long-term ones (so-called secular variations). However, we cannot use parameters typical of the Earth's core in our models. The real parameter values cannot be used for computational reasons – so far there is no supercomputer in the world that could solve a Geodynamo model with such values. However, we are gradually approaching these values as computers become more and more powerful.
Researcher: Sergiu Arapan
Project: Controlling the exchange coupling with ferromagnetic impurities
Allocation: 3 420 000 core hours
Abstract: Interlayer exchange coupling in metallic multilayers between two ferromagnetic layers across a nonmagnetic spacer layer has been of great interest for fundamental studies and applications. The magnetic phenomena in these systems form the basis of a new field of synthetic antiferromagnetic spintronics. The tunability of magnetic multilayers allows for optimization of properties that are desirable for applications including magnetic field sensing and magnetic random access memory. The interlayer exchange coupling may be altered by the presence of the ferromagnetic elements in the nonmagnetic spacer. Here we investigate the role of magnetic impurities in controlling the interlayer exchange coupling by first-principle calculations. The value of this exchange coupling influence the magnetization dynamics and may decrease the switching time of the memory cells.
Researcher: Stepan Sklenak
Project: Periodic DFT studies of zeolite-based catalysts
Allocation: 3 900 000 core hours
Abstract: Zeolite based catalysts are the most important industrial catalysts used mainly in petrochemistry (production of fuels, various hydrocarbons), and furthermore, for N2O decomposition and N2O selective catalytic reduction (SCR), for direct NO decomposition as well as for SCR of NOx using various reducing agents. In addition, zeolites are also employed as catalysts in synthesis of fine chemicals and as well as in the processing of biomass. Zeolites are also used as ion exchange agents (e.g., they replaced harmful phosphates in washing powder detergents). Zeolites are crystalline microporous aluminosilicates with a unique microporous nature, where the shape and size of a particular pore system exerts a steric influence on the reaction, controlling the access of reactants and products. Thus, zeolites are often said to act as shape-selective catalysts. Increasingly, attention has focused on fine-tuning the properties of zeolite catalysts in order to carry out very specific syntheses of high-value chemicals. Periodic DFT methods permit investigations of properties of zeolite-based catalysts which are needed for their fine-tuning. DFT calculations are complementary to experimental examinations and together they can provide more complex knowledge of the properties of the studied catalysts and the reactions they catalyze.
Researcher: Michal Otyepka
Project: Photoexcited electrons in the complex structures of carbon dots
Allocation: 3 000 000 core hours
Abstract: Carbon dots (CDs) are technologically interesting nanomaterials owing to their tunable photoluminescence (PL) properties, biocompatibility, low toxicity, easy synthesis, etc. Although the structural and PL properties of CDs have been theoretically examined,1–4 the so far published studies only utilized small ad hoc models not capturing the high complexity of real CDs. Additionally, no theoretical study addressing dynamical processes upon excitation, i.e., structural relaxation and charge transfer between core, surface and molecular states, has been published yet. The goal of this project is to unravel processes occurring after photoexcitation of CDs using quantum methods applied to realistic structural models. To this end, we plan to utilize program packages for simulating the dynamics of electronically excited molecules and molecular assemblies, namely Newton-X, in combination with ab initio and semi-empirical computational chemistry software packages. We expect that these calculations will elucidate the roles of specific structural (core, shell, molecular fluorophores etc.) regions in PL of CDs. Such knowledge can guide experimentalists to efficiently tune CDs' PL by modifying their chemical composition and structure via controllable bottom-up synthetic procedures.
Researcher: Azin Shahsavar
Project: Tunable Magnetic Interactions on Graphene
Allocation: 800 000 core hours
Abstract: In this project, we aim to provide a theoretical understanding of the electrical tuning of magnetic coupling between magnetic centers within metal-organic networks synthesized on graphene substrates. The metal-organic networks allow precise positioning of magnetic atoms into long-range-ordered lattices on the graphene substrate. We will exploit the possibility of an external tuning of the graphene Fermi level through a gate voltage and explore the possibility of modulation of the magnetic coupling between the magnetic centers. We aim at a theoretical description and understanding of the tunable magnetic interactions with the help of ab-initio DFT calculations. To provide a correct description, we will utilize a range of experimental data that will provide a solid background for theoretical calculations. The project is of great fundamental interest, as many conflicting predictions regarding strength and distance dependence of such indirect magnetic couplings exist in the literature, which are not easily discerned without the experimental data.
Researcher: Rajko Cosic
Project: Incrorporating the anharmonicity in the Wigner function-based sampling
Allocation: 333 000 core hours
Abstract: Wigner function-based sampling is a convenient way to obtain the initial conditions for the molecular dynamics simulations. However, the construction of Wigner functions for the Hamiltonian with general, anharmonic, potential energy surface (PES), is still a challenging task. In our approach, the Wigner functions are constructed from the vibrational self consistent field (VSCF) generated wavefunctions. To lower the computational complexity, the ab initio PES is represented by its expansion in the normal modes of its Hessian matrix calculated at the minimal-energy configuration. The combination of single normal modes and couplings up to n-th order enables us to incorporate the anharmonic corrections into the computationaly cheap potential. The project aims at two main kinds of calculations: a) the validation of the approach via the comparison of the configuration sampling between the Wigner approach and fully anharmonic benchmark calculations provided by the path integral Monte Carlo (PIMC) simulations. Due to computational complexty of the PIMC method, these simulations are to be performed with computationally cheap potential. b) the pilot calculations with the computationally expensive ab
initio methods will be performed.
Researcher: Martin Beseda
Project: Modeling of Collision Processes in Low-Temperature Plasma
Allocation: 2 039 000 core hours
Abstract: This project is meant to cover the majority of the necessary computational expenses of the DGS project Modeling of Collision Processes in Low-Temperature Plasma. Considering its similar content to OPEN-20-20, the relevant parts will be cited in this proposal.
Nowadays, applications of low-temperature plasma are a topic of interest in areas like surface treatment [1], food industry [2], and medicine [3,4,5,6] with the last one motivating our current research efforts. This project will be a direct continuation of our previous research efforts, following up Van de Steen's work [7,8,9,10] and several OPEN projects. As our current understanding of processes occurring in medicinal applications is not sufficient to tune the plasma for specific purposes[11], we aim to describe the processes in detail from the creation phase to the moment of application in the future. As the ionization and collisions of atomic and molecular ions have been covered already, we aim to continue with a) calculations of transport properties of collision complexes created by interactions of carrier-gas ions with the air and b) modeling of molecular ions formation. Both of these goals are tightly entangled with two double-degree Ph.D. theses co-directed by VSB-TUO and Université Toulouse III - Paul Sabatier, France, following the existing long-time collaboration of our team with LAPLACE, LCPQ, and LCAR research laboratories of the French university.
Researcher: Karel Sindelka
Project: Mesoscopic simulations of colloidal aqueous solutions in steady shear flow
Allocation: 900 000 core hours
Abstract: Aqueous solutions are omnipresent in nature, industrial processes and daily life. Understanding their behaviour in inhomogeneous environments (nanopores, self-assembled systems) and in equilibrium as well as non-equilibrium conditions (shear flow) is important in many key applications such as medicine or environmental protection. In this project, we focus on two aqueous solutions: surfactant adsorption on soft surfaces and solubilisation and release of small molecules into and from polymeric structures. In both systems, we also investigate the effect of shear flow on the equilibrium structure and behaviour. We use mesoscopic simulations to provide molecular-level insights into these systems as well as to fill gaps in our understanding of the physical and chemical behaviour of the studied systems.
Researcher: Michal Krumnikl
Project: Fiji Bioimage Informatics on HPC - Path to Exascale
Allocation: 900 000 core hours
Abstract: Aqueous solutions are omnipresent in nature, industrial processes and daily life. Understanding their behaviour in inhomogeneous environments (nanopores, self-assembled systems) and in equilibrium as well as non-equilibrium conditions (shear flow) is important in many key applications such as medicine or environmental protection. In this project, we focus on two aqueous solutions: surfactant adsorption on soft surfaces and solubilisation and release of small molecules into and from polymeric structures. In both systems, we also investigate the effect of shear flow on the equilibrium structure and behaviour. We use mesoscopic simulations to provide molecular-level insights into these systems as well as to fill gaps in our understanding of the physical and chemical behaviour of the studied systems.
Researcher: Vladyslav Larionov
Project: The search for the genetic cause of selected Mendelian forms of endocrinopathies: a re-analysis of next-generation sequencing data
Allocation: 50 000 core hours
Abstract: Our group has shown long-term dedication and commitment towards the research of polygenic and multigenic endocrine diseases of childhood. Of these, growth deterioration and childhood-onset diabetes mellitus stand out due to their clear and strong genetic component that however remains only incompletely explored. The search for single gene defects is warranted in families with rare Mendelian inheritance of diabetes, and in familial short stature. We have utilized Next Generation Sequencing (NGS) techniques for a decade and have currently tested DNA of 658 patients. Despite novel findings that we published on both groups of disorders, a large proportion of clinically well-characterized cases is still awaiting their correct genetic diagnosis. To this end, we have prepared an enhanced bioinformatic pipeline following the Genome Analysis Toolkit with emphasis on the detection of structural gene abnormalities. The aim of the proposed project is to re-analyse all our NGS datasets using this pipeline, assuming that additional causative factors will be identified among classes of small deletions, insertions and other structural gene variants. Secondary aims include a formal comparison of several bioinformatic approaches towards mutation detection. The project will contribute to the understanding of physiology and pathology of human growth and pancreas development, and can help modify treatment modalities and outcome.
Researcher: Petra Sukova
Project: Stellar transits through accretion flow
Allocation: 300 000 core hours
Abstract: We will study how the repetitive transits of a star in the close neighbourhood of the supermassive black holes in the centers of Galaxies can influence the process of accretion of plasma. We will use general-relativistic magneto-hydrodynamical code to follow the motion of the gas, while it is being pushed by the star along its trajectory. Some of the gas may be expelled from the accretion disc into the pollar region, where a strong magnetic field ordered along the axis of rotation of the black hole accelerates the gas away from the center up to relativistic speed. On the other hand, the density waves caused by the motion of the star in the medium are spreading both downwards and outwards and hence influence the accretion rate on the black hole and the density distribution in the disc. To see the complicated structure of the outflow and inflow, robust 3D simulations are needed with resolution high enough to capture the details of the flow close to the black hole horizon while stretching far enough to see the outflow on larger scales. We will find the observational traces of the presence of the star close to the nucleus, which will help to find promising targets for the upcoming European space gravitational observatory LISA (Laser Interferometer Space Antenna).
Researcher: Lubomir Riha
Project: SCAlable LAttice Boltzmann Leaps to Exascale (SCALABLE)
Allocation: 1 700 000 core hours
Abstract: Computational Fluid Dynamics (CFD) is the most computationally demanding topic for both aeronautics and automotive industry. For example, over 80% of Airbus and 50% of Renault global High Performance Computing (HPC) resources are today devoted to CFD simulations. Despite efforts in software and hardware improvements, turnaround times and thus induced electric power consumption are still large (up to 2GWh per month for Airbus HPC).
Lattice-Boltzmann Methods (LBM) have emerged as a trustworthy solution for CFD, even reaching a leading position in some key research areas like automotive aerodynamics. A scientific consensus tends to settle, evidencing that in a comparison, i.e. same data retrieved, with same sampling rate, LBM algorithms are roughly one order of magnitude faster than Navier-Stokes-based algorithms for a given accuracy [1], no matter whether Finite Differences, Finite Volumes, or Finite Elements are employed.
Due to algorithmic specificities, LBM has a great potential for achieving better performance with respect to other comparable CFD approaches once properly optimized for today' (x86-type) and tomorrow's (ARM, EPI ...) generation of massively parallel machines.
The primary goal of this project is to provide the computational resources needed to fulfill the objective of related H2020 SCALABLE project which is to develop an industrial grade LBM-based CFD solver capable to exploit current and future extreme scale architectures, expanding capabilities of existing industrial LBM solvers by at least two orders of magnitude in terms of processor cores and in terms of lattice cells, while preserving its accessibility from both an end-user and a physical developer point of view.
Researcher: Zdenek Futera
Project: Redox protein interactions with charged electrodes and their conductivities
Allocation: 1 056 400 core hours
Abstract: Redox-active metalloproteins are designed by Nature to efficiently transfer charge in living organisms where they participate for example in processes like photosynthesis or respiration cycle. Thanks to the unique properties of these biomolecules, metalloproteins can be used in various applications like accurate biosensors, electrocatalysts, or components of nanobioelectronic devices where the proteins are in contact with metallic electrodes. However, complex atomistic details, interfacial electronic structure, and even the exact electron-transfer mechanism on such bio-metallic interfaces are not fully understood yet. While electron hopping is supposed to be a dominant regime in solution, recent single-protein conductivity measurements point to rather controversial coherent long-range tunneling as the undergoing transport mechanism in vacuum protein junctions. Here, we propose a computational project focusing on the comparison of Azurin and small-tetraheme cytochrome (STC) transport properties based on state-of-the-art density functional theory (DFT) calculations. We hope that the project can elucidate some of the raised questions and help to interpret the experimental data.
Researcher: Tomas Martinovic
Project: ACROSS WP7 - Energy and Carbon Sequestration Pilot
Allocation: 82 000 core hours
Abstract: Simulation of fluid flow in porous underground rocks is of fundamental importance for applications such as ground water, oil and gas production, and geothermal energy extraction. CO2 sequestration is a newer application, that is likely to become much more important in the future. Separating CO2 from the exhaust of industrial processes and then injecting that CO2 in the underground will be necessary on a large scale to reach the CO2 emission goals called for by the Paris agreement. This is called CCS, for Carbon Capture and Storage, and our concern here is with the storage part. The most suitable place to store the CO2 is in large aquifers that are not freshwater sources. One prominent such example is the Utsira formation in the North Sea. For more than 20 years, CO2 extracted from the produced natural gas of the Sleipner field has been injected into this aquifer, at a rate of approximately 1 megaton (million ton) per year. In the future, to make a significant contribution to the reduction of CO2 emissions, we must scale this up to the gigaton range (billions of tons).
In this project, we seek to improve the open-source simulator program OPM Flow used to simulate large-scale CO2 storage scenarios. We want to ensure it is capable of handling the much larger storage scenarios needed to scale up the storage rates. Also, we want to improve the ability to simulate directly on processed seismic data with high resolution, reducing human effort in pre-processing.
Researcher: Tomas Martinovic
Project: ACROSS WP5 - Aeronautical pilot
Allocation: 2 146 000 core hours
Abstract: Avio Aero has launched a challenging research activity aimed at significantly improving the feasibility and exploitation of advanced numerical modeling capabilities for critical engine components. The synergy between new generation HPC platforms, Innovative AI techniques and HPDA methods will open new scenery for the design and optimization of aeroengines, enabling innovative investigation strategies and providing unprecedented levels of accuracy and detail.
Avio Aero will leverage the state-of-the-art HPC resources available at IT4Innovations to verify the feasibility of these ambitious objectives. The industrial applicability of last generation HPC, AI techniques and HPDA methods will be investigated by means of two aeronautical engineering case studies: one regarding the combustor and the one referring to aeronautical Low-pressure turbines design. Both numerical investigations are based on complex Computational Fluid Dynamics (CFD) analyses and rely on CPU-intensive and time-consuming routines. The objective is to demonstrate the speed-up opportunities given by state-of-the-art Computing Systems and to develop and deploy an efficient management of numerical results leading to extremely accurate performance prediction never achieved before.
Researcher: Vitezslav Hanzal
Project: AutopilotData2
Allocation: 143 000 core hours
Abstract: As the autonomy and human machine interaction are one of most demanded research areas, development of advanced autonomous flight control system (AFCS) of jet combat/trainer aircraft was extensively investigated in recent projects. As a result, support of autopilot functions using advanced flight control system will be one of the innovative features of L-39NG fast jet trainer.
Essential part of the task is to gather required aerodynamical characteristic of the plane. Extensive campaign of computer simulation of the flow dynamics (CFD) is needed.
Researcher: Vitezslav Hanzal
Project: AutopilotData2
Allocation: 143 000 core hours
Abstract: As the autonomy and human machine interaction are one of most demanded research areas, development of advanced autonomous flight control system (AFCS) of jet combat/trainer aircraft was extensively investigated in recent projects. As a result, support of autopilot functions using advanced flight control system will be one of the innovative features of L-39NG fast jet trainer.
Essential part of the task is to gather required aerodynamical characteristic of the plane. Extensive campaign of computer simulation of the flow dynamics (CFD) is needed.
Researcher: Erik Andris
Project: Neural network potential for zinc-binding sites in peptides
Allocation: 2 292 000 core hours
Abstract: Proteins perform multitude of biological functions in the cell; however, their primary role is to catalyze biochemical reactions. Since they consist of thousands of individual atoms, theoretical studies of their structure and function must employ approximate methods. Among these, molecular mechanics, where the proteins are treated by simple mechanical models, have been very successful. However, in the catalysis of biochemical reactions, proteins often employ metal atoms and these are difficult to be described by the simple models based on classical mechanics. Thus, the enzymes usually have to be treated by quantum mechanics methods, which require lengthy and expensive calculations. Recently, methods based on machine learning have shown great promise in providing fast approximations of complicated behavior, such as that shown by metals in proteins, if they enough training data is provided. In this project, we plan to use reliable quantum-chemical density functional theory and state-of-the-art solvation models to generate such geometrical and thermochemical data for all possible spatial arrangements simulating interactions of amino acids with zinc. We have chosen zinc in particular as the simplest representative of functional metal atoms in proteins. These data will serve as the training data for neural networks and other machine learning methods to create a fast approximate model of interaction of zinc within proteins. This will further improve the understanding of structure and function of zinc-containing proteins.
Researcher: Andreas Erlebach
Project: Reactive transformations of the zeolite-water interface
Allocation: 2 590 400 core hours
Abstract: Zeolites are environmentally friendly, cheap, and commercially available catalysts produced at the Megatonne scale. Therefore, zeolites play a crucial role in developing new catalysts in the emerging field of sustainable chemistry. Their large external and internal interface to water makes the hydrothermal (in)stability the most critical factor for their application in many chemical processes involving water at high temperatures and pressures, e.g., in the biomass conversion. Detailed understanding of the complex chemical processes at the zeolite-water interfaces is a tremendous challenge for theoretical and experimental studies. Therefore, we use a combination of state-of-the-art machine learning models and quantum mechanical simulations in this project to unravel the zeolite interface chemistry in water. The development of neural network potentials will enable the required large-scale simulations with unprecedented accuracy to understand the reaction pathways of the Si-O and Al-O bond hydrolysis. We will apply the new potentials to advanced atomistic simulations for the most industrially relevant zeolites: MFI, BEA, and FAU. Additionally, we will perform the first highly accurate simulations of two-dimensional MFI nanosheets using realistic models. The simulations will bring new insights into catalyst stability under operating conditions, and together with the developed machine-learning models, stimulate future research for the design of more durable catalysts.
Researcher: Andrzej Kadzielawa
Project: Thermal properties of Cerium Titanides
Allocation: 3 457 000 core hours
Abstract: The study of light-active compounds, such as the semiconductors without the doping of the noble transition metals, is the crucial element in designing novel solar panels, harvesting the power of the sun, thus reducing our dependency on the fossil-fuel and effectively reducing the atmospheric CO2 emission. One of the prospective fields are the Cerium - Titanium - Oxygen and Lanthanum - Tantalum - Oxygen systems, exhibiting the susceptibility to the visible light (e.g., the plethora of the compound colors). In this study we examine the monoclinic and orthorombic phases of exemplary CeTiO4, CeTaO4, and LaTaO4 and both its thermal and optical properties, using the ab-initio modeling techniques.
Researcher: Amina Gaffour
Project: Phosphorylation induced chemical shift changes in disordered protein
Allocation: 1 826 000 core hours
Abstract: Intrinsically disordered proteins (IDPs) are post-translationally modified by attaching a phosphate group to serine, threonine, and tyrosine. Phosphorylation or de-phosphorylation of the amino acids causes conformational changes or leads to transitions between ordered and disordered states. Through these mechanisms, many processes at a cellular level are affected including those that result in neurodegenerative diseases, e.g. the Parkinson and Alzheimer disease. We apply a combination of computational techniques to study how phosphorylation affects nuclear magnetic resonance (NMR) chemical shifts. Our computational approach combines molecular dynamics, fragmentation techniques and density functional theory calculations. The project aims to revealthe capacity of the state-of-the-art multiscale computational protocols to reliablypredict CS changes upon phosphorylation and thereby aid the interpretation of experimental NMR spectroscopy data. It will facilitate the structure characterization of IDPs and thus assist in the process of designing drugs for IDPtargets.
Researcher: Jiri Klimes
Project: Accuracy and precision for extended systems VII
Allocation: 4 500 000 core hours
Abstract: Computer simulations have become indispensable for understanding the majority of experimental observations and thus understanding our world. This is especially true at the atomic and molecular level where experimental observations can become very difficult to interpret. However, if simulations are to be useful, they need to model the system reliably. For example, if there are different pathways for a chemical reaction, the simulation needs to be able to correctly predict their relative importance so as to increase our understanding of the system. The world at the atomic level is governed by the complex laws of quantum mechanics. In practice, this means that increasing the reliability of simulations, i.e., using less approximations, increases the computational cost substantially. While we understand well the impact of some of the approximations, there are some for which our knowledge is limited. In some cases, these can lead to unexpectedly large loss of the reliability of calculations. In our project we analyse some of the problematic cases occurring for calculations of intermolecular interactions, either in small clusters, or in molecular crystals.
Researcher: Karel Tuma
Project: A phase-field study of the microstructure evolution and the role of phase compatibility in ternary NiTiPd shape memory alloys
Allocation: 540 000 core hours
Abstract: Martensitic transformation is a first-order solid-solid diffusionless transformation that occurs between the parent phase (austenite) and the product phase (martensite) and is characterized by the microstructure evolution, which is accompanied by the formation, propagation and annihilation of the interfaces. In shape memory alloys, the microstructure evolution is the mechanism behind the properties such as pseudoelasticity and shape memory effect. In general, due to the incompatibility between austenite and a single variant of martensite, the martensitic transformation is realized via the formation of complex twinning microstructure. The resulting transition layer is highly stressed and causes the creation of defects. Macroscopically, this is manifested by a large hysteresis loop. Upon cyclic loading, the defects are accumulated and cause the degradation of the functional fatigue [1].
Experimental findings suggest that by tuning the lattice parameters of SMA, a nearly compatible interface between austenite and martensite can be achieved, which results in a significant reduction of the hysteresis and thus improves the fatigue life. In addition, dramatic changes in the microstructure, e.g. twinless martensite domains, have been reported [2].
Here, the goal is to study the microstructure evolution and the hysteresis behavior of NiTi-based SMAs using a phase-field model. In these alloys, the third component is added and its composition is tuned to achieve phase compatibility.
Researcher: Tomas Martinovic
Project: ACROSS WP6 - Weather & Climate use case
Allocation: 411 000 core hours
Abstract: In the context of the EuroHPC ACROSS project, we are developing new workflows demonstrating pre-exascale scalability.In particular, during this first phase of the project, we will focus on two workflows:The first is hydro-meteorological workflow with Deltares WFLOW executed on 1km model of Rhine and Meuse forced by global deterministic and probabilistic NWP provided by ECMWF.The second workflow is a WRF and WRF-DA regional down-scaling over Europe and Greece of deterministic and probabilistic NWP provided by ECMWF.In addition to those workflows, we will install and configure the MPI ICON model for the execution of high-resolution climatological simulations.
Researcher: Jiri Jaros
Project: Photoacoustic tomography of the breast – Final Evaluation
Allocation: 400 000 core hours
Abstract: Photoacoustic tomography (PAT) is a biomedical imaging modality based on the photoacoustic effect. In PAT, non-ionizing laser pulses are delivered into biological tissues. Some of the delivered energy will be absorbed and converted into heat, leading to transient thermoelastic expansion and thus wideband ultrasonic emission in the low MHz range. The generated ultrasonic waves are detected by ultrasonic transducers and then analyzed to produce images. Since the optical absorption is closely associated with physiological properties, such as hemoglobin concentration and oxygen saturation, the PAT is used to visualize vasculature inside tumors with a very high resolution.
The purpose of this allocation is to evaluate and compare the photoacoustic images of 40 volunteer patients scanned by the Pammoth imager during last year. This study will include tissue heterogeneity into the reconstruction model and use four different laser wave lengths to separate veins and arteries based on the blood oxygen saturation. This study is part of the final image evaluation and moves us towards the deployment in a real PAT system for breast mammography.
Researcher: Kateřina Ruzickova
Project: Slope analysis from geologist view
Allocation: 10 000 core hours
Abstract: The possibilities of different ways, how to move in terrain are done by terrain morphology and kind of soil in space. There should be relationship between kind of soil and terrain slopes, because different soils have different compactness. As a result of morphological processes various shapes of terrain appeared on surface. These project should prove this relationship via slope analysis in geological districts.
Researcher: Mauricio Maldonado Dominguez
Project: Transduction of the kinetic energy redistributed during C-H activation reactions
Allocation: 3 580 000 core hours
Abstract: Nowadays, most organic reactions in solution are visualized through the prism of Eyring's transition state theory, assuming that chemistry occurs through minimal energy pathways with equilibration over the free-energy landscape. Deviations from this statistical description can be, however, commonplace even in solution, where the times for vibrational energy redistribution from a hot substrate can take tenths of picoseconds or even longer. Nature efficiently translates chemical energy into motion in the protein kinesin, and vice versa as in ATP-synthase, and serves as inspiration to leverage on nonequilibrium processes as means to use energy released during chemical reactions.
Albeit appealing, translating motion into useful energy directly from the transition state of simple chemical reactions remains mostly unexplored. In this project, we will validate a computational method developed by us as a tool to diagnose reactions where excess kinetic energy at the TS can be transduced into useful motion, which is a blooming field where only a handful of reactions have been studied and no theoretical tools exist for facile diagnosis on general chemical reactions.
Researcher: Roman Fanta
Project: Toward an accurate description of a benzene dimer by fixed-node diffusion Monte Carlo method
Allocation: 3 800 000 core hours
Abstract: Noncovalent interactions (NI) play key roles in many areas of physics, chemistry, materials science, and beyond. A "gold standard" of quantum chemistry, coupled-cluster (CC), has been used for NI benchmarks. However, due to the prohibitive CPU cost scaling with the system size and poor parallelism, its applications have been limited to comparatively small systems.
A promising competitor of CC for NI in larger systems is the single Slater determinant (SD) fixed-node diffusion Monte Carlo (FNDMC). The single SD FNDMC is a massively parallel low-order polynomial scaling method that meets the benchmark level of CC for many NI. For the use of SD approximation, it is important to uncover the limits of SD FNDMC method.
In this context, we focus here on one of the long-standing questions, that is understanding of the current disagreement between CC and SD FNDMC in stacked parallel displaced benzene dimer complex (Bz2), a representative medium sized dispersion-bound system. To this end, we plan improvement of the CC reference, as well as a screening of to date, overlooked technical parameters that recently emerged as sources of SD FNDMC method bias. The results have the potential to boost FNDMC as a genuine benchmark method scalable to large noncovalent systems, and, in turn, enable an accurate many-body description of NI in systems where CC is out of reach, like large/extended systems.
Researcher: Pavel Jungwirth
Project: Ab-initio molecular dynamics of solvated electron in liquid ammonia with alkali metal cation
Allocation: 2 157 000 core hours
Abstract: Alkali metals dissolve readily in anhydrous liquid ammonia yielding intensely colored solutions, characterized by a fine blue color for dilute solutions and a copper/bronze color for concentrated solutions. The process results from the alkali metal valence electron being detached from the metal and being dissolved by the solvent molecules. These solutions are widely used as reducing agents in organic chemistry, for example in Birch reduction. Ab initio molecular dynamics simulations (AIMD) have already proven to be a useful tool for studying properties of metal-ammonia solutions, successfully describing electron spatial structure and localization in bulk systems as well as its dynamics and vertical detachment energy. All the previous theoretical studies contained only the excess electrons and the solvent molecules. For a better understanding of the real systems, we need to simulate more complex systems, containing not only the ammoniated excess electrons but also alkali metal counter-cation. Note that phase-diagrams of metal-ammonia solutions differ significantly for different alkali metals; this indicates that interaction of alkali metal cations with solvent are important and ion-specific. Because of this, we need to study a series of alkali metals in order to understand, where the excess electron prefers to localize, i.e., in the vicinity of the cation or not. Simulation of the ammoniated excess electrons with one alkali metal cation is the first step to the complete theoretical description of more concentrated metal-ammonia solutions.
Researcher: Marek Pecha
Project: ml4py – distributed machine learning tools (dev/stage)
Allocation: 250 000 core hours
Abstract: Current research related to machine learning commonly prefers deep learning techniques using (deep) neural networks. An associated training phase could be considered as a stochastic process that converges to an expected model. Generally, this stochastic nature arises from employing underlying solvers based on iteration technique using passing randomly selected batches of training samples successively. It helps to decrease a computational cost of a predictor training phase. On the other hand, these techniques lack determinism.
In contrast to deep learning, exploiting conventional techniques allows us to design training pipelines directly and explain the qualities of an attained learning model and numerical behaviour of the underlying solver more straightforwardly. Unlike stochastic optimization solvers, employing deterministic ones provides strictly settled and reproducible pipelines for training predictors. However, these approaches are expensive in the sense of computational cost, and they can efficiently solve up to medium scale tasks by the state-of-the-art. Thus, novel approaches to conventional machine learning techniques, which exploit data parallelism efficiently, are required to reduce the time related to the training phase of predictors.
The concept of the machine learning library called ml4py, developed at VSB-TU Ostrava and Institute of Geonics, provides effective data parallelism using Message Passing Interface (MPI) for conventional technique; the training phase could be accelerated employing stochastic optimization methods as well. It helps to attain high-quality explainable models within a reasonable time. Basically, we find a good candidate for a solution using stochastic approaches expeditiously, and this candidate is used as a hint for deterministic and parallel predictor training subsequently. Unlike standardly used machine learning libraries, the tools in the ml4py library are based on data parallelism natively. It takes the full potential of high-performance computing by means of utilizing the MPI standard instead of a commonly employed orchestration based on running multiple container instances of an application.
Researcher: Jakub Stosek
Project: Towards the prediction of chemical shifts in intrinsically disordered proteins phosphorylated at threonine.
Allocation: 410 000 core hours
Abstract: The objective of the present project is to develop a database of chemical shift (CS) patterns for tripeptide models of intrinsically disordered proteins (IDPs) phosphorylated at threonine residues. The database presents an essential building block in the development of a software (SW) tool for the prediction of NMR CSs. Prediction of CSs is a mandatory step in the structure characterization of IDPs by NMR spectroscopy. The project designs SW that predicts CSs for phosphorylated residues based on CSs computed by quantum mechanical methods. Phosphorylated IDPs can influence regulation of neurodegenerative processes that cause e.g. Alzheimer and Parkinson's diseases. This project helps to accelerate the research on neurodegenerative diseases and contributes to reducing the related research costs and thus it can improve the available medical care.
Researcher: Dominika Maslarova
Project: Enhancement of ultrashort directional X-ray pulses
Allocation: 1 150 000 core hours
Abstract: Short directional X-rays have been extensively studied due to their high potential in many applications such as high-quality phase-contrast imaging of biological samples and spectroscopy at femtosecond timescale. Interaction of very intense laser pulses with plasma medium enables the production of high energy photons in X- and gamma-ray range by several mechanisms, such as betatron radiation, Compton scattering and bremmstahlung. A basic principle is that a relativistic electron radiates light at a very short wavelength, which is a consequence of a relativistic Doppler shift. In the frame of this project, the betatron radiation from laser wakefield acceleration (LWFA) of electrons will be studied. In LWFA, electrons are injected into a plasma wave (wakefield), generated and dragged by a few-tens-of-fs, ultra-intense laser pulse (driver) in optically transparent plasmas. The betatron radiation is naturally generated due to the transverse oscillatory motion of electrons during the acceleration. In order to introduce the LWFA X-ray source fully for practical purposes, the production of more photons is required. In order to address this problem, the aim is to study the radiation enhancement by adding another short laser pulse that collides with the wakefield of the driver and boost the transverse oscillations and energy of the electrons which are accelerated in this wakefield. The research will be carried out by numerical particle-in-cell simulations. Moreover, the results will be experimentally verified in the Extreme Light Laboratory in The University of Nebraska-Lincoln in the USA in the frame of an international collaboration.
Researcher: Marek Stepan
Project: GW Approximation for Conductance of Molecular Junctions
Allocation: 298 000 core hours
Abstract: Miniaturization of the electronic circuits used in digital devices lies at the heart of current technological revolution. A promising strategy is to employ molecules as efficient building blocks, replacing transistors, resistors or contacts by objects as small as few atoms. This effort gave birth to a new interdisciplinary research field of molecular electronics. The challenge for researchers is now to understand and quantify quantum-mechanical effects, such as tunneling or uncertainty principle, which constitute the basic laws that govern electron transport through the molecular devices. These fascinating quantum aspects make research in this field challenging. Supercomputers become a necessary tool for simulating the prospective nanoscopic molecular devices, understanding and predicting their behavior.
Researcher: Dominique Geffroy
Project: Dynamics of the exotic Hunds-Mott transition in antiferromagnetic materials
Allocation: 3 000 000 core hours
Abstract: Antiferromagnetic metals are raising to prominence in spintronic applications. Considering that they are relatively rare in nature, their physics has only been marginally explored so far. On the one hand, they are metals, so that their low-energy behavior involves one-particle (electrons, holes) excitations. On the other hand, they are magnetically ordered, so that the two-particle collective modes (magnons) are relevant as well. The simultaneous description of one- and two-particle excitations in correlated materials is a challenging task. Dynamical mean-field theory has the capability to reach it, while being sufficiently material specific. Nevertheless, the calculation of two-particle spectra in material specific models requires innovative numerical techniques, which support the multi-index high-dimensional tensor objects involved in the solution of the so called Bethe-Salpeter equations. In this project, we will employ our recently developed solver to study simple antiferromagnetic metals and materials with magnetically driven metal-insulator transition.
Researcher: David Zihala
Project: Comprehensive transcriptomic study of extramedullary disease in multiple myeloma
Allocation: 40 000 core hours
Abstract: Multiple myeloma is the second most common blood cancer caused by the expansion of antibody-producing immune cells in the bone marrow. The symptoms include anaemia, bone lesions, renal failure, and patients eventually die despite the treatment. The modern treatment is focused on the molecular pathways specific for the cancer cells and helped to extend patients life significantly. However, the more aggressive stages of this disease have a chance to develop during the patients' lifetime. Extramedullary myeloma (EMD) is a more aggressive form of myeloma accompanied by extensive drug resistance and short survival. The understanding of molecular pathways typical for this stage could help to design better therapies. We have unique access to the tumor samples of EMD, and we believe that a comprehensive analysis of gene expression will allow us to identify novel therapeutic targets.
Researcher: Alberto Marmodoro
Project: GatMag
Allocation: 700 000 core hours
Abstract: We intend to study from first principles the problem of magnon spectroscopy and its functionalization in possible thin film devices, in combination with a gating electric field.
Researcher: Martin Matys
Project: Laser-driven ion acceleration using structured targets II
Allocation: 851 000 core hours
Abstract: Laser-plasma ion accelerators are currently receiving particular scientific attention as promising source of accelerated charged particles, since they are able to generate much stronger electric fields in comparison with conventional accelerators and can possibly replace them in future in several impressive applications, including proton therapy for the treatment of the cancer cells, production of PET (positron emission tomography) medical isotopes, generation of ultrashort neutron pulses, radio isotope source, etc. Currently, laser-driven ion acceleration still needs to face several challenges, like further improvement of produced particle beam quality and properties. Therefore, a novel scheme for ion acceleration is investigated in this project. Interaction of high-intensity laser pulse with thin overdense double layer targets with initial corrugation on the interface, results into controlled rupturing of the foil. The remaining bunches are then accelerated as whole well-collimated structures, exhibiting monoenergetic behavior. Results from our demanding 2D and 3D simulations will also be visualized in 3D virtual reality web application to understand the whole picture of multidimensional effects and used for promotion of laser-plasma physics to the general public.
Researcher: Martin Å Rejber
Project: Engineered lipids in vaccine research
Allocation: 1 937 000 core hours
Abstract: In the last decades, liposomal drug delivery became one of the most investigated delivery systems in multiple fields like cancer chemotherapy, gene therapy, liposome-entrapped drug delivery or vaccines. This unprecedented interest in drug delivery reached a new urgency due to the global COVID-19 pandemic. In an unparalleled effort uniting scientific community, several vaccine mechanisms were proposed, one of them being mRNA vaccines. The immunogenic messenger RNA coding SARS-CoV-2 spike protein is enveloped by a liposome consisting of a mixture of natural and engineered lipids. These lipid nanoparticles (LNPs) are tailored to attain desired features of antigen carriers. However, the full nature of the liposome-mRNA interactions is not yet understood - and can become a focus of in silico studies. The main aim of this project is to provide a rationale for the complex action of LNPs self-assembly in presence of mRNA, mimicking the process of vaccine preparation. Thus, we plan to focus on investigating the effect of clinically designed lipids on the membrane properties. Furthermore, we focus on rationalizing the specific mRNA-lipid nanoparticle interaction that presumably plays crucial role in stabilization of immunogenic messenger RNA.
Researcher: Alberto Marmodoro
Project: RE-TM_magnetism
Allocation: 600 000 core hours
Abstract: We will study magnetic materials containing both rare-earth and transition metal elements by means of a multi-scale combination of first-principles electronic structure and atomistic spin dynamics methods. We will explore suitable corrections on top of standard exchange-correlation DFT functionals, and the range of validity for approximations and numerical treatments for the solution of problems of laser induced, heat-assisted switching.
Researcher: Marta Cudova
Project: Offloading of Workflows Executions to Remote Computational Resources
Allocation: 271 000 core hours
Abstract: In recent years, the therapeutic ultrasound has grown in a number of applications such as tumor ablation, targeted drug delivery or neurostimulation. Precise preoperative treatment planning tailored individually to each patient is crucial for the maximalization of the treatment's outcome.
The fundamental challenge shared by all applications of therapeutic ultrasound is that the ultrasound energy must be delivered accurately, safely, and noninvasively to the target region within the body identified by a medical doctor. The estimation of treatment outcome heavily depends on the computationally very intensive ultrasound, thermal and tissue models which are only realizable with the use of HPC facilities.
Since there is a lack in expertise of the clinical end-users to use HPC resources efficiently, we developed the k-Plan software simplifying the everyday use of the HPC resources without a need to specify task execution parameters, dependencies, and their monitoring. k-Plan also reacts to the problem of task execution parameters selection, e.g., number of nodes, estimated execution time, storage space, etc. The end-users have no knowledge about the strong and weak scaling of the software being used, yet these characteristics have great impact on the calculation cost and overall execution time.
The goals of this project are to (1) deploy the k-Plan software and offer it to a small set of pilot end users from clinical environment to execute realistic treatment plans, (2) to tailor the task submission planning logic to IT4Innovations clusters, (3) and investigate methods that allow k-Plan to automatically tune execution parameters for individual tasks.
Researcher: Tomas Panek
Project: Characterization of mitochondrial metabolism and phylogenetic position of a novel deep-branching eukaryote
Allocation: 215 000 core hours
Abstract: Eukaryotes are currently classified into several supergroups, most of them containing exclusively unicellular lineages referred to as protists. Because protists constitute majority of deep eukaryotic lineages, they are crucial for understanding the eukaryotic evolution. Advances in molecular approaches achieved during the last 20 years allowed scientists to study some of them in a great detail. However, our knowledge of protistan diversity and cell metabolism is still poor compare to typical model eukaryotes (animals, fungi, and land plants). This significantly limits our understanding of the evolution of the eukaryotic cell, its organelles, and metabolic pathways. We isolated and cultured a novel, deep-branching eukaryotic lineage (strain SUM-K), which has a high number of unique traits, including possibly plesiomorphic mitochondrion. Using IT4i computational resources, we will carry out a detailed in silico investigation of transcriptomic and genomic data to predict mitoproteome of the organism and reveal its phylogenetic position within eukaryotes, using methods of molecular phylogenetics and phylogenomics. The results should significantly advance our knowledge of the eukaryote evolution.
Researcher: Martin Zeleny
Project: Ab initio study of exchange interactions at grain boundaries
Allocation: 143 000 core hours
Abstract: As the autonomy and human machine interaction are one of most demanded research areas, development of advanced autonomous flight control system (AFCS) of jet combat/trainer aircraft was extensively investigated in recent projects. As a result, support of autopilot functions using advanced flight control system will be one of the innovative features of L-39NG fast jet trainer.
Essential part of the task is to gather required aerodynamical characteristic of the plane. Extensive campaign of computer simulation of the flow dynamics (CFD) is needed.
Researcher: Libor Veis
Project: Electronic structure calculations of novel polycyclic aromatic hydrocarbons
Allocation: 2 500 000 core hours
Abstract: In this project, we will study the electronic structure properties of novel polycyclic aromatic hydrocarbons (PAHs) prepared via the ultra-high vacuum (UHV) on-surface chemical synthesis [1] and characterized experimentally by means of the scanning probe microscopy with the unprecedented sub-molecular resolution [2]. The main computational tool will be our massively parallel implementation of the density matrix renormalization group (DMRG) method [3], which is a method of choice for strongly correlated systems requiring very large active spaces. In particular, we will study different rhombene-like derivates whose molecular topology largely influences their electronic (and magnetic) properties, and super-ring annulene-within-annulene (AWA) structures exhibiting peculiar global aromatic character, which contradicts the well-known Huckel's [4n + 2] rule [4]. Our state-of-the-art calculations are expected to provide a thorough theoretical insight into the structure-properties relationship and help to explain the unusual electronic properties of these PAHs.
Researcher: Ales Vitek
Project: Atomic and molecular clusters: computationally demanding Monte Carlo simulations
Allocation: 889 000 core hours
Abstract: This project will be focused on computationally demanding classical Monte Carlo simulations of atomic and molecular clusters. Particularly, we will focus on mercury clusters and water clusters. Recently, we have published a few papers about mercury clusters [1,2,3], we computed their photo absorption spectra or thermodynamic properties, but now, we have found very interesting phenomena on Hg13 cluster, which has thanks to the crossing of energy levels two most stable isomers with the same structure, but different size and only the external pressure decides, which isomer will be the global minimum on the enthalpy hypersurface. The second our aim is the simulation of big water clusters under various thermodynamic properties and testing and developing of the inclusion of the external pressure for finite systems.
Researcher: Jakub Sebesta
Project: Magnetic ordering and magnetoelastic behavior of Fe,Co based Laves phases
Allocation: 3 000 000 core hours
Abstract: Magnetic materials stand for a crucial part of contemporary electronic engineering, lots of the strong magnetic materials rely on rare-earth elements. They bear strong localized magnetism and allows one to make small magnets or to achieve high magnetic fields. However, the rare-earth elements occur in ores only with low concentration. This makes their extraction demanding and expensive. Thus, we'd like to focus on the d-elements based Laves phases, which seem to be able to provide strong magnetic behavior with great application potential. Particularly, we are interested in the Fe, Co based intermetallic compounds (Fe2Ta, Fe2Ti derivatives etc.) formation of the so-called Laves phases, their ground-state magnetic ordering employing the ab-initio calculations. In addition, the magnetoelastic properties, the magnetostriction, and magnetic anisotropic energy of the discussed compounds will be studied since their importance for practical applications. Finally, calculating the magnetic pair exchange interactions, the magnetic ordering temperatures will be evaluated to determine whether the magnetic ordering is achievable at the room temperatures.
Researcher: Jiri Kaleta
Project: Self-Assembled Arrays of Molecular Machines on the Water/Air Interface
Allocation: 800 000 core hours
Abstract: Rational design of sophisticated molecular-level machines is a key strategy towards the new generation of smart materials with unique properties and applications, for example in nanoelectronics, information storage, artificial propulsion units of microscopic object, solar energy harvesting etc. The term molecular machines involves various (photo)switches, (unidirectional) molecular motors, turnstiles, etc., which can repetitively respond on the external stimuli. Some of these materials are intensively studied in our laboratories, with a focus on 2-D films: systems which are on the borderline between solids and liquids. They combine the ability of unrestricted switching characteristic for molecules in solution with a high degree of organization typical for crystalline materials. Also, special attention is placed on the 3-D structure and possible intramolecular interactions of individual molecules, which deeply impact their self-assembly properties and structures of final 2-D self-assembled monolayers. One of our approaches is based on the Langmuir-Blodgett technique generating SAMs on the water/air interface. Herein proposed molecular dynamic simulations will provide crucial information regarding these SAMs; namely polymorphism, area per molecule, tilting/bending of individual molecules, impact of ions on monolayer stability, etc.
Researcher: Tomas Martinek
Project: Benchmarking Sulfonate-Cation Force Field Parameters to Improve the Modeling of Glycosaminoglycans
Allocation: 1 861 000 core hours
Abstract: Glycocalyx is located above the cell membrane and consists of long polysaccharide chains called glycosaminoglycans. The sulfation of glycosaminoglycans determines their activity, selectivity, and functionality. Proteins rich in amines and positively charged cations like calcium are suggested to have high affinity for these glycosaminoglycans, particularly for specific sulfated motifs. However, we know little about these interactions because current molecular models suffer from inaccurate description of electrostatics. Here, we aim to address this shortcoming by benchmarking molecular models of methylammonium and methylsulfonate that mimic protein and glycosaminoglycan charged groups, respectively. We will use ab initio molecular dynamics simulations to obtain the free energy profiles of methylsulfonate-methylammonium and methylsulfonate-calcium interactions. These data will be adopted for testing and improving the parameters of classical molecular dynamics force fields as no experimental counterpart is available. The newly derived parameters will enable more accurate large-scale molecular dynamics simulations of glycosaminoglycans providing a more realistic picture of outer-cell interactions.
Researcher: Jan Novotny
Project: Paramagnetic NMR spectroscopy scrutinized by relativistic quantum chemistry
Allocation: 2 907 000 core hours
Abstract: The relativistic effects are known to influence dramatically magnetic-resonance properties of heavy-elements compounds, but their chemical interpretations are notably underexplored. The paramagnetic NMR spectroscopy is becoming indispensable analytical technique in modern biological and materials research of open-shell systems (used as redox catalysts, metallodrugs, contrast agents). However, its further development calls for novel tools, concepts, and transparent interpretations of relativistic effects on the NMR shifts.
In this project we will use computational resources provided by IT4I infrastructure to apply relativistic methodology for calculation and analysis of magnetic response properties in paramagnetic coordination compounds of various complexity (starting from small model molecules up to ensembles of supramolecules with explicit solvation). We are planning to employ new protocols for characterization of complexes in higher spin state or exhibiting strong spin-orbit interaction and analyse fundamental terms contributing to hyperfine shielding.
Researcher: Luigi Cigarini
Project: Effects of strain on the stability of beta-PN
Allocation: 285 000 core hours
Abstract: Theoretical physics can be useful for technological applications in the field of electronics, energy harvesting and conversion, and even new generation devices which could be able to use light as a carrier of information. Optical computing, could be an exciting result of this kind of studies and increase the performances of current computers, using light instead of electron to carry information inside the CPU of computers. A huge number of futuristic applications could be listed, but a deep understanding of the physics involved is necessary, and it will be reached only by numerical simulation, which can be performed in the current electronic calculators.
Researcher: Radovan Vselicha
Project: Advanced jet engine for aircraft use
Allocation: 171 000 core hours
Abstract: This project is focused on design a turbine gas engines for aerospace applications, where VZLU is responsible for combustion chamber design. Application of the experimental method is limited due to the conditions inside the chamber(e.g. shape of the chamber, temperature over 2000 K, high pressure, ...). Therefore, the application of the computational method is necessary. Modeling of liquid fuel combustion is quite complex because it is necessary to simulate the highly unsteady turbulent flow field, heat transfer, particle motion, and chemistry at the same time. There is also a strong interaction between simulated phenomena which are limiting the application of a suitable computational method. For this applications are asked accurate model of turbulent flow field which can be obtained by using complex turbulence models such as LES or hybrid RANS-LES turbulence model. These project is supported by by the Technology Agency of Czech Republic (grant no. FW01010519).