Výsledky 17. kola Veřejné grantové soutěže

Všem žadatelům o výpočetní čas v rámci 17. Veřejné grantové soutěže děkujeme za podání žádostí.

Uchazeči o výpočetní zdroje požádali v rámci této výzvy do následujícího alokačního období o 115 673 068 jádrohodin. Požadované zdroje převyšují vyhrazenou kapacitu 66 000 000 o 75%. S ohledem na vysoký zájem o výpočetní čas v poměru k nabízeným zdrojům, přistoupila alokační komise ke snížení alokací hodnocených projektů . Redukce se úměrně dotkly všech projektů.

Všechny projekty prošly technickým hodnocením a hodnocením počtu registrovaných publikačních výstupů k počtu předchozích žádostí PI.

O alokacích komise rozhodla na základě výsledků hodnocení vědecké excelence, výpočetní připravenosti a společenské užitečnosti. Váhou 50 % přihlédla také k poměru počtu registrovaných publikačních výstupů k počtu předchozích žádostí. V průměru komise očekává 0.75 publikace na finalizovaný projekt. Projekty v rozsahu 7% požadovaných zdrojů nebyly předmětem peer-review nezávislými hodnotiteli. V těchto případech komise předpokládala nejvyšší možné bodové hodnocení vědecké excelence, výpočetní připravenosti a společenské užitečnosti. U projektů nových uživatelů komise předpokládala nejvyšší možné hodnocení počtu registrovaných publikačních výstupů k počtu předchozích žádostí.
Nízký počet registrovaných publikací v poměru k počtu předchozích projektů byl hlavní příčinou výrazného krácení alokací.

Komise konstatovala velmi dobrou vědeckou a technickou úroveň hodnocených projektů, kde z maximálního dosažitelného počtu 30 bodů bylo průměrně dosaženo 24,8 bodů, minimálně 20 bodů. Deset z hodnocených projektů dosáhlo více než 28 bodů.

Mezi 65 projektů, z toho 7 multi-year, bylo rozděleno celkem 85 003 000 jádrohodin. Rozdíl oproti, pro veřejnou grantovou soutěž vyhrazené kapacitě, byl vykompenzován rezervami ředitelství IT4Innovations a kapacitou nově instalovaných výpočetních systémů. Projekty, které nebyly podrobeny peer-review, zastupují celkem 7 206 000 (8%) z celkového počtu alokovaných jádrohodin.

Zpoždění rozdělení výpočetních zdrojů oproti plánu bylo způsobeno zdržením peer-review procesu.

Alokační komise v 17. kole Veřejné grantové soutěže IT4Innovations rozdělila výpočetní zdroje takto:

Hlavní řešitel: Jiri Klimes
Projekt: Accuracy and precision for extended systems III
Alokace: 1 873 000 jádrohodin
Abstrakt: Materials bound by non-covalent interactions are important both in nature and industries. From methane clathrates at the bottom of the sea, over pharmaceuticals in pills, togm layered systems such as graphite. Many of them have also some peculiar properties. For example, even at the same conditions, many pharmaceuticals can exist in different crystal structures, called polymorphs. One of these polymorphs is the most stable, but the others are usually very close in energy. One of the long term goals of our project is to develop a theoretical modelling scheme that would allow a reliable description of the stability of the different polymorphs or different phases of materials in general. The problem is that to get the tiny energy differences between different phases, we need to use quantum mechanics. Solving the equations of quantum mechanics is only possible for simple systems and for extended systems we need to use approximations. To reach our goal we want to combine one of the most accurate schemes currently available for the treatment of extended systems with a method used to calculate reference quality binding energies of molecules. Moreover, we want to develop methods that will ensure that our results are very precise and thus reproducible. This will enable us to obtain highly reliable binding energies of extended systems.

Hlavní řešitel: Karel Tuma
Projekt: Large scale three-dimensional numerical simulations of the microstructure evolution in the NiAl shape memory alloy
Alokace: 292 000 jádrohodin
Abstrakt: Shape memory alloys are very interesting materials known for their ability to exhibit two extraordinary phenomena. The first one is the shape memory effect in which the material “remembers” its original shape and it is able to return back to it after it is deformed. The second is the superelasticity effect when the material acts as an elastic material that can undergo reversible severe deformation upto several percents. Both of these effects are caused by so-called martensitic transformation when during the deformation one phase changes to another phase. This is the reason why it is important to understand what is happening on the microscopic level. The material can exist in different phases that can transform between each other. In order to capture several phases we are using the phase-field method in which every phase is described by the evolving order parameter. The main advantage of this method is that we do not need to track the interfaces and they are obtained automatically by solving the corresponding differential equations. On the other hand, this additional unknown increases the size of the discrete problem. Thus, to obtain a good resolution of the microstructure, a big numerical problem has to be solved.

Hlavní řešitel: Karel Tuma
Projekt: Large scale three-dimensional numerical simulations of the microstructure evolution in the NiAl shape memory alloy: test of preconditioners
Alokace: 125 000 jádrohodin
Abstrakt: Shape memory alloys are very interesting materials known for their ability to exhibit two extraordinary phenomena. The first one is the shape memory effect in which the material “remembers” its original shape and it is able to return back to it after it is deformed. The second is the superelasticity effect when the material acts as an elastic material that can undergo reversible severe deformation upto several percents. Both of these effects are caused by so-called martensitic transformation when during the deformation one phase changes to another phase. This is the reason why it is important to understand what is happening on the microscopic level. The material can exist in different phases that can transform between each other. In order to capture several phases we are using the phase-field method in which every phase is described by the evolving order parameter. The main advantage of this method is that we do not need to track the interfaces and they are obtained automatically by solving the corresponding differential equations. On the other hand, this additional unknown increases the size of the discrete problem. Thus, to obtain a good resolution of the microstructure, a big numerical problem has to be solved.

Hlavní řešitel: Martin Beseda
Projekt: Investigation of significant Nitrogen compounds’ properties
Alokace: 1 400 000 jádrohodin
Abstrakt: This project is a collaboration of IT4Innovations and UFCH-JH on the problems, they‘ve been both researching in their previous projects and which are focused on Nitrogen compounds present in the two of nowadays Physical Chemistry „hot topics“ – Plasma Medicine and Planetary Atmospheric Physics. Considering the former one, it was shown experimentally, that rare-gas plasmas are well-working in medical applications. To understand the healing properties of cold rare-gas plasmas, however, detailed knowledge of processes is of crucial importance. We are mainly focused on the interaction of He with N2, i.e. this part is a direct continuation of the OPEN-14-25 project. Considering the computational demands of [N2/He]+ complex, the project will contain a thorough convergence analysis, taking different active spaces, basis sets and point groups available for a given geometry, into consideration. Talking about the latter part of the project, the inspiration there arose from the very successful Cassini-Huygens mission to Saturn´s moon Titan. One of the most surprising results of this endeavor was the existence of large negative ions in Titan’s upper atmosphere. Thus, the project aims for the theoretical treatment of selected important reactions of negatively charged molecular ions, mostly C2N- and C4N- with acetylene. These processes have already been studied experimentally at the UFCH-JH and the proposed calculations are necessary to rationalize the obtained findings.

Hlavní řešitel: Martin Beseda
Projekt: HPC-oriented machine-learning representation of potential energy surfaces
Alokace: 167 000 jádrohodin
Abstrakt: This project is focused on potential energy surfaces (PESs) representations using machine-learning methods, specifically artificial neural networks. As such it is a direct continuation of previous OPEN-14-25 and OPEN-16-14 projects. The main motivation of the project are common troubles with potential energy surfaces computed ab initio: 1) missing points and 2) excessive computational demands. Machine learning seems to be a plausible way to overcome both of these obstacles – if a model is trained on a sufficiently dense grid of ab initio data, then we are able to predict even the values among those training points, so we can “fill” numerically problematic areas and predict as many points as needed, i.e. get arbitrarily dense grid. Utilizing different approaches we will test them on two types of Helium compounds – [N2/He]+ collision complex and He21+ cluster. The former is very important in a field of plasma medicine and represents a smaller heterogeneous molecule, while the latter is a medium-to-large sized homogeneous cluster, needing a different modeling approach, presumably. Our software libraries lib4neuro and neuron4dyn will be used and the implementations will be performed with a focus on shared-memory parallelization, utilizing both OpenMP and OpenACC and comparing the two parallelization approaches subsequently.

Hlavní řešitel: Huseyin Sener Sen
Projekt: Strength Tuning Atomistic Research
Alokace: 2 877 000 jádrohodin
Abstrakt: Technologically, super-hard materials are of great interest due to their importance in a wide range of industrial applications, such as aerospace and nuclear technology. However, it is still challenging to theorethically describe the resistance of a material to deformation – namely hardness, due its inherent mechanical comlexity. The search of super-hard materials is driven also by scientific curiosity as it is important for understanding the fundamental correlations between microscopic characteristics of inter-atomic interactions and macroscopic properties. Henceforth, today, the quest to novel superhard materials is of great interest for modern science and technology. Our project falls under the umbrella of developing new metal alloys with enhanced and/or multi-functional properties. We aim to explore different strategies to optimize material performance by collectively tailoring chemistry and microstructure in our systems. Much of the science lies in understanding the nanoscale mechanisms that govern the properties we measure at the macroscale. This project will allow us to understand such mechanisms and enable us to build materials with extreme properties, in this case super-hard materials.

Hlavní řešitel: Fabien Jaulmes
Projekt: Computational modelling of ion orbits in tokamak plasmas.
Alokace: 767 000 jádrohodin
Abstrakt: 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. Among the approaches to fusion – tokamak (russian abrev.: ”toroidal chamber with magnetic coils”) seems to be the most promising one. The 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 with first plasma currently scheduled for December 2025. If successful, the device would be the first one of its kind to produce net energy. COMPASS is a small tokamak located in Prague, Czech Republic. It allows scientific investigation of various physics related to the operation of the future ITER. In particular, a new 80keV Neutral Beam Injection system is planned to be tested next year at Institute of Plasma Physics (IPP). The study and modelling of NBI-born particle behaviour is of great relevance and might impact future design of the system and its integration in COMPASS, as well as in the planned upgrade of the machine in 2022 (COMPASS-Upgrade). An example of calculation performed in 2019 with the support of the IT4I cluster is attached as image with this request.

Hlavní řešitel: Dominika Maslarova
Projekt: Betatron radiation enhancement in laser wakefield accelerators
Alokace: 792 000 jádrohodin
Abstrakt: Laser wakefield acceleration (LWFA) is currently considered as one of the most promising mechanisms to potentially reduce the size and cost of future electron accelerators. In this technique, plasma electrons are injected into a plasma wave (wakefield), generated and dragged by an ultra-short, ultra-intense laser pulse in optically transparent plasma. Such electrons gain relativistic energy within a few millimeters, which is three orders of magnitude lower than the contemporary technology. In addition, the accelerated electrons are also capable of producing ultra-short, coherent, so-called betatron radiation in the X-ray domain. Such X-ray pulses are interesting in the terms of applications, e.g., high-quality phase-contrast imaging of biological samples. In order to introduce LWFA X-ray sources 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 modifications in plasma density. The research will be carried out by numerical particle-in-cell (PIC) simulations for standard parameters feasible with current sub-100 TW laser systems, which ensures a fast introduction of this technology to applications. This approach enables developing and optimizing the design of laser wakefield accelerators without repetitive expensive and time demanding experiments.

Hlavní řešitel: Jan Řezáč
Projekt: Benchmark database for the development of next-generation approximate computational chemistry methods
Alokace: 2 255 000 jádrohodin
Abstrakt: To apply computational chemistry to real-world chemical problems, it is often necessary to work with large systems with thousands of atoms. This is especially true in the two currently most prominent research directions, in the applications of computational methods to biochemistry and to (nano)materials. This requires the use of approximate methods, often including empirical parameters. The development of such method then relies on accurate reference data that can be used for their parametrization and validation. Also, the emerging applications of machine learning to molecular systems require enormous amounts of data for their development This proposal is a part of a larger project that aims to build a state-of-the-art database of accurate calculations that can serve for this propose. Here, we focus on non-covalent interactions, an effect of key importance in larger molecular systems. The new database will be an order of magnitude larger than these available today (such as the successful S66 I have developed in the past), and provide systematic coverage of broader chemical space. It will feature hundreds of systems, with multiple geometries for each, calculated using the CCSD(T)/CBS methodolgy which is considered to be the “gold standard” in the field. Based on the success of the smaller data sets I have published in the past, It can be assumed that the new database will become widely adopted, and that the improvements will enable even more applications.

Hlavní řešitel: Mojmir Sob
Projekt: From antiphase boundaries to new rare-earth-free magnets
Alokace: 7 964 000 jádrohodin
Abstrakt: Our modern industrialized society critically needs new magnetic materials in order to address numerous challenges related to, for example, electromobility, environmentally-friendly refrigeration as well as other industrial sectors. These new materials should preferably not contain rare-earth elements as their production is environmentally unfriendly and quite unpredictable due to complex geopolitical situation. Therefore, there is a renewed interest in conventional magnetic materials based on iron and its alloys. Importantly, new mechanisms are needed to enhance properties of these conventional magnets. There is an experimental evidence (Nature Comm. 5 (2014) 4133, Ref. [1]) that antiphase boundaries (APBs) in Fe-Al alloys have their local magnetic flux density by 60% higher than the surrounding APB-free material but the underlying physical mechanism is unclear. The proposed project will use theoretical tools based on quantum-mechanical calculations to identify the origin of this phenomenon.

Hlavní řešitel: Jakub Sistek
Projekt: Advanced incompressible flow simulations for vortex identification
Alokace: 333 000 jádrohodin v první periodě
Abstrakt: The main aim of the project is performing high-resolution computational fluid dynamics simulations of prototype problems of incompressible viscous flows. The primary goal of these simulations is to generate high-resolution 3D data with vortical structures, which will be subsequently used for development of new methods for flow-field analysis and vortex identification and visualization supporting a research project granted by the Czech Science Foundation. Unsteady flows considering very fine computational meshes are required for this purpose. The computations will be performed using an in-house parallel finite element solver based on the pressure correction method and multilevel domain decomposition, with the aid of an existing parallel implementation of the method in the open-source BDDCML library. A subsequent goal of the project is further development of the computational method and optimization of the BDDCML library for large numbers of computer cores and GPU accelerators.

Hlavní řešitel: Athanasios Koliogiorgos
Projekt: Quantum dot and nanorod perovskite nanostructures for photovoltaics (QDNAP)
Alokace: 2 418 000 jádrohodin
Abstrakt: In the quest for clean and abundant sources of energy to substitute the limited and environmentally burdening fossil fuels, photovoltaic (PV) materials play a central role. The research for suitable materials for efficient PVs is at a peak. Quantum dots, or nanoparticles, are among the structures that can be used in solar cell devices, exploiting their exceptional electronic and optical properties derived from their quantum confinement phenomena. At the same time, hybrid halide perovskites are among the most intensively studied materials for use in PVs, thanks to their characteristics as high absorption, suitable band gaps and low cost of synthesis. The synthesis of perovskite QDs has gained ground, but the computational literature on this field is still quite limited. In this project, we study the combined properties of nanostructures and halide perovskites, through ab-initio calculation of the structural, electronic and optical properties of a wide array of halide perovskite nanostructures (QDs and nanorods), with focus on lead-free materials, in order to determine suitable candidates for use in PV applications.

Hlavní řešitel: Antonio Cammarata
Projekt: Assisted Layer EXfoliation of transition metal dichalcogenides (ALEX)
Alokace: 3 232 000 jádrohodin
Abstrakt: After the low-hanging graphene fruits have already been harvested, layered transition metal dichalchogenides (TMD) represent a new avenue to traverse, in order to exploit the vast and exciting properties of mono- and few-layer two-dimensional materials. To this aim, high-quality large-scale production of TMD thin films with controlled number of layers is mandatory1-6. In this perspective, the present project will advance the knowledge on the atomic-scale mechanism governing the exfoliation phenomenon in TMDs. The focus will be on the comprehension of atomic interactions governing friction at the nanoscale in TMDs, in order to control the frictional behaviour hence layer exfoliation by using external electrostatic fields and exploiting the presence of small contaminant molecules. Density functional calculations will be used together with advanced methods of electronic and phonon analysis, in order to predict and suggest new structure/composition configurations to ultimately design novel 2D TMD-based materials. Project outcomes will allow to formulate novel experimental frameworks to produce mono- and few-layer TMD films on a large scale with controlled number of layers. Thanks to the features of the elementary physical processes involved in the layer exfoliation phenomenon, the acquired knowledge can be promptly exploited for thin film production of other inorganic graphene analogues, like boron nitride, borocarbonitrides, metal oxides, and metal-organic frameworks.

Hlavní řešitel: Miroslav Šoóš
Projekt: Desolvation kinetics of pharmaceutical solvates
Alokace: 417 000 jádrohodin
Abstrakt: The active pharmaceutical ingredients (APIs) are commonly prepared by their crystallization from suitable solvents. An API might crystalize while keeping the solvent molecules embedded in the crystal structure, producing thus API solvates. Moreover, the desolvation of such solvates can result in polymorphic transformations of the original crystal lattice. Understanding the conditions leading to polymorphic transformations is of great industrial and scientific interest since different polymorphs of an API can exhibit different physicochemical properties. Therefore, modeling desolvation processes may provide a first insight to the consequent formation of the various polymorphs. In this work, we will use molecular dynamics (MD) to investigate the desolvation of four different Ibrutinib (an anti-cancer API) solvates. We will study temperature and initial structure transformation effects on the desolvation kinetics, calculate solvent diffusion coefficients and elucidate the very early stages of the polymorphic transformations. Finally, we will combine this microscopic information with the macroscopic solid-state reaction kinetics modeling performed on experimental desolvation kinetics data obtained by powder X-ray diffraction (XRPD) and thermogravimetric analysis (TGA).

Hlavní řešitel: Lubomir Rulisek
Projekt: Conformational Behaviour of Small Peptide Fragments Studied by the Quantum Chemical Methods
Alokace: 5 416 000 jádrohodin
Abstrakt: To What Extent Conformational Strain in Proteins Determine Their Three-dimensional Structure? Large-scale quantum chemical calculations coupled with modern solvation methods represent unique set of tools to elucidate key determinants of the biomolecular structure ab initio. Understanding conformational strain in proteins and in their ligands may represent a new and computationally tractable way how to significantly deepen our understanding of protein folding and of protein-ligand interactions. In addition to computed data, an experimentally verifiable set of tests is defined in the proposal to provide evidence for the suggested hypotheses.

Hlavní řešitel: Michal Podhoranyi
Projekt: PaReTran4
Alokace: 417 000 jádrohodin
Abstrakt: The main objective of the project is to develop a reactive-transport model able to utilize HPC resources. The primary purpose is constructing a mathematical representation of a proposed reactive-transport system in order to simulate the potential risk of environmental contamination. Additionally, the contribution of the project is not only associated with HPC usage but also with new model features implemented during the developing phase. Overall, the Transport-Reaction Model (TRM) will be developed to include complex functionality that is necessary in order to solve specific transport-reaction issues. TRM will be based on coupling the PhreeqcRM geochemical library with 2D solute species transport in water on a regular rectangular network of elements. Compared to the other similar models, our model will offer a unique feature that is associated with the 2D mesh. This feature represents an innovative component that improves modelling results.

Hlavní řešitel: Rene Kalus
Projekt: Formation of molecular ions in rare-gas plasmas
Alokace: 417 000 jádrohodin
Abstrakt: Molecular ions, ionic dimers in particular, play an important role in cold rare-gas plasmas and a lot of attention has been paid in our group to their collisions with carrier gas atoms. Both electronically ground-state dimers and electronically excited dimers have been studied within a series of previously solved OPEN projects with all the rare gases (He – Xe) taken into account. The preceding computations revealed, however, a principal question, not considered up to now theoretically, on the way how molecular (dimer) ions are formed from atomic ions. In particular, the knowledge of typical electronic and/or rotational-vibrational excitations in their nascent populations is of principal importance and represents a real challenge for the theory since while there are some experimental data on the rate of molecular ions formation, the nascent excitations cannot be extracted from experiments and numerical modeling represents the only practicable way. The main intention of the present project is to initialize research in this field, i.e. a) to calculate rate constants of the dimer ions formation and compare them with experiments (to validate theoretical approaches used) and b) to perform a deep analysis of computational data to extract the information about the electronic and/or rotational-vibrational states of the dimers immediately after their formation.

Hlavní řešitel: Michal Krumnikl
Projekt: Fiji Bioimage Informatics on HPC – „Path to Exascale“
Alokace: 278 000 jádrohodin
Abstrakt: “Bioimage Informatics on HPC” allows IT4Innovations to be involved in research on a completely new topical area of big biological image data processing on HPC. This specific research is focused on parallelization of key steps in lightsheet microscopy data processing as well as analysis of big data generated from other microscopic modalities. Particularly, multi-dimensional microscopy acquisitions present one of the main primary data sources in modern biological sciences and deployment of HPC in these areas is a necessary condition for making biologically meaningful conclusions. This project is a continuation of the previous OPEN-15-12 call, solving particular VP3 subtasks of the Path to Exascale project. In the previous period, SPIM Workflow Manager was developed and published under Apache License. The project aims at further development and dissemination of HPC-aware plugins for the Fiji community. In this call we would like to specifically focus on utilization of a newly developed library for seamless parallel execution of SciJava plugins – SciJava Parallel and an automatic image segmentation and region labeling plugin – Labkit. As a new research topic we are starting to parallelize a simulator of artificial time-lapse images from developmental biology – EmbryoGen.

Hlavní řešitel: Petr Nachtigall
Projekt: Stability of Zeolite Catalysts in the Presence of Water
Alokace: 6 132 000 jádrohodin v první periodě
Abstrakt: Zeolites are the most important industrial catalysts produced on the Mtonne scale worldwide. Zeolites are microporous aluminosilicates stable in water at standard conditions and stable even at high temperatures during the crude oil processing and petrochemistry. These processes are inherently connected with generation of coke (carbon blocking the transport to the catalytically active sites) that must be periodically burned out using oxidizing atmosphere and high temperature. A substantial amount of water formed during the catalyst regeneration is responsible for partial zeolites hydrolyses and catalyst deactivation. It is therefore absolutely critical to understand the zeolite (in)stability in water and on steam. It is the aim of proposed project to investigate the factors influencing the zeolite (in)stability in aqueous environment by means of ab initio molecular dynamics performed at conditions closely mimicking relevant experimental situations. Molecular simulations planned within the project will bring new insight and understanding of the factors critical for catalyst stability in operando conditions at the atomistic level. They will serve experimental and industrial partners to design more durable catalysts.

Hlavní řešitel: Jan Zálešák
Projekt: EM studies of protein complexes and nucleoprotein filaments involved in homologous recombination
Alokace: 417 000 jádrohodin
Abstrakt: DNA is a carrier of the heritable information and it is under constant thread by multiple internal and external agents. These agents commonly cause numerous lesions and damages to our DNA. In our research, we investigate the processes which are responsible for reparation of such damages. These repair processes are carried out by complex protein machineries and they help maintain the stability of DNA within living organisms. DNA stability is a must for functional manifestation of the heritable information. In this project, we investigate the activation mechanism, which triggers the action of the key proteins of homologous recombination. These proteins, Rad51 and DMC1, have the ability to search and find highly similar patches within the cellular DNA and then use these patches as templates during DNA repair. We use electron microscopy to reveal three-dimensional details of these proteins in their activated or inactivated states. Our efforts are directed towards using structural information to facilitate the development of new therapeutic approaches.

Hlavní řešitel: Vojtech Mrazek
Projekt: Evolutionary Techniques for Hardware-Aware Hyperparameter Tuning and Architectural Search of Deep Neural Network Accelerators
Alokace: 167 000 jádrohodin
Abstrakt: Due to their state-of-the-art performance in many applications, deep neural networks (DNNs) have become an essential part of resource-constrained embedded devices. These devices are equipped with specialized accelerators to perform the underlying operations in a DNN inference efficiently. The accelerators have several architectural parameters that, when selected carefully, can lead to optimal performance and energy-/power-efficiency. Similarly, at software-level a number of different types of DNN architectures are available and the right selection of the type of architecture along with right set of hyperparameters can lead to optimal accuracy and performance . In this work, we will focus on an emerging type of deep neural networks, i.e., CapsuleNets. These networks are considered significantly better than traditional CNNs at modeling hierarchical relationships and, thereby, offer state-of-the-art accuracy, specifically in classification and recognition applications. We envision to use the computational resources for testing a methodology which can conjointly explore the hyperparameters of the CapsuleNets and architectural parameters of the CapsuleNet hardware accelerator to build a system which can offer both high accuracy and high resource efficiency.

Hlavní řešitel: Martin Zeleny
Projekt: Ab initio study of oxygen octahedra rotations in La2/3Sr1/3MnO3
Alokace: 2 022 000 jádrohodin
Abstrakt: High scientific interest is focused on the hole-doped manganite La2/3Sr1/3MnO3 (LSMO) with perovskite-type structure which possesses the high Curie temperature (TC ∼ 370 K) and almost 100 % spin polarization. Such a combination of physical properties makes LSMO interesting candidate for spintronics applications. The strength of ferromagnetism and metallic conductivity of LSMO are results of double exchange interaction between manganese Mn3+ and Mn4+ ions. Probability of such electron transfer is driven by Mn-O-Mn bond geometry. Therefore, the main factors influencing the magnetic properties are rotations and distortions of MnO6 octahedra, which are induced in thin films by an epitaxial strain coming from lattice mismatched substrates. The goal of the proposed theoretical investigation is to reveal the impact of epitaxial strain on structural changes in MnO6 octahedra and subsequently on changes in electronic structure of LSMO described by density of states. Magneto-optical (MO) Kerr spectroscopy reveals important details about the magnetic and electronic structure of materials. We are going to calculate also MO spectra, which allow us to directly compare our theoretical results with experimental measurements. Good theoretical understanding of relationship between structural changes induced by epitaxial strain and MO properties will pave the way for controlling the electronic structure of ultrathin perovskite films.

Hlavní řešitel: Marta Cudova
Projekt: Adaptive execution planning of biomedical workflows
Alokace: 21 000 jádrohodin
Abstrakt: Realistic ultrasound treatment planning and screening require very powerful computers on the daily basis. Their execution is described by a workflow consisting of a number of different cooperating tasks. The manual execution of individual tasks is even tedious and time consuming for expert users, but more importantly, preventing most inexperienced clinicians from the use of HPC recourses. The k-Dispatch tool offers a ‘run and forget’ approach where the users are completely screened out from the complexity of HPC systems. k-Dispatch provides task scheduling, execution, monitoring, and fault tolerance. Since the execution configuration strongly affects the final tasks mapping on the computational resources, the execution planning is of the highest priority. Unlike other tools, k-Dispatch considers a variable amount of computational resources per individual tasks. Since the scaling of the individual codes is never going to be perfect, k-Dispatch may find such a good solution even an experienced user would miss. The proposed adaptive execution planning is based on collected performance data and the current cluster utilization monitoring. The goal of this project is to use k-Dispatch to execute various ultrasonic, thermal and tissue modeling codes with different execution parameters on the cluster in order to obtain performance data. Since the collected datasets will be incomplete, interpolation and machine learning methods will be used to adapt the job scripts to previously unseen data. In the last stage, the planning algorithms will be evaluated in dedicated queues as well as in common production queues.

Hlavní řešitel: Andrzej Kadzielawa
Projekt: Mapping Density Functional Theorem onto Heisenberg model
Alokace: 2 799 000 jádrohodin
Abstrakt: Design of new materials for use in scientific and technological endeavors became an incentive for the development of novel computational methods based of First Principles of Quantum Mechanics. One of the most variously applicable approach is the Density-Functional Theorem (DFT), used via the both the commercially available and open-source code. In this project we propose a compact and systematic method of obtaining the magnetic properties (e.g. Néel and Curie temperatures) by using our original code (github.com/Mellechowicz/JorG) for mapping the DFT results to the Heisenberg model, allowing the calculation of the (A)FM-PM transition temperature. To obtain satisfying consistency of the code, we propose to first verify our results with both experimental measurements and other approaches (e.g. using the Green Functions), and finally to apply the approach to a set of newly proposed structures (e.g. iron-based permanent magnets). This in turn, will allow us to provide the community with an useful and fast (in comparison with existent codes) method of assessing the validity of time-consuming and (conceivably) expensive experimental trials.

Hlavní řešitel: Michal Merta
Projekt: Development of parallel BEM-based solvers II
Alokace: 333 000 jádrohodin
Abstrakt: One can choose from several numerical methods for modelling natural phenomena occurring in the real world, let us mention, e.g., the finite element method or the finite volume method. The main features of the boundary element method (BEM) make it well suited for problems stated on unbounded domains (such as sound or electromagnetic wave scattering) or shape optimization problems. Within the previous projects, we have focused on optimization of BEM solvers for Haswell and Skylake microarchitecture and its acceleration using Intel Xeon Phi coprocessors which are installed at infrastructure of IT4Innovations National Supercomputing Center. The current project aims at their further optimization and at development of parallel solvers for time-dependent heat equation. The global space-time approach will lead to the possibility of parallelization both in space and time and will improve the scalability on current and future supercomputers. This is a continuation of the project Development of parallel BEM-based solvers.

Hlavní řešitel: Vladislav Pokorny
Projekt: Correlation effects in superconducting quantum dot junctions
Alokace: 120 000 jádrohodin
Abstrakt: Conventional Josephson junctions are miniature devices in which two superconducting electrodes are separated by a weak link, usually an insulator barrier. These junctions had become a standard building blocks of various devices including rapid single flux quantum (RSFQ) electronics and qubits in quantum computing. If the insulating layer is replaced by a quantum dot, e.g. a carbon nanotube or semiconducting nanowire, we obtain a hybrid device in which various quantum mechanical phenomena as superconductivity, electron correlations and quantum tunneling can be separately tuned. Understanding the complex interplay of these phenomena is necessary step in developing a new generation of superconducting devices as quantum supercurrent transistors or monochromatic single-electron sources. Supercomputers are now a necessary tool for simulating the prospective superconducting devices, understanding the available experimental results and predicting their behavior.

Hlavní řešitel: Hugo Semrad
Projekt: Benchmarking MD/DFT calculations of 31P chemical shifts in intrinsically disordered proteins.
Alokace: 1 820 000 jádrohodin
Abstrakt: In the project, we perform benchmarking calculations of 31P nuclear magnetic resonance (NMR) chemical shifts (CSs) in intrinsically disordered proteins (IDPs). The proteins are important factor in the regulation of molecular mechanisms that cause neurodegenerative diseases. For humans, the understanding the structure of proteins and their interactions with other biomolecules is essential. Due to the high flexibility of IDPs, their structural characterization by experimental techniques such as NMR spectroscopy is made more difficult. The problem can be substantially alleviated by the use of computational methods. Using molecular dynamics simulations from which the coordinates of molecular clusters are obtained for further quantum mechanical calculations, the IDPs flexibility can be reflected. The project is aimed at DFT calculations of 31P chemical shifts in IDPs to provide a benchmark against which lower-level QM calculations will be compared to. The benchmarking calculations will employ (i) large basis sets and (ii) extensive statistical sampling of molecular structures. The results obtained will be used to evaluate the extent of systematic errors stemming from the basis set and statistical sample size. The benchmarking calculations will employ the human tyrosine hydroxylase 1 (hTH1) that serves as an example of an IDP. The project contributes to the development of approaches for computer-aided structural characterization of intrinsically disordered proteins.

Hlavní řešitel: Jakub Stosek
Projekt: Towards the prediction of chemical shifts in phosphorylated intrinsically disordered proteins: Design of chemical shift databases for phosphorylated tripeptide models.
Alokace: 1 518 000 jádrohodin
Abstrakt: The principle objective of the present project is to develop a database of chemical shift (CS) patterns for tripeptide models of proteins phosphorylated at serine residues. The database presents an essential building block in the development of a software (SW) tool for the prediction of NMR CSs in phosphorylated intrinsically disordered proteins (IDPs). The ability to predict CSs is a prerequisite for the structure characterization of IDPs by a combination of computational methods and NMR spectroscopy. While fast semiempirical CS prediction tools exist, they are neither tailored to the needs of IDPs, nor are they suited to handle non-standard residues, such as phosphorylated amino acids. The proposed project thus aims to address this lack in the current state of the art and design SW that predicts CSs for phosphorylated residues based on CSs computed by quantum mechanical methods. Phosphorylated IDPs can influence regulation of neurodegenerative processes caused by Alzheimer and Parkinson’s disease however the laboratory testing is extremely costly and time-demanding. Thus, this project will help to accelerate the research on neurodegenerative diseases and contribute to reducing the related research costs and thus it can improve the available medical care.

Hlavní řešitel: Jan Benacek
Projekt: Particle-in-cell simulations of the kinetic processes in the transition region in the solar atmosphere
Alokace: 530 000 jádrohodin
Abstrakt: The Sun is our closes star and represents an essential condition for life on Earth. The influence is mainly done by radiation, especially in the form of visible, infrared and ultraviolet light. The energy, which created in solar core by the process called nuclear fusion, propagates up through the solar interior and is radiated by the atmosphere. The solar atmosphere is composed of three main layers: photosphere, chromosphere, and corona. The layers influence not only the formation of the light spectrum but also the neighborhood of the Sun and whole solar system. In the research of solar atmosphere, there are two main processes which are not very well understood: heating of the solar corona, and the problem of the transition region. In our research, we study the plasma processes in the transition region. In this area, which has thickness only about 500 km, is a steep jump in the plasma temperature from approximately 10^4 K in chromosphere to 10^6 K in the corona. From the classical physical view, these temperatures should very quickly equilibrate but that is not measured. It is supposed that the plasma kinetic processes must be taken into account. We want to study this process using Particle-in-cell (PIC) simulations, where we want to create a similar temperature step and analyze the associated phenomena.

Hlavní řešitel: Zbynek Sokol
Projekt: Electrification of thunderstorms – simulations of selected events
Alokace: 1 041 000 jádrohodin
Abstrakt: Thunderstorms rank among severe weather phenomena since they are usually responsible for heavy rainfall, including hail, and strong lightning activity. Although they can lead to huge economic losses related to damages as well as many casualties, the thunderstorms and especially processes within thunderstorms have still not been fully described and understood making their prediction uncertain. One of the processes in thunderstorms that remains unclear is the process of electrification. Electrification is necessary to produce lightning. As soon as the thundercloud gets sufficiently electrified and charge well separated into positively and negatively charged layers, the lightning can be triggered. Several numerical models enable to simulate electrification using diverse concepts. This project aims at simulating the electrification of several thunderstorms that occurred in the Czech Republic using a Cloud Electrification Model that we implemented in COSMO Numerical Weather Prediction model. Similar to other alike models, our model requires high computation time. Our results will be validated with observed lightning during the thunderstorms and the simulations of the thunderstorm electrification will help us better understand the process of electrification, which can contribute to an improvement of severe weather prediction in future.

Hlavní řešitel: Rajko Cosic
Projekt: Utilization of artificial neural networks in path integral Monte Carlo simulations II
Alokace: 375 000 jádrohodin
Abstrakt: The aim of present project is to continue the work on potential energy surface (PES) evaluation via the artificial neural networks (ANNs) started within preceding project (OPEN-16-21). The main task is to provide reliable training sets for the ANNs’ learning process. Since our other research projects involve simulations of the photoabsorption spectra of charged helium clusters, the ANNs representation of charged systems is of high importance. Since clusters we are interested in are of medium sizes (N=21 and N=30 for which the experimental data exist), the simulations using standard diatomics in molecules (DIM) based models are almost impossible. This project is part of broader scientific intention and focuses only on generating the training sets, test sets and test simulations using the path integral Monte Carlo (PIMC) method. Another complementary projects with intention to deal with the ANNs construction, implementation of permutational invariants and incorporating the charge distribution into the ANN have been (and, in the future calls, will also be) submitted.

Hlavní řešitel: Marek Lampart
Projekt: Non-autonomous model of collisions of a constrained body on a moving belt
Alokace: 42 000 jádrohodin
Abstrakt: The main aim of the project is to analyse the dynamic properties of a mechanical system. The investigated system consists of a cylinder hanging on a flexible rope and of a moving belt. Such a system with impacts and dry friction is an image of many industrial applications, like stones falling on a conveyor moving belt. The mathematical model of the system has three degrees of freedom from which two correspond to the position of the cylinder centre and the last one to its angular rotation. The studied system is excited by a slider moving in the vertical direction and by impacts between the cylinder and the belt. As a main result, it will be observed that the cylinder exhibits movement with both regular (periodic) and irregular (chaotic) patterns depending on the excitation amplitude and frequency. The goal of the research is to qualify and quantify the movement character. For this purpose, the 0-1 test for chaos together with approximate entropy will be applied to find the regions of parameters for which chaos or regularity will be observed.

Hlavní řešitel: Jiri Jaros
Projekt: Photoacoustic tomography of the breast
Alokace: 320 000 jádrohodin
Abstrakt: 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. This simulation study is focused on the optimization and validation of the PAT software being developed as part of the H2020 project Pammoth on a set of breast phantoms. The resolution, accuracy, sharpness, motion and noise artifact, and the depth of penetration will be investigated and optimized. This study moves us towards the deployment in a real PAT system for breast mammography.

Hlavní řešitel: Vitezslav Hanzal
Projekt: AutopilotData
Alokace: 250 000 jádrohodin
Abstrakt: Autonomy and human machine interaction are one of most demanded research areas. This project is intended to allow development of demonstrator of the advanced autonomous flight control system (AFCS) of jet combat/trainer aircraft. Advanced AFCS of such high maneuverability, transonic airplane must handle not only flight on predefined trajectory, takeoff and landing, but also auto-recovery from unusual positions, adaptability on varying flight characteristics due to changing plane configurations, pods, varying Mach number. Essential part of the task is to gather required flight characteristic of the plane. Extensive campaign of computer simulation of the flow dynamics (CFD) is needed.

Hlavní řešitel: Pavel Balaz
Projekt: Influence of Dzyaloshinskii-Moriya interaction on static and dynamic properties of magnetic structures
Alokace: 764 000 jádrohodin
Abstrakt: Dzyaloshinskii-Moriya interaction (DMI) is an antisymmetric exchange interaction between two magnetic moments induced by spin-orbit coupling, a relativistic interaction of electron’s spin with its orbital momentum. In magnetic materials, DMI violates collinear magnetic configurations since it favors spin canting leading to complex magnetic textures. One, of the most important features of DMI is, it can stabilize magnetic skyrmions – topological defects in magnetic materials with particle-like behavior in a sense, they can be manipulated by external magnetic fields electric current as a particle carrying information in a form of magnetic momentum. Therefore, skyrmions might create a basis for future logic devices used in spintronics. On one hand side, DMI can be observed in magnetic multilayers, where a magnetic layer is attached to a heavy metal with strong spin-orbital interaction. On the other hand, DMI is present in some multiferroic materials causing their electric polarization induced by magnetic ordering. In this project we shall focus on both. First, we shall revisit localized 90-degree magnetic domain walls, which can be prepared in multiferroic heterostructures. By means of micromagnetic simulations we shall describe influence of DMI on their static and dynamic properties. Second, we shall focus on a novel ferroelectric material GaV4S8 featuring DMI acting between V4 clusters and studying possibility of stabilizing skyrmion lattices under various physical conditions.

Hlavní řešitel: Stanislav Palacek
Projekt: Computations of [CH40]+ potential energy surfaces
Alokace: 1 555 000 jádrohodin
Abstrakt: This project is proposed for computations of potential energy surfaces as well as their gradients, transition dipole moments and nonadiabatic couplings of [CH4O]+ system. These physical quantities will be interpolated by neural networks using my own nuron4dyn libraries. Learned neural networks will be used on semiclassical dynamics simulations using program MULTIDYN [1] in version 3.0. Quantum chemistry calculations will be done by reliable quantum chemistry programs provided by third parties.

Hlavní řešitel: Christopher Heard
Projekt: 2D Layered Oxyhydroxides for Catalytic Water Splitting
Alokace: 2 170 000 jádrohodin
Abstrakt: Low dimensional materials show remarkable electronic properties and are becoming employed in a wide variety of applications, ranging from sustainable energy production and storage, to spintronics. The wide variety of possible structures, compositions, surface functionalization and electronic band structures make these materials extremely amenable to optimization. However, a rigorous understanding of the local structure of these materials is lacking, both experimentally and computationally, which currently limits the possibility to design 2D materials with optimal stability and catalytic properties. In this project, we will investigate two important groups of related 2D layered oxyhydroxide materials: i) iron-based layered double hydroxides (LDH), and ii) oxygen/hydroxide terminated transition metal carbides (MXenes), under realistic conditions, taking into account the important effects of temperature and solvent (water). We will determine the true atomistic structures of these materials and determine how they vary as a function of several coupled variables: intra-layer metal identity, oxidation state and layer surface termination. The structure will be related to the catalytic activity by determining the reaction mechanisms of OER and HER reactions on LDHs and MXenes respectively, via a thorough investigation of possible reaction steps with ab initio biased dynamical simulations. We will locate the promising modifications of the materials for catalysis, and further the understanding of the structures needed for application of 2D materials in practice.

Hlavní řešitel: Jiří Kaleta
Projekt: Surface-bound molecular machines as building blocks in nanoscience
Alokace: 491 000 jádrohodin
Abstrakt: A very important goal of the rapidly advancing field of molecular machines is the production of highly organized nanoscale systems capable of propelling themselves in a controlled manner. Such systems could have many uses, such as cargo transport, drug delivery, nanosurgery or environmental remediation. In our group, we are developing machines with propulsion capabilities based on light propelled molecular motors attached to metallic surfaces of nanoparticles. The aim of this project is to find an optimal molecular shaft interconnecting the rotor and the metallic surface. Such shaft needs to fulfill three requirements: (i) it must have a high barrier for rotation, effectively functioning as a stator; (ii) it needs to pack efficiently in 2D films, to maximize the surface density of rotors; and (iii) it should have high affinity to the metallic surface, to ensure good coverage of the surface. The first two requirements can be satisfied by interlocking the shafts together, while the third requires careful choice of an anchoring group. The proposed computations would explore different shaft-anchor-surface possibilities to find optimal combinations, explain experimental data and guide further synthetic efforts.

Hlavní řešitel: Michal Bidlo
Projekt: Evolutionary Approach to Approximate Component-Based Circuit Design
Alokace: 292 000 jádrohodin v první periodě
Abstrakt: The aim of this project is to investigate advanced techniques based on the principles of natural computation for designing approximate computing systems. The recent research in this area has shown that in some cases various features of the system may be optimized if a certain error is tolerated on output. Such systems utilize approximative parts (sub-circuits) which enable this capability. For example, power consumption may be reduced in multiplier circuits if the result is allowed to exhibit a small error measured using a suitable metric. An extensive library of approximate components has been developing at BUT FIT during recent years which is going to be utilized within this project. The goal is to investigate various approaches how to design useful computing systems, composed of approximative components from the library, that can benefit from the approximate computing concept. In particular, bio-inspired techniques and evolutionary algorithms are supposed to be used in this research.

Hlavní řešitel: Stepan Sklenak
Projekt: Periodic DFT studies of zeolite-based catalysts
Alokace: 947 000 jádrohodin
Abstrakt: 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. We propose periodic DFT investigations of various properties of metal cation (Li+, Na+, Cs+, Fe2+, Co2+, and Cu2+) exchanged zeolites (e.g. ferrierite, ZSM-5, beta, mordenite, and TNU-9).

Hlavní řešitel: Tomas Oberhuber
Projekt: Development of multilevel BDDC
Alokace: 208 000 jádrohodin
Abstrakt: The domain decomposition method can be employed to the solution of linear algebraic systems resulting from the underlying partial differential equations. We focus on the non-overlapping subdomains approach which leads to the solution of the Schur complement system. There are various approaches to the pre-conditioning of the iterative solution of the Schur complement system. In this work, we employ the balancing domain decomposition based on constraints (BDDC). The BDDC consists of two problems. The local problem that requires only the data available at the subdomain and can be done in parallel and the global (coarse) problem. The coarse problem requires the data from all the subdomains and, therefore, its parallelization is more complicated. With the increasing number of subdomains, the coarse problem becomes the bottleneck for scalability of the two-level BDDC. This can be overcome by the introduction of next level or levels to the BDDC. The multilevel BDDC takes advantage of the structure of the coarse problem. Its structure is the same as the original problem and therefore the domain decomposition can be employed again. On this level, the unknowns are coarse degrees of freedom of the previous level. Therefore, instead of the exact solution of the coarse problem, the solution is only approximated by a single BDDC step. If the new coarse problem is still too large, these steps can be repeated recursively.

Hlavní řešitel: Marketa Tkadleckova
Projekt: Numerical modelling of metallurgical processes
Alokace: 100 000 jádrohodin
Abstrakt: The technology for clean metal production in modern metallurgical plant is analyzed using the ANSYS Fluent software. Three main areas are solved. As first, the refining of molten aluminum alloy in ladle by an inert gas injected through a rotary impeller. The second area of the research is focused on the numerical modelling of molten steel flow during the inclusions removal in the tundish. The third area represents the steel flow modelling in ladle during the continuous casting and the risk of slag emulsification into molten steel. In all three cases, the main aim is to optimize the boundary conditions of technology to produce the clean metal without impurities. The impurities cause the lower quality of the final metal products. But in the field of metal production, the priority of every manufacturing company is to increase the productivity, quality and production safety of the produced metal with simultaneous minimization of the environmental burden. Modelling processes is a method that aims at capturing the behavior of a real system using a physical or mathematical model without having to interfere with a real technological process and thereby limit the standard production process. Based on the results obtained on the model, it is possible to predict the behavior of the real system in the course of various process changes.

Hlavní řešitel: Ivo Peterek
Projekt: Parallel-in-Time Methods
Alokace: 42 000 jádrohodin
Abstrakt: A parallel solution to boundary value problems for partial differential equations (PDEs) by means of domain decomposition methods (DDM) is nowadays well-established. Extensions towards time-dependent problems are much less understood and any progress in this direction is appreciated from both scientific and practical points of view. In this project, we propose to combine the Parareal method with spatial domain decomposition. We rely on an MPI-C++ inhouse code of one of the investigators.

Hlavní řešitel: Dominique Geffroy
Projekt: Metal to insulator transition in relativistic Mott insulators
Alokace: 2 177 000 jádrohodin v první periodě
Abstrakt: Over the last decade, relativistic Mott insulators have become a hot topic in the field of strongly correlated electrons. In these compounds, spins and orbitals become entangled by a large spin-orbit coupling, thereby producing new effective degrees of freedom. Their complex structure may get reflected in the interactions and give rise to novel forms of quantum magnetism, a prominent example being Kitaev materials. This project will study the physics of systems with d4 transition-metal ions such as Ru4+ where the multiplet structure contains a nonmagnetic singlet groundstate and a low-lying triplet. Magnetic moments are formed by a superposition of these states and hence are soft. We will focus on the peculiar excitation spectra involving spin-length fluctuations (Higgs mode). Strong emphasis will be put on the interpretation of Iinelastic neutron scattering (INS) and Resonant Inelastic X-ray Scattering (RIXS) experiments. DFT+DMFT calculations will be used, so as to describe real materials in as precise a manner as possible with today’s technology.

Hlavní řešitel: Lukas Halagacka
Projekt: Optical modelling of textured optical surfaces with CST Microwave Studio
Alokace: 167 000 jádrohodin
Abstrakt: Micro- and nano-scaled texturation is the efficient approach how to increase active area of the devices without need of increasement of volume of materials. Proper modification of surfaces by detailed texturation or precisely defined multilayer stacking represents mostly used approaches to control optical response of the surface. Nevertheless, with the broadening of the device operational spectral range, the complexity of the surface modulation increases. Therefore, a modification of surfaces with flat structures or stack of layers is not sufficient anymore. This opens possibilities to explore new types of structures with wide range of possible applications such as optical sensors, solar cells, electrochemical devices for hydrogen production, or new generations of LEDs.

Hlavní řešitel: Ales Prachar
Projekt: KruegerDyn
Alokace: 250 000 jádrohodin v první periodě
Abstrakt: The project deals with modelling of the unsteady aerodynamics during extension of the Krueger flap on the leading edge of the wing with advanced CFD methods. This includes the possibility of moving geometries by newly implemented overlapping mesh approach. The better and more accurate aerodynamic characterization of loads during the Krueger flap deployment contributes to more adequate structural sizing, thus reducing complexity and weight of the Krueger flap device. Furthermore, Kruger flap qualifies as enabler for the laminar wing technology, which is attractive due to reduction of aerodynamic drag and consequently of fuel burn. Both contribute to the reduction of environmental footprint of the aviation industry.

Hlavní řešitel: Ivan Kolos
Projekt: Numerical modeling of load of structures in quasi-static effect of wind
Alokace: 69 000 jádrohodin
Abstrakt: The project is focused on numerical modeling of flow around objects in the atmospheric boundary layer. This issue is complicated mainly due to the atmospheric turbulence, which requires the use of advanced numerical models of the flow coupled with detailed computational mesh of the domain. This research will contribute to bigger efficiency in design of building structures.

Hlavní řešitel: Jaroslav Resler
Projekt: Fine-scale air quality modelling of scenarios in Prague-Dejvice using PALM-4U
Alokace: 2 888 000 jádrohodin
Abstrakt: PALM-4U is a newly developed complex urban climate model. Our team on ICS CAS significantly contributed to model development and implemented many of urban-specific processes (see Resler et al. 2017, Maronga et al. 2019). The model has been validated against observations collected during measurement campaigns in Prague-Holesovice and Prague Dejvice. We also developed unique detailed geodatabases for fine-scale modelling for Prague-Dejvice, part of Prague-Holesovice, and historical centre of the Prague (in progress). The pilot simulations of these areas proved to be of great interest for local municipalities, as well as for selected departments of the Ministry of the Environment. Regular cooperation between ICS and Prague municipality was established to perform modelling assessment of their development scenarios for comparison of different adaptation strategies. The study will focus on air quality in the street canyons and the first chosen region is the Prague-Dejvice area. The modelling setup will follow the new setup developed and tested during the validation works. It will include an area of extent 4 km x 4 km around the square Vítězné náměstí. We plan to calculate one base case and 8 scenarios for two typical meteorological situations. Scenarios will include significant landcover modifications and adaptations of the transportation. The results of this analysis will not only serve Prague municipality but we also plan to publish them in a relevant scientific journal.

Hlavní řešitel: Michael Owen
Projekt: Alzheimer’s Lipids III
Alokace: 1 273 000 jádrohodin v první periodě
Abstrakt: Neuronal membranes can enhance or prevent the formation of oligomers of the amyloid-β (Aβ) peptide, the neurotoxic species implicated in Alzheimer’s disease. Membrane gangliosides have an important role in brain development, regeneration, and in the progression of Alzheimer’s disease, however, the specific role of gangliosides in Ab oligomerisation remains unclear. We previously tested the validity of various force fields in their ability to accurately represent the structural behaviour of GM1 and the Ab peptide using all-atom molecular dynamics (MD) simulations. We have further investigated the properties of membranes comprised of GM1 and five other gangliosides, GM2, GM3, GD1a, GD1b, GT1b, at both high and low ganglioside concentration to mimic the ability of the gangliosides to form clusters. In this Project Call, we outline our study on the effect of each ganglioside, in both their clustered an unclustered configurations, to interact with a dimer of Ab42 that has been previously characterized as “on-pathway” with respect to its ability to form toxic oligomers. We will characterize the interactions between each membrane and the Ab dimer to evaluate the ability of each ganglioside type to destabilize the dimer structure, and thus protect the neuron from the larger, more toxic oligomers that can form from the on-pathway dimer. This innovative research project is supported by the South Moravian Project (SoMoPro) and Marie Skłodowska-Curie Actions.

Hlavní řešitel: Josef Michl
Projekt: Porphene mono- and bilayers as tunable 2D materials
Alokace: 1 197 000 jádrohodin
Abstrakt: Design of new periodic two-dimensional polymers is a topic of immense research interest, with possible applications in optoelectronics, molecular magnetism, catalysis, energy storage, sensing and others. Graphene and metal organic frameworks have shown much promise, but one of the biggest issues remains the possibility of tunability, i.e. the modification of properties in a predictable and controllable manner. This project is focused on the investigation of porphene: a synthetically available, but relatively unexplored nitrogen containing 2D polymer that is built of porphyrine units and is closely related to graphene. It is expected that its nitrogen-based acceptor sites can bind metals or small molecules without destroying the underlying conjugated electronic structure. This means that the properties of porphene sheets and multilayers could be finely tuned by binding appropriate ions or small molecules, making porphene based materials very attractive.

Hlavní řešitel: Lukas Grajciar
Projekt: Neural network potentials for in silico design of zeolites
Alokace: 2 013 000 jádrohodin
Abstrakt: Computer aided (in silico) design of novel catalysts such as zeolites is of fundamental importance for development of new technologies for sustainable (green) chemistry. It demands both accuracy and speed which is difficult to satisfy simultaneously due to prohibitive costs of first principles-only treatment or due to unsuitability/unreliability of available empirical force fields. This project provides a computational tool providing the best compromise of accuracy and computational costs by combining artificial intelligence and quantum mechanical simulations. State-of-the-art artificial neural networks are used to model the interatomic interactions of complex zeolite materials, facilitating large scale simulations with unprecedented accuracy. The simulation results provide deeper insights into the structure and stability of existing and hypothetical zeolites that have not yet been synthesized. In addition, the prediction of the atomic structure of germanium containing zeolite frameworks allows to guide experimental studies for time and cost efficient fabrication of new, unfeasible zeolites with improved catalytic properties.

Hlavní řešitel: Pablo Nieves
Projekt: Towards discovering novel magnets with high anisotropy under extreme conditions
Alokace: 3 819 000 jádrohodin
Abstrakt: At present, there is a demand for permanent magnets (PMs) which can operate at temperatures up to 450°C. This has arisen from a drive to replace hydraulic systems in aeronautics with electromechanical devices, to develop systems such as airborne motor/generators integrated with turbine engines and hybrid magnetic bearings, as well as the need for high performance high temperature magnets for electric vehicles. Typically, the properties of PMs rapidly degrade as temperature is increased. Nowadays, there isn’t any high-performance magnet with maximum energy product greater than 30 MGOe above 300°C, although important potential applications exist in the fields of aeronautics, the space industry, electronics, and the automotive industry. Additionally, current high-performance PMs are totally dependent on not abundant and expensive Rare-Earth (RE) materials like Nd and Sm, which have now become critical raw materials. Within current proposal we intend to calculate the temperature behavior of magneto-crystalline anisotropy energy (a key property of high-performance PMs) for a set of interesting uniaxial theoretical Fe-based compounds (Fe-X, X=Ta, Hf, B, P, C, N), previously screened and studied at zero-temperature, in order to identify new RE-free/lean PMs suitable for high-temperature applications.

Hlavní řešitel: Sergiu Arapan
Projekt: Helium in Liquid Metals (HELM)
Alokace: 1 944 000 jádrohodin
Abstrakt: Fusion power is probably the most sought-after technological goal in the pursuit of clean energy. Nuclear fusion, the nuclear reaction that powers the Sun and the stars, is a potential source of safe, non-carbon emitting and virtually limitless energy. Replicating that process on Earth at sufficient scale could unleash a tremendous amount of energy. The problem is (i) creating the extreme conditions necessary for such reactions to occur, (ii) harnessing the resulting energy in a useful way, and (iii) controlling the reactions once they have been induced. The future power plants are supposed to operate on fusion of two isotopes of hydrogen, deuterium and tritium.Deuterium can be extracted from seawater in virtually boundless quantities. The supply of available tritium, however, is limited and currently estimated at only twenty kilos. Fortunately, a second source of tritium exists. Tritium can be produced within the reactor when neutrons escaping the plasma interact with lithium contained in the blanket. There is also helium He produced during the reaction, which can alter the liquid properties. We have limited experimental knowledge of the behavior of He and T inside the liquid alloy. Molecular Dynamics is an appropriate numerical tool to study such kind of processes. However, to our knowledge, an accurate interatomic potential able to describe the whole system (PbLi+He+T) is not available. The main goal of the proposed project is to develop a machine learning interatomic potential for the PbLi+He+T system for an accurate study of the system at different temperature, pressure, and He and T concentrations conditions.

Hlavní řešitel: Jan Heyda
Projekt: QM/MM molecular dynamics simulations of electron transfer processes in blue copper azurine mutants
Alokace: 463 000 jádrohodin
Abstrakt: Electron transfer (ET) reactions [Beratan] are essential in life processes and play important role in biology e.g. in respiration and photosynthesis. Long-range electron transfer through proteins is essential to deliver or remove electrons over distances varying from few Ångströms to tens of Ångströms. The number of tryptophan units and their position in the protein structure strongly influence the rate of ET. The carbonyl stretching frequencies of Re carbonyl diimine complex incorporated into protein chains as a sensitizer monitors kinetics of ET transfer. The mechanisms of photoinduced ET will be studied by time-resolved spectral techniques combined with molecular dynamics (MD) theoretical calculations. The aim of this project is to complete our previous computational study (project OPEN-14-49) of the electron transfer properties of the blue copper protein Pseudomonas aeruginosa azurin with attached ReI photosensitizer. Calculations performed by combined quantum chemical – molecular mechanical methods (QM/MM) indicated the necessity to treat several excited states simultaneously by time-dependent DFT (TD-DFT) techniques and account for explicit solvent surroundings. Number of TD-DFT QM/MM molecular dynamic simulations will start from DFT QM/MM-MD simulations performed along long classical MM/MD runs in order to obtain statistically sound results. Geometrical factors and electron transfer coupling obtained along the MD trajectories will be used for the interpretation of the experimentally proposed mechanism of electron transfer kinetics.

Hlavní řešitel: David Schwarz
Projekt: Fluid structure interaction analysis of arterial tree
Alokace: 29 000 jádrohodin
Abstrakt: The vascular tree is used for the distribution of oxygenated blood that comes out of the heart during heart cycle. Due to the elasticity of the walls, the arteries can reduce the pressure to such an extent that a constant blood flow occurs at the periphery of the vascular tree. However, over time, arterial walls become stiff, resulting in a change in the shape and amplitude of the pulse wave, which is directly related to blood pressure which has been strongly connected to various vascular diseases. Hypertension has been shown to cause various types of diseases such as coronary heart disease, heart failure, peripheral arterial diasease, stroke, or renal disease. Therefore, this project aims to create a Fluid-structure interaction (FSI) analysis that shows the parameters significantly affecting the course of the pressure in the vascular tree. The vascular tree is obtained from a CT scan and is printed from silicone to validate and tune the FSI analysis. Thanks to the combination of FSI analysis and experimental measurement, it is possible to obtain a more comprehensive overview of the pressure course when changing the searched parameters.

Hlavní řešitel: Radek Vitasek
Projekt: Computational modelling of abdominal aortic aneurysms
Alokace: 17 000 jádrohodin
Abstrakt: Abdominal Aortic aneurysm (AAA) is a permanent diameter enlargement of abdominal aorta. Its rupture is a life-threatening event. On the other hand the non-ruptured AAA does not bring any complication to patient in most cases, therefore operations should be done only in cases with AAA close to rupture. Mostly used criterion for operation is maximum diameter. However it is not very accurate since not all large AAAs rupture while some small does. Therefore new criterions have been searched by scientific teams. One of the criterions is based on a wall stress analysis on the AAA. This project aims at analyzing of different mean arterial pressures affecting AAA.

Hlavní řešitel: Libor Veis
Projekt: Benchmark tests of the newly developed massively parallel quantum chemistry DMRG code
Alokace: 292 000 jádrohodin
Abstrakt: Quantum chemical computations have, over the years, become essential tools of the basic as well as applied research in chemistry and material science. However, due to the difficulty of the electronic structure problem, there is a need for novel approaches, which will allow faster computation of desired properties. One of the new trends in computational chemistry is a massive parallelization and modern implementations are indeed designed to be highly scalable. In previous projects, we have developed the MOLMPS program, a novel and highly scalable C++ implementation of the quantum chemical density matrix renormalization group method (QC-DMRG), which is one of the most successful computational methods for treatment of strongly correlated (multireference) molecules and materials. In this project we would like to apply this method to large scale benchmark tests and verify current limits of the state-of-the-art QC-DMRG calculations. The most challenging benchmark tests, which we would like to perform, are electronic structure computations of FeMoCo, the nitrogenase cofactor, the true holy grail of bio-inorganic chemistry.

Hlavní řešitel: Ales Vitek
Projekt: Methane adsorption on graphene II
Alokace: 333 000 jádrohodin
Abstrakt: Our group is developing own computer code for simulation of thermodynamics on graphene surface under different thermodynamics condition. We test various description of interaction models and different approaches to simulate particular thermodynamics condition. Methane capture has gained a lot of interest because of two main reasons. First of all there is an increasing urgency to reduce greenhouse gas emissions, as was highlighted by the recently signed Paris agreement. While less methane is emitted than carbon dioxide, its higher energy-uptake makes it a big contributor to the greenhouse effect, and thus materials capable of filtering methane from exhaust gas mixtures are highly welcomed. On the other hand, methane is often suggested as a transition fuel until alternative energy sources become feasible for large-scale use. In our computations, we will use classical isobaric-isothermal Monte Carlo simulation to predict possibilities of adsorption of methane on graphene surface under different thermodynamics condition.

Hlavní řešitel: Jozef Hritz
Projekt: Conformational changes of Tau protein
Alokace: 2 048 000 jádrohodin v první periodě
Abstrakt: Although the molecular mechanism of Alzheimer’s disease is poorly understood, there is a strong agreement that pathologically altered tau protein plays an important role in the process. The main motivation of the proposed project is to study the structural ensembles and conformational changes induced by truncation and phosphorylation of tau protein – the family member of microtubule associated proteins. The methodological challenge for planned study is the fact that tau protein belongs to the group of intrinsically disordered proteins (IDPs). It means that they do not have a single, well-defined structure, and are rather represented by an ensemble of structures distinctly different from a random coil. The disordered character of tau protein makes it difficult to study by current methods in structural biology and biophysics, which were developed and optimized for well folded proteins. We will study tau protein variants by computational simulations based on molecular dynamics. Due to the need of extensive conformational sampling of studied proteins we will be also testing performance of enhanced sampling methods. Multiple force-field parameters will be validated against experimental data obtained from nuclear magnetic resonance (NMR), small angular X-ray (SAXS) and circular dichroism (CD) spectroscopy data. Most of these experimental data will be measured in our research group or in the labs of our collaborators allowing for interactive approaches when needed. The direct interplay of computational and experimental techniques is the key point of this proposal.

Hlavní řešitel: Jan Zemen
Projekt: Modeling Magneto-Optical Spectra in Antiperovskite Nitrides Subject to Strain
Alokace: 1 365 000 jádrohodin
Abstrakt: The aim of the project is to calculate Magneto-optical spectra across a range of antiperovskite nitrides, compare them to available spectra measured in thin films, and identify magnetic phases with potential for high sensitivity to external stimuli (mainly lattice strain) at room temperature. The magnetic structure of Mn-based antiperovskite nitrides was first studied by neutron diffraction in bulk samples in 1970s. Antiferromagnetic (AFM) coupling between neighbouring magnetic moments of three Mn atoms in the unit cell results in a geometrically frustrated triangular AFM structure highly sensitive to applied strain which hosts a range of useful effects even in bulk form, e.g., the invar effect. In recent years, several groups have attempted to grow these materials in thin-film form aiming for applications in non-volatile memory devices, sensors, and actuators. We have shown by comparing computed and experimental data that the lattice mismatch between the epitaxial antiperovskite film and the substrate leads to canting of the non-collinear Mn moments and may induce a collinear ferrimagnetic (FIM) phase. Large Magneto-optical Kerr Effect (MOKE) has been measured in thin films which is explained by the large local magnetic moments and the symmetry of the structure. The sensitivity of MOKE to changes in magnetic and electronic structure makes it an ideal tool to identify the strain-induced magnetic phases and the chemical composition of alloys with desirable spintronic or solid-state cooling functionalities.

Hlavní řešitel: Luigi Cigarini
Projekt: Interplay between electronic, conformational and phononic structure in group V 2D materials phosphorene and arsenene
Alokace: 250 000 jádrohodin
Abstrakt: Phosphorene is an example of new generation material which could have exceptional applications in electronics and energy management, because it can switch from being electrical insulator to conductor when, e.g. reaching a certain temperature or exposed to some light, color or electromagnetic impulse. Research is still in progress to increase our knowledge on how phosphorene works at the atomic level. In this research project we aim to simulate how electrons and atomic nuclei behave in phosphorene, in order to be able, in future to tune its switching properties for real applications.

Hlavní řešitel: Dominik Legut
Projekt: Novel 2D magnetic semiconducting materials
Alokace: 7 360 000 jádrohodin
Abstrakt: For the spintronic applications like large data storages (high capacity HDD) the industry search for ferromagnetic insulators and at the same time a very thin materials. Recently the discovery of Bi2O2Se/Te phases could exist as 2D material and still be semiconducting. Here we investigate these novel layer-structured materials with high electron mobility, while its efficiency could be greatly improved by doping different elements to introduce a magnetic spin order employing larege quantum-mechanical calculations. We explore the electronic and magnetic properties of various ferromagnetic (e.g. Fe) or antiferromagnetic (e.g. Mn) transitional metals doped Bi2O2Se/Te within framework of quantum chemical calculations. We start with the magnetic order of the bulk phase in which the magnetic atoms form interlayer coupling would vary with the type and concentration of doped atoms and go towards the 2D materials. As a result of the competitions of magnetic interactions we will calculate magnetic anisotropy energy as a crucial quantity (our first main goal) to achieve a potential spintronic material. Next, in combinations with Monte Carlo simulations, solve the exchange interaction constants for the Heisenberg model and further evaluate the Curie temperatures (second main goal) of Bi2-nXnO2Se/Te to see if these types of materials allow to desgn new dilute magnetic semiconductors for spintronic applications at room and above temperatures.

Hlavní řešitel: Jan Heyda
Projekt: Collapse kinetics of the themoresponsive polymer chain in light of proximal water density fluctuations
Alokace: 467 000 jádrohodin
Abstrakt: Polymer poly(N-isopropylacrylamide) (PNIPAM) is the most widely studied temperature-responsive material, which undergo completely reversible volume phase transition in water, upon heating above 32°C. [Stuart] Polymer collapse is driven by worsening the solvent quality with temperature, which strengthens the hydrophobic interactions. In this project we aim to quantitatively study the kinetics of the collapse transition of a single PNIPAM chain, employing transition interface sampling techniques developed for investigation of rare events. In our previous kinetic study (OPEN-12-27) we found that solely polymer variables are insufficient for accurate description of the collapse pathway and solvent coordinate is required. Recently, water density fluctuations in the vicinity of polymer cavity were suggested as the key solvent property, which drives kinetic and thermodynamic properties of the polymer. Extensive sampling of water density fluctuations will be performed on the ensemble of transient state configurations of the polymer. We will compare the polymer collapse pathway and evaluate solvent kinetics at temperatures below, at, and above the volume phase transition temperature.

Hlavní řešitel: Martin Matys
Projekt: Ion acceleration driven by current high-repetition laser systems and effects of preplasma in this interaction
Alokace: 333 000 jádrohodin
Abstrakt: With the commissioning of laser facilities like ELI-Beamlines and ELI-ALPS, the experiments with high-repetition (kilohertz) lasers with energy around 100 mJ are now available. With this type of lasers, ions from solid/fluid targets can be accelerated to energies suitable for several applications and fundamental researches, like nuclear research, g.e., neutron generation via fusion reactions. In this project we will examine this interaction with the help of demanding 2D and 3D particle-in-cell simulations. We will also investigate the effects of preplasma, which is usually generated by the prepulse accompanying the main laser pulse, on the ion acceleration in this case.

Hlavní řešitel: Libuse Horackova
Projekt: Development of parallel implementation of feedforward neural networks and recurrent neural networks with the main focus on solving multidimensional Schrödinger equation
Alokace: 333 000 jádrohodin
Abstrakt: The Schrödinger equation is a motion equation of nonrelativistic quantum theory. It was formulated by Erwin Schrödinger in 1925. It describes the temporal and spatial development of the wave function of a particle moving in the field of forces. The main aim of this research is to investigate the use of classical and recurrent neural networks for solving multidimensional Schrödinger equation. Using this approach, it is possible to obtain a solution to the Schrödinger equation while avoiding numerically challenging calculations needed in classical computation to this equation. The outcomes of this project will help scientists investigate atomic states using less computational power and will also open the door to investigation more complex atomic systems that could not be solved using classical methods of numerical mathematics.