Ingenieurwissenschaften und zugeordnete Tätigkeiten
Filtern
Dokumenttyp
- Forschungsdatensatz (11)
- Vortrag (6)
- Posterpräsentation (3)
- Zeitschriftenartikel (2)
- Sonstiges (2)
- Video (1)
Referierte Publikation
- nein (25) (entfernen)
Schlagworte
- Automation (12)
- Bonding analysis (11)
- Computational Chemistry (11)
- Database (11)
- Materials Informatics (11)
- SAXS (4)
- Simulation (4)
- Digitalisierung (3)
- PDF (3)
- XRD (3)
Organisationseinheit der BAM
- VP Vizepräsident (25) (entfernen)
Eingeladener Vortrag
- nein (6)
In the pursuit of discovering materials with desirable properties, extending the available material libraries is crucial. High-throughput simulations have become an integral part in designing new materials in the past decades. However, there is no straightforward way of distinguishing synthesizable materials from all the proposed candidates. This project focuses on employing AI-driven methods to estimate synthesizability of materials.
Up to now, material scientists and engineers have relied on domain knowledge as well as empirical heuristics to guess the stability and synthesizability of molecules and crystals. The famous Pauling rules of crystal stability are an example of such heuristics. However, after the accelerating material discovery in all the years since Pauling, these rules now fail to account for the stability of most known crystals. A new predictive set of heuristics for crystal stability/synthesizability is unlikely to be uncovered by human perception, given the magnitude and dimensionality of crystallographic data. Hence, a data-driven approach should be proposed to find a predictive model or set of heuristics which differentiate synthesizable crystal structures from the rest. The main challenge of this research problem is the lack of a negative set for classification. Here, there are two classes of data: the positive class which contains synthesizable materials and the negative class which contains materials which are not synthesizable. While the data from the positive class is simply the data of crystals which have been experimentally synthesized, we do not have access to data points which are certainly unsynthesizable. Strictly speaking, if an attempt of synthesizing a crystal fails, it does not necessarily follow that the crystal is not synthesizable. Also, there is no database available which contains the intended crystal structures of unsuccessful synthesis attempts.
This project proposes a semi-supervised learning scheme to predict crystal synthesizability. The ML model is trained on experimental and theoretical crystal data. The initial featurization focuses on local environments which is inspired by the Pauling Rules. The experimental data points are downloaded through the Pymatgen API from the Materials Project database which contains relaxed structures recorded in Inorganic Crystal Structure Database – ICSD. The theoretical data is queried from select databases accessible through the Optimade project’s API.
Mit der Initiative MaterialDigital fördert das BMBF seit dem letzten Jahr ein wichtiges Instrument zur Digitalisierung der Materialwissenschaft und Werkstofftechnik in Deutschland. In der ersten Phase dieser Initiative wird seit Juli 2019 die Plattform MaterialDigital von einem Konsortium aufgestellt. Ziel dieser Plattform ist der Aufbau eines virtuellen Materialdatenraums, um zusammen mit allen Interessenten die Systematisierung des Umgangs mit Werkstoffdaten voranzutreiben. In einer zweiten Phase werden akademische F&E-Projekte voraussichtlich Anfang 2021 gefördert. In jedem dieser Projekte wird an einer konkreten Fragestellung das Themenfeld „Digitalisierung der Materialwissenschaft und Werkstofftechnik“ adressiert und multidisziplinär bearbeitet. In einer dritten Phase werden im Rahmen industriegeführter vorwettbewerblicher Verbundprojekte mit der gleichen Zielsetzung gefördert.
A deep insight into the chemistry and nature of individual chemical bonds is essential for understanding materials. Bonding analysis is expected to provide important features for large-scale data analysis and machine learning of material properties. Such information on chemical bonds can be calculated using the LOBSTER (www.cohp.de) software package, which post-processes data from modern density functional theory computations by projecting plane wave-based wave functions onto a local atomic orbital basis. We have performed bonding analysis on 1520 compounds (insulators and semiconductors) using a fully automated workflow combining the VASP and LOBSTER software packages. We then automatically evaluated the data with LobsterPy (https://github.com/jageo/lobsterpy) and provide results as a database. The projected densities of states and bonding indicators are benchmarked on VASP projections and available heuristics, respectively. Lastly, we illustrate the predictive power of bonding descriptors by constructing a machine-learning model for phononic properties, which shows an increase in prediction accuracies by 27 % (mean absolute errors) compared to a benchmark model differing only by not relying on any quantum-chemical bonding features.
Continuing progress in the field of X-ray scattering methods empowers scientists with new possibilities to capture the most important piece of information about the structure of the sample - its 3D electron density.
Although the first methods appeared almost a century ago, recovering the density structure of a sample is still very problematic. Most avail-able imaging techniques transform a 3D electron density of a realspace structure into the 2D Fourier Transform of the intensity of scattered waves in the reciprocal space. This process causes a loss of information.
Firstly, instead of a 3D sample, a 2D image is created, and secondly, the phase information of the scattered waves is lost. The latter is known as the ”phase problem” and poses a serious obstacle on a way to recover a 3D electron density. In this work, we draw attention to the problem of forward and inverse Small Angle X-Ray Scattering. In the first, forward, part, we rethink the existing pipelines to computationally simulate such scattering experiments. Although there are efficient implementations of fast Fourier transformation, they often have some drawbacks. For instance, to calculate a 3D fast Fourier transform it is required to place its density in the RAM. For high-resolution structures of size > 1024 3 , this becomes very problematic, as the whole density structure requires more than 16 GB of memory.
CUDA solution allows for a very fast and parallelizable implementation of high-resolution data on hundreds of last-generation machines.
Such computations are very pricy and inaccessible for most scientists.
To bypass this limitation, we propose a solution for a split-up 3D fast Fourier transform, which is implemented as a sequence of 2D and 1D operations. We compare our implementation on the simulated 3D shapes and show the result of a proof-of-concept on 4096 3 Metallorganic framework density structure. In the second, inverse problem, we train an invertible neural network, that given scattering data can predict the shape and its parameters. The architecture is built such, that the inverse problem is learned together with the forward process - the Fourier Transformation. We achieved very good results with this architecture, nonetheless, further testing is required, as the current training set only encompasses three simple shapes: sphere, hard sphere and cylinder.
All code to reproduce and analyze the results is available at https: //github.com/sofyalaski/SAXS-simulations.
Im Auftrag des Bundesministeriums für Wirtschaft und Klimaschutz haben DIN und DKE im Januar 2022 die Arbeiten an der zweiten Ausgabe der Deutschen Normungsroadmap Künstliche Intelligenz gestartet. In einem breiten Beteiligungsprozess und unter Mitwirkung von mehr als 570 Fachleuten aus Wirtschaft, Wissenschaft, öffentlicher Hand und Zivilgesellschaft wurde damit der strategische Fahrplan für die KI-Normung weiterentwickelt. Koordiniert und begleitet wurden diese Arbeiten von einer hochrangigen Koordinierungsgruppe für KI-Normung und -Konformität.
Mit der Normungsroadmap wird eine Maßnahme der KI-Strategie der Bundesregierung umgesetzt und damit ein wesentlicher Beitrag zur „KI – Made in Germany“ geleistet.
Die Normung ist Teil der KI-Strategie und ein strategisches Instrument zur Stärkung der Innovations- und Wettbewerbsfähigkeit der deutschen und europäischen Wirtschaft. Nicht zuletzt deshalb spielt sie im geplanten europäischen Rechtsrahmen für KI, dem Artificial Intelligence Act, eine besondere Rolle.
In this study, a wound dressing composed of an alginate dialdehyde−gelatin (ADA-GEL) hydrogel incorporated by astaxanthin (ASX) and 70B (70:30 B2O3/CaO in mol %) borate bioactive glass (BBG) microparticles was developed through 3D printing. ASX and BBG particles sti.ened the composite hydrogel construct and delayed its in vitro degradation compared to the pristine hydrogel construct, mainly due to their cross-linking role, likely arising from hydrogen bonding between the ASX/BBG particles and ADA-GEL chains. Additionally, the composite hydrogel construct could hold and deliver ASX steadily. The composite hydrogel constructs codelivered biologically active ions (Ca and B) and ASX, which should lead to a faster, more e.ective wound-healing process. As shown through in vitro tests, the ASX-containing composite hydrogel promoted fibroblast (NIH 3T3) cell adhesion, proliferation, and vascular endothelial growth factor expression, as well as keratinocyte (HaCaT) migration, thanks to the antioxidant activity of ASX, the release of cell-supportive Ca2+ and B3+ ions, and the biocompatibility of ADA-GEL. Taken together, the results show that the ADA-GEL/BBG/ASX composite is an attractive biomaterial to develop multipurposed wound-healing constructs through 3D printing.
Weltweit stellen Waldbrände eine große Bedrohung für Umwelt, Wirtschaft und menschliches Leben dar. Der fortschreitende Klimawandel verstärkt Trockenheit und Dürre, wodurch die Größe und Intensität von Waldbränden sowie das daraus resultierende Gefahrenpotential zusätzlich erhöht werden. Um im Risikofall eine schnelle und effektive Waldbrandbekämpfung zu gewährleisten, sollen Einsatzkräfte in Zukunft von KI-basierten Ausbreitungsmodellen unterstützt werden.
Der Einsatz von maschinellem Lernen ermöglicht dabei schnelle und zielgerichtete Ausbreitungsvorhersagen in Echtzeit, die in den Prozess der Brandbekämpfung als zusätzliche Information einfließen können. Im Rahmen des durch das EU-Förderprogram Horizon 2020 finanzierte Projekt TREEADS wird ein solches Modell für Europa entwickelt.
Dazu wird mit verschiedenen Satellitendaten des Copernicus-Programms und der NASA ein auf räumlicher und zeitlicher Ebene aufgelöster Datensatz zu europäischen Waldbränden aufgebaut. Zu den erfassten Waldbränden werden wichtigen Faktoren, die Einfluss auf die Brandausbreitung haben, ergänzt. Dazu gehören das von Copernicus bereitgestellte, digitale Geländemodell Europas und verschiedene, von Sentinel-2 Messungen abgeleitete Informationen zu Vegetation und Landnutzung. Dieser 2D-Trainingsdatensatz zu realen Waldbränden wird durch die zeitlich hochaufgelösten, meteorologischen Reanalyse-Produkte – Copernicus European Regional ReAnalysis (CERRA) und ERA5-Land – vervollständigt und ist die Grundlage für ein Modell zur Ausbreitungsvorhersage von Waldbränden.
Methoden des Maschinellen Lernens können komplexe Muster in den gesammelten Daten erkennen und so zu einer verbesserten Vorhersage der Brandausbreitung beitragen. Die vielfältigen, von Copernicus bereitgestellten Daten sind dabei ein unabdingbarer Bestandteil des Modells, mit dem den Einsatzkräften ein zusätzliches Tool für den effizienten Einsatz ihrer limitierten Ressourcen zur Verfügung steht.
Understanding the chemistry and nature of individual chemical bonds is essential for materials design. Bonding analysis via the LOBSTER software package has provided valuable insights into the properties of materials for thermoelectric and catalysis applications. Thus, the data generated from
bonding analysis becomes an invaluable asset that could be utilized as features in large-scale data analysis and machine learning of material properties. However, no systematic studies exist that conducted high-throughput materials simulations to curate and validate bonding data obtained from LOBSTER. Here we present an approach to constructing such a large database consisting of quantum-chemical bonding information.
This database consists of bonding data computed using Lobster for 1520 solid-state compounds consisting of insulators and semiconductors. It consists of two kinds of json files. Smaller lightweight JSONS consists of summarized bonding information for each of the compounds. The files are named as per ID numbers in the materials project database.
Here we provide also the larger computational data json files for 700 compounds. This files consists of all important LOBSTER computation output files data stored as dictionary.
This database consists of bonding data computed using Lobster for 1520 solid-state compounds consisting of insulators and semiconductors. The files are named as per ID numbers in the materials project database.
Here we provide the larger computational data JSON files for the rest of the 820 compounds. This file consists of all important LOBSTER computation output files data stored as a dictionary.
This talk highlights a proof-of-concept that demonstrates the ability to calculate high-resolution Fourier transforms. These can be combined with multi-scale modeling to simulate scattering over a wide range, from small-angle scattering to XRD and PDF.
The preprint documenting this is available on the ArXiv here:
https://doi.org/10.48550/arXiv.2303.13435
The Jupyter notebook, VASP calculation details and MOUSE measured scattering patterns are available from this Zenodo repository: https://dx.doi.org/10.5281/zenodo.7764045
We demonstrate a strategy for simulating wide-range X-ray scattering patterns, which spans the small- and wide scattering angles as well as the scattering angles typically used for Pair Distribution Function (PDF) analysis. Such simulated patterns can be used to test holistic analysis models, and, since the diffraction intensity is on the same scale as the scattering intensity, may offer a novel pathway for determining the degree of crystallinity.
The "Ultima Ratio" strategy is demonstrated on a 64-nm Metal Organic Framework (MOF) particle, calculated from Q < 0.01 1/nm up to Q < 150 1/nm, with a resolution of 0.16 Angstrom. The computations exploit a modified 3D Fast Fourier Transform (3D-FFT), whose modifications enable the transformations of matrices at least up to 8000^3 voxels in size. Multiple of these modified 3D-FFTs are combined to improve the low-Q behaviour. The resulting curve is compared to a wide-range scattering pattern measured on a polydisperse MOF powder. While computationally intensive, the approach is expected to be useful for simulating scattering from a wide range of realistic, complex structures, from (poly-)crystalline particles to hierarchical, multicomponent structures such as viruses and catalysts.
## Summary:
This notebook and associated datasets (including VASP details) accompany a manuscript available on the ArXiv (https://doi.org/10.48550/arXiv.2303.13435) and hopefully soon in a journal as short communication as well. Most of the details needed to understand this notebook are explained in that paper with the same title as above. For convenience, the abstract is repeated here:
## Paper abstract:
We demonstrate a strategy for simulating wide-range X-ray scattering patterns, which spans the small- and wide scattering angles as well as the scattering angles typically used for Pair Distribution Function (PDF) analysis. Such simulated patterns can be used to test holistic analysis models, and, since the diffraction intensity is presented coupled to the scattering intensity, may offer a novel pathway for determining the degree of crystallinity.
The ``Ultima Ratio'' strategy is demonstrated on a 64-nm Metal Organic Framework (MOF) particle, calculated from $Q<0.01$\,$\mathrm{nm}^{-1}$ up to $Q\approx150$\,$\mathrm{nm}^{-1}$, with a resolution of 0.16\,\AA. The computations exploit a modified 3D Fast Fourier Transform (3D-FFT), whose modifications enable the transformations of matrices at least up to $8000^3$ voxels in size. Multiple of these modified 3D-FFTs are combined to improve the low-$Q$ behaviour.
The resulting curve is compared to a wide-range scattering pattern measured on a polydisperse MOF powder.
While computationally intensive, the approach is expected to be useful for simulating scattering from a wide range of realistic, complex structures, from (poly-)crystalline particles to hierarchical, multicomponent structures such as viruses and catalysts.
Die Bereitstellung von zuverlässigen Werkstoffdaten stand für die Werkstoffprüfung seit dem 19. Jahrhundert in zentraler Stelle. Mit der zunehmender Digitalisierung unserer Gesellschaft gewinnt das Datenmanagement insbesondere im Hinblick auf die Modellierungs- und Simulationsaktivitäten an Bedeutung.
In unserem Beitrag beschreiben wir die gegenwärtigen Aktivitäten in Deutschland – insbesondere die NFDI-MatWerk und die BMBF-Initiative MaterialDigital – sowie in Europa und Übersee. Abschließend heben wir die Auswirkungen auf die technische und akademische Aus- und Weiterbildung.
Fatigue strength estimation is a costly manual material characterization process in which state-of-the-art approaches follow a standardized experiment and analysis procedure. In this paper, we examine a modular, Machine Learning-based approach for fatigue strength estimation that is likely to reduce the number of experiments and, thus, the overall experimental costs. Despite its high potential, deployment of a new approach in a real-life lab requires more than the theoretical definition and simulation. Therefore, we study the robustness of the approach against misspecification of the prior and discretization of the specified loads. We identify its applicability and its advantageous behavior over the state-of-the-art methods, potentially reducing the number of costly experiment.