BAM-Publica - Publikationsserver der Bundesanstalt für Materialforschung und -prüfung
Not a member yet
58839 research outputs found
Sort by
Towards Safety and Accuracy of Hydrogen Refuelling Stations through Digital Twins
No full textThis paper presents digital quality infrastructure methods for hydrogen refueling stations using the Asset Administration Shell as a standardized digital twin. Implemented at BAM’s test platform, it integrates real-time sensor data, calibration certificates, and compliance documents to support traceable, interoperable asset management. In combination with AI and semantic tools, the system will enable predictive maintenance, remote audits, and improved safety. This approach reduces downtime, enhances transparency, and offers a scalable model demonstrating the potential of digital twins in advancing metrological traceability and operational efficiency in hydrogen technologies
Robot-Assisted Automated Serial-Sectioning and Imaging for 3D Microstructural Investigations
No full textComprehensive materials characterization requires precise structural knowledge beyond traditional methods. The robot-assisted automated serial-sectioning and imaging (RASI) platform, developed at BAM, provides automated 3D metallographic reconstructions, enabling detailed microstructural analysis of technical materials. This article showcases RASI’s capabilities through several case studies, including characterization of lamellar graphite in gray cast iron, porosity in sintered steel, melt pool morphology in additively manufactured 316L stainless steel, defects in metal-ceramic packages, and oxidation behavior in an Fe-12Cr-2Co alloy. By automating sample handling, mechanical serial-sectioning, etching, and optical imaging, RASI captures complex 3D microstructures with high precision and at high speed. This approach reveals microstructural features missed by 2D analysis, even using stereological assumptions. Specifically, statistically rare and large microstructural features, such as secondary phases or interconnected pores, become apparent, which 2D methods cannot reveal. The generated volumetric data can furthermore serve as quantitative reference datasets (i.e., the ‘ground truth’) essential for validating other 3D characterization techniques and computational models, helping to bridge the gap between predictive simulations and real-world material behavior. RASI’s modular design makes it a flexible tool that provides realistic 3D insights into materials, which can be used for advanced materials research, process optimization, and quality control
The evolution of symbiont and immune system interactions in Blattodea
No full textImmunity plays an important role in evolutionary ecology. The immune system interacts with both pathogens, which can act as important selective forces, and symbionts, which are regulated by the host‘s immune system and can also play a role in the host‘s immune defenses.
The Blattodea consist of what is commonly known as cockroaches and the termites, who are themselves specialized eusocial cockroaches. Blattodea have two symbiont types that are characteristic to either the cockroaches or the termites (Table 1). These symbiont transitions open up the opportunity to characterize and investigate host-symbiont-pathogen interactions and how they evolved in Blattodea
Hydrogen enhanced dislocation cross-slip in polycrystalline nickel
No full textHydrogen embrittlement (HE), degradation of the mechanical properties of metals due to the presence of hydrogen, is a persistent problem that has been attracting the attention of the material science community for about fifteen decades. Extensive experimental observations indicate the presence of nanovoids and the increase of free volume at the grain boundaries in hydrogen contaminated metals. This rate-dependent phenomenon motivates theoretical investigations of the underlying mechanisms. Here, a hydrogen enhanced cross-slip (HECS) mechanism in the close vicinity of the grain boundaries is demonstrated by direct molecular dynamics simulations and theoretical calculations. To this end, the interaction of screw dislocations with a variety of symmetric tilt grain boundaries in H-charged and H-free bicrystalline nickel is examined. The presence of segregated H atoms at the grain boundaries induces a stress field in their vicinity, and thus,- the barrier for cross-slip of screw dislocations considerably decreases. The enhanced cross-slip of dislocations facilitates the formation of jogs on bowedout dislocations. These jogs can form vacancies during the glide process. This mechanism of defect production shows nanoscale evidence of enhanced vacancy formation and subsequent increase in the free volume along the grain boundaries in the presence of H
Multispektrale optische Tomografie (MK-OT) trifft MachineLearning
No full textDas pulverbettbasierte Laserstrahlschmelzen von Metallen (PBF-LB/M) zählt zu den am weitesten verbreiteten additiven Fertigungsverfahren für metallische Bauteile. Trotz zunehmender industrieller Relevanz, etwa in Luft- und Raumfahrt, Medizintechnik und Energieanwendungen, bleibt eine umfassende Prozessüberwachung und Qualitätssicherung eine zentrale Herausforderung. Die hohe Komplexität des Verfahrens, bedingt durch eine Vielzahl interagierender Prozessparameter, sowie die resultierende Datenfülle erschweren die direkte Korrelation zwischen Prozessanomalien und resultierenden Bauteildefekten wie Poren oder Rissen.
Zahlreiche Monitoring Ansätze, insbesondere thermografische Verfahren, wurden bereits intensiv untersucht. Dennoch fehlt es bislang an praxistauglichen, wirtschaftlich skalierbaren Lösungen. Insbesondere die Kosten und begrenzte Sichtfelder vieler Systeme hemmen eine breite industrielle Anwendung.
Oster et al. [1] konnten zeigen, dass die Analyse kurzwelliger Infrarotstrahlung (SWIR-Thermografie) mittels auf künstlicher Intelligenz (KI) basierender Auswertung zur ortsaufgelösten Detektion von Porosität im PBF-LB/M-Prozess geeignet ist. Ihr Ansatz erlaubt erstmals die Identifikation nicht gezielt induzierter Poren, zeigt jedoch Einschränkungen hinsichtlich des erfassten Sichtfelds (~9 × 10 mm²) und der Systemkosten (~30k€).
Eine vielversprechende Alternative stellt die Multikanal Optische Tomografie (MK-OT) dar [2], die auf kostengünstigen visuellen Kameras basiert. Durch gleichzeitige Erfassung mehrerer Spektralbereiche lassen sich Schichtbilder mit hoher räumlicher Auflösung und verbesserter Robustheit, verglichen mit marktüblicher monospektralen OT, gegenüber prozessbedingten Schwankungen erzeugen [2]. Die auch hierbei anfallenden großen Datenmengen erfordern eine automatisierte Auswertung. Bereits Ero et al. [3] und Feng et al. [4] konnten zeigen, dass mittels KI Porositätsvorhersagen aus herkömmlichen monochromatischen OT Daten gewonnen werden können.
Ziel dieses Beitrags ist es, das Vorgehen zur Untersuchung der Übertragbarkeit des in [1] vorgestellten KI-basierten Auswerteverfahrens auf das MK-OT-System vorzustellen. Hierdurch könnte eine kosteneffiziente und skalierbare Prozessüberwachung mittels Porositätsbestimmung für den industriellen Einsatz ermöglicht werden. Der aktuelle Stand der Arbeit und das geplante Vorgehen wird dargelegt
Phenotype and function of human monocytes remain mainly unaffected by very small superparamagnetic iron oxide particles
No full textThe field of medical application of organic or inorganic nanoparticles is extensive. Medical nanoparticles offer benefits but pose risks. For safe use in diagnostics and therapy, they should be inert, non-immunogenic, non-aggregating, and avoid long-term accumulation in sensitive tissues like bone marrow or the brain. We have developed in-house very small superparamagnetic iron oxide nanoparticles (VSOP), 7 nm in size, which have been successfully used in preclinical magnetic resonance imaging (MRI) to detect intestinal inflammation, neuroinflammation and atherosclerosis. This study examines nanoparticle effects on human blood cells focusing on monocytes in vitro as a first step toward clinical application. Whole blood and monocytes from healthy donors and patients with inflammatory bowel disease were treated with VSOP in vitro and analyzed for changes in their transcriptome, phenotype and function. RNA sequencing of monocytes identified the transferrin receptor as one of the most significantly downregulated genes after VSOP treatment, likely to limit iron uptake. Whereas whole blood RNA sequencing showed significant changes only in three non-coding genes. CyTOF analysis confirmed that VSOP-treated monocytes remain inactive, with no increased proliferation or altered migration. Metabolically, VSOP uptake enhanced the oxygen consumption rate. This effect was likely due to phagocytosis rather than effects mediated by the VSOP itself, as phagocytosis of latex beads showed comparable results. In summary, the analysis of peripheral blood mononuclear cells and monocytes suggests that VSOP treatment has no major impact on immune cell phenotype or function indicating VSOP as a promising diagnostic tool in MRI for inflammatory bowel disease
Near-real-time in-situ powder bed anomaly detection using machine learning algorithms for high-resolution image analysis in PBF-LB/M
No full textIn-situ captured visual images of the laser powder bed fusion process (PBF-LB/M) provide valuable insights into process dynamics. Automatic analysis of after-recoating images using machine learning algorithms enables the detection of process deviations to reduce scrap production. However, current industrial monitoring systems for PBF-LB/M are limited by low image resolution. While higher resolutions enable the system’s ability to capture smaller features, they increase storage and computational demand.
Edge devices offer a solution by enabling near-real-time, on-premises image analysis within the machine and company network. In this study, high-resolution after-recoating images, captured with a spatial resolution of 17 µm/pixel and an image size of 9344 x 7000 pixels, were processed on an Nvidia Jetson Orin NX16 edge device. The images were downscaled, and anomaly detection algorithms were used to identify regions of interest for segmentation and classification at full resolution. To address computational constraints, state-of-the-art anomaly detection algorithms were evaluated and an appropriate downscaling factor for the on-edge implementation was determined. The EfficientAD algorithm achieved promising results, detecting anomalies within an inference time of less than 10 seconds. The presented framework enables anomaly detection with a maximum delay of one layer. This lays the foundation for the future development of near-real-time intervention in the PBF-LB/M process
A Python workflow definition for computational materials design
No full textNumerous Workflow Management Systems (WfMS) have been developed in the field of computational materials science with different workflow formats, hindering interoperability and reproducibility of workflows in the field. To address this challenge, we introduce here the Python Workflow Definition (PWD) as a workflow exchange format to share workflows between Python-based WfMS, currently AiiDA, jobflow, and pyiron. This development is motivated by the similarity of these three Python-based WfMS, that represent the different workflow steps and data transferred between them as nodes and edges in a graph. With the PWD, we aim at fostering the interoperability and reproducibility between the different WfMS in the context of Findable, Accessible, Interoperable, Reusable (FAIR) workflows. To separate the scientific from the technical complexity, the PWD consists of three components: (1) a conda environment that specifies the software dependencies, (2) a Python module that contains the Python functions represented as nodes in the workflow graph, and (3) a workflow graph stored in the JavaScript Object Notation (JSON). The first version of the PWD supports directed acyclic graph (DAG)-based workflows. Thus, any DAG-based workflow defined in one of the three WfMS can be exported to the PWD and afterwards imported from the PWD to one of the other WfMS. After the import, the input parameters of the workflow can be adjusted and computing resources can be assigned to the workflow, before it is executed with the selected WfMS. This import from and export to the PWD is enabled by the PWD Python library that implements
the PWD in AiiDA, jobflow, and pyiron
Mikrostrukturspezifische Wasserstoffdiffusion in UP-Schweißverbindungen hochfester Stahlgrobbleche
No full textHochfeste Baustähle sind für den sogenannten, modernen Stahlleichtbau im Gebäude-, Anlagen- oder Mobilkranbau nicht mehr wegzudenken. Durch den Einsatz dieser Stähle mit Streckgrenzen ≥ 690 MPa können durch die Reduzierung der Wanddicke kann eine erhebliche konstruktive Gewichtsreduzierung erreicht werden. Dies führt zudem zu weiteren sekundären Vorteilen, wie geringeren Schweißverarbeitungskosten, da u.a. die zu füllenden Nahtquerschnitte kleiner sind. Für hochfeste Grobbleche kommt dabei insbesondere das Unterpulverschweißen (UP) zum Einsatz, das durch seine hohe Abschmelzleistung gekennzeichnet ist. Allerdings haben hochfeste Stahlgrobbleche aufgrund ihrer Mikrostruktur eine von vornherein begrenzte Duktilität ggü. niederfesten Stählen und sind per se anfälliger für verzögerte, wasserstoffunterstützte Kaltrissbildung Zudem führt die große Bleckdicke einerseits zu hoher konstruktiver Steifigkeit der Komponenten (mit der Folge erhöhter Eigenspannungen) und andererseits durch die dicken Schweißlagen zu langen Diffusionswege für den Wasserstoff, welcher bspw. durch feuchtes Schweißpulver in die Naht gelangen kann. Hieraus ergeben sich zwei Schwierigkeiten: (1) bis zu welcher Zeit mit einer verzögerten Rissbildung bei Raumtemperatur zu rechnen, wenn keine weitere Wärmebehandlung zur Reduktion des Wasserstoffes erfolgt bzw. (2) wenn diese notwendig ist, bei welche Temperatur dies erfolgen sollte. Dazu sind abgesicherte Diffusionskoeffizienten für den Wasserstoff in UP-Schweißungen notwendig, u.a. für numerische Simulationen. Diese Koeffizienten sind bisher nur äußert lückenhaft verfügbar. Aus diesem Grund wurden mikrostrukturspezifische, elektrochemische Permeationsversuche (nach ISO 17081) und Warmauslagerungsversuche mit TGHE an UP-Schweißverbindungen durchgeführt. Dazu wurden ein thermomechanisch (TM) gewalzter bzw. vergüteter (QT) Grundwerkstoff betrachtet, sowie das Schweißgut und die Wärmeeinflusszone (WEZ). Interessanterweise (und im Gegensatz zu Effekten des Wärmebehandlungszustandes auf die Diffusion in MSG-Schweißverbindungen) zeigten die UP-Schweißmikrostrukturen kaum signifikante Unter-schiede im Diffusionsverhalten von WEZ und Schweißgut. Dies ist auf den positiven Ef-fekt des mehrfachen Anlassens der Mikrostruktur durch die Mehrlagenschweißung zu-rückzuführen. Aus praktischer Anwendersicht können daher dickwandige UP-Verbindungen ausschließlich anhand der einfach ermittelbaren Diffusionskoeffizienten für den Grundwerkstoff beurteilt werden. Zudem zeigte sich, dass der Walzzustand (QT vs. TM) in Grobblechen gegenüber Dünnblechen eine untergeordnete Rolle für die Wasserstoffdiffusion bildet. Ergänzende numerische Simulationen zeitabhängigen Wasserstoffdiffusion bestätigten das Verhalten
Numerical and Experimental Investigation of Deformation Induced Martensitic Transformation in Fused Filament Fabricated Austenitic Stainless Steel for Cryogenic Applications
No full textCryogenic structural components, such as collars, bladders, keys for superconducting magnets, and elements of liquid hydrogen storage systems like hoses and valves, are frequently constructed from austenitic stainless steel due to its favorable properties. However, manufacturing these components using traditional methods is challenging due to their complex geometries. Additive manufacturing emerges as a promising solution, though a comprehensive understanding of the associated material behavior under extrem e conditions is still developing.
This study aims to explore the deformation induced martensitic transformation (DIMT) in fused
filament fabricated (FFF) 316L stainless steel through both experimental testing and numerical
simulation. The research focuses on predicting the material’s respo nse under tensile stress at ambient, 77K, and 4K temperatures.
Numerical simulations employ a finite element approach to incorporate the constitutive model and its temperature dependent phase transformation kinetics, enabling detailed investigation of stress and strain distributions at various cryogenic temperatures. These simulations are systematically calibrated and validated against corresponding experimental datasets, ensuring that the computational predictions mirror the observed microstructural evolution and macroscopic response under tensile loading. By comparin g simulation results to experimental findings obtained at temperatures from room temperature down to 4K, the reliability of the model can be assessed, and its predictive capabilities can be refined.
Ultimately, the research seeks to expand the understanding of DIMT in additively manufactured 316L components, supporting the development of advanced, simulation driven material models tailored for demanding cryogenic structural applications