Filtern
Dokumenttyp
- Posterpräsentation (305) (entfernen)
Referierte Publikation
- nein (305)
Schlagworte
- Additive manufacturing (20)
- Computed tomography (17)
- LIBS (15)
- Additive Manufacturing (14)
- Thermography (13)
- Ultraschall (13)
- Ultrasound (10)
- Mobile Robot Olfaction (9)
- X-ray refraction (9)
- BAMline (8)
Organisationseinheit der BAM
- 8 Zerstörungsfreie Prüfung (305) (entfernen)
Current capabilities for full-scale field testing are highly resource intensive. Reliable small-scale experiments are an effective alternative. Characterization of the dynamic response and damage of RC elements to scaled blast loads was investigated in scaled-down field experiments. Spatially resolved information on the dynamic structural response using distributed fiber optic acoustic sensing (DAS) and acceleration as well as blast loading by piezoelectric pressure sensors.
Bei der Betonage von Abdichtbauwerken können Temperaturen bis 110 °C und hohe Drücke bis 5 MPa erreicht werden. Mittels Technikumsversuchen wurde nachgewiesen, dass Mikroakustik- und Ultraschallsensoren zum Einbau in Demonstrationsbauwerke geeignet sind.
Die Untersuchungen wurden bei der BAM (Bundesanstalt für Materialforschung und -prüfung) in Berlin durchgeführt. Es ist vorgesehen, die Sensoren in Morsleben am Demonstrationsbauwerk im Anhydrit einzusetzen.
Im Rahmen des Forschungsprojekts "Artificial Intelligence for Rail Inspection" (AIFRI) wird ein KI-Algorithmus entwickelt, um die Fehlererkennung und Bewertung bei der Auswertung von Schienenprüfungen mittels Ultraschall- und Wirbelstromprüfverfahren zu verbessern. Die Bandbreite möglicher Defekte und die Menge an Einflussgrößen auf die Schienenprüfung ist sehr groß, aber die Prüfdaten aus dem Feld bilden diese Bandbreite nicht balanciert ab und sind unzureichend gelabelt. Durch Simulationen werden große Mengen detailliert gelabelter Daten für relevante Schienenschädigungen und Artefakte bereitgestellt. Aus diesen Daten werden wiederum virtuellen Prüffahrten erstellt, die für das Training und die Validierung der KI genutzt werden können.
In unserem Poster stellen wir die Simulation dieser Ultraschall- und Wirbelstromprüfdaten vor. Dabei gehen wir im Detail auf die Themen CAD-Modellierung, Modellparameter, Simulationssoftware und Signalverarbeitung ein. Wir stellen ausgewählt die Simulation von Head Checks für Wirbelstromprüfverfahren und Bohrungsanrissen für Ultraschallprüfverfahren dar, da das Erkennen dieser Schädigungstypen bei der Schienenprüfung hohe Priorität innehat.
Für die Verifikation der Simulationsergebnisse verweisen wir auf den ebenfalls eingereichten Vortrag „Verifikation von Ultraschall- und Wirbelstromsimulationen bei der Schienenprüfung“.
Das Projekt AIFRI wird im Rahmen der Innovationsinitiative mFUND unter dem Förderkennzeichen 19FS2014 durch das Bundesministerium für Digitales und Verkehr gefördert.
Wir präsentieren unsere Forschungsergebnisse zur Entwicklung eines faseroptischen Sensorsystems mit polymeroptischen Fasern integriert in Geokunststoffe für die Überwachung von geotechnischen Anlagen. Das Sensorsystem dient der räumlichen Auflösung lokaler Dehnungsereignisse im für Anwender signifikanten Größenordnungsbereich von 3 % bis 10 %. Das gesamte Sensorkonzept besteht aus einem praxistauglichen Messsystem basierend auf der digitalen inkohärenten optischen Frequenzbereichsreflektometrie (I-OFDR) und einer Sensormatte in Form eines Geotextils mit integrierten faseroptischen perfluorierten polymeroptischen Fasern (PF-POFs).
Currently, mandatory requirements and recommendations for the detection of irregularities in laser beam welded joints are based on classic micrographs as set out in the standard ISO 13919-1:2019. Compared to classic micrographs, computed tomography enables a non-destructive, three-dimensional and material-independent mode of operation, which delivers much more profound results. Even in building material testing, methods with limited informative value can be checked and supplemented by CT examinations.
Communities worldwide face significant threats from Explosive Remnants of War (ERW), which endanger lives and restrict land usage. From forest fires due to ERWs or in ERW-contaminated areas (e.g., in Jüterbog, Germany) to broader global challenges (e.g., the Ukrainian conflict), the need for efficient detection and removal of these remnants, especially for humanitarian demining, is paramount. Traditional methods, like manual demining, have severe limitations in safety and efficiency. Here, we introduce an innovative solution to these challenges: “Chemosensing Smart Dust.” This technology uses chemoselective dyes that change their fluorescence properties when exposed to explosives like 2,4,6-trinitrotoluene (TNT). Fluorescence-based detection offers superior sensitivity, reduced likelihood of false positives, and enhanced accuracy of explosive detection. Drones, equipped with excitation lasers or LEDs, deploy the Chemosensing Smart Dust over areas of interest and actively detect the fluorescence changes using high-resolution cameras, offering a rapid, safe, and adaptable detection method. Beyond demining, this innovative approach has potential applications in monitoring polluted areas, homeland security, and emergency response.
The deployment of machine learning (ML) and deep learning (DL) in structural health monitoring (SHM) faces multiple challenges. Foremost among these is the insufficient availability of extensive high-quality data sets essential for robust training. Within SHM, high-quality data is defined by its accuracy, relevance, and fidelity in representing real-world structural scenarios (pristine as well as damaged). Although methods like data augmentation and creating synthetic data can add to datasets, they frequently sacrifice the authenticity and true representation of the data. Sharing real-world data encapsulating true structural and anomalous scenarios offers promise. However, entities are often reluctant to share raw data, given the potential extraction of sensitive information, leading to trust issues among collaborating entities.
Our study introduces a novel methodology leveraging Federated Learning (FL) to navigate these challenges. Within the FL framework, models are trained in a decentralized manner across different entities, preserving data privacy. In our research, we simulated several scenarios and compared them to traditional local training methods. Employing guided wave (GW) datasets, we distributed the data among different parties (clients) using IID (independent, identically distributed or in other words, statistically identical) mini batches of dataset, as well as non-IID configurations. This approach mirrors real-world data distribution among varied entities, such as hydrogen refueling stations.
In our methodology, the initial round involves individualized training for each client using their unique datasets . Subsequently, the model parameters are sent to the FL server, where they are averaged to construct a global model. In the second round, this global model is disseminated back to the clients to aid in predictive tasks. This iterative process continues for several rounds to reach convergence.
Our findings distinctly highlight the advantages of FL over localized training, evidenced by a marked improvement in prediction accuracy . This research underscores the potential of FL in GW-based SHM, offering a remedy to similar challenges tied to data scarcity in other SHM approaches and paving the way for a new era of collaborative, data-centric monitoring systems.
Due to their increasing energy density, lithium-ion-batteries (LIBs) play a key role in the traffic energy transition. Regarding their safety behavior, the main challenge of LIB-cells remains the thermal runaway (TR) process. In situ/operando investigations of the TR on commercial cells is possible with radiographic and computer tomographic measurements. Nonetheless, high resolution visualization of the TR persists as a challenge due to the high progression speed of the TR-process itself. Generally, performing abuse tests at cryogenic temperatures allows to slow down or even prevent the TR. Nevertheless, not all abuse methods are suitable for TR investigations at low temperatures. Nail penetration is an appropriate option, however, contains numerous unknown parameters and therefore suffers regarding reproducibility.
Herein, a self-developed high precision nail-penetration-setup is introduced, approaching the necessary mechanically reproducibility with controlled temperatures down to -190°C. The setup allows the preparation of critically abused, however, at cryogenic temperatures stable LIB-cells. These cells were controlled rethermalized to room temperature during synchrotron x-ray computer tomography (SXCT) with a pixel size up to 0.7 μm. During this measurement, the temperature and voltage of the cell is monitored allowing the visualization of the initial internal cell reactions. This study reveals the relation between internal reactions and cell voltage. Finally, the developed set-up enables in-depth analysis of thermal runaway behavior down to material level for various commercial battery cells in the future.