8.1 Sensorik, mess- und prüftechnische Verfahren
Filtern
Erscheinungsjahr
- 2024 (10) (entfernen)
Dokumenttyp
- Beitrag zu einem Tagungsband (3)
- Vortrag (3)
- Posterpräsentation (3)
- Zeitschriftenartikel (1)
Sprache
- Englisch (10) (entfernen)
Schlagworte
- Gas Tomography (3)
- Plume (3)
- TDLAS (3)
- Cellulose (2)
- Colorimetric Sensor (2)
- Confetti (2)
- Drone (2)
- Gas distribution mapping (2)
- Graph neural networks (2)
- Indicator (2)
Organisationseinheit der BAM
- 8 Zerstörungsfreie Prüfung (10)
- 8.1 Sensorik, mess- und prüftechnische Verfahren (10)
- 1 Analytische Chemie; Referenzmaterialien (3)
- 1.5 Proteinanalytik (2)
- 9 Komponentensicherheit (2)
- 1.9 Chemische und optische Sensorik (1)
- 2 Prozess- und Anlagensicherheit (1)
- 2.1 Sicherheit von Energieträgern (1)
- 3 Gefahrgutumschließungen; Energiespeicher (1)
- 3.5 Sicherheit von Gasspeichern (1)
Eingeladener Vortrag
- nein (3)
We introduce a passive smart dust concept as a novel solution for environmental monitoring. Utilizing chemical reagents like colorimetric indicators and other chemosensors, these particles detect varying environmental conditions. We developed paper-based sensors that are both cost-effective and eco-friendly. In practical tests, these sensors, dispersed over a designated area, successfully identified hazardous substances by changing their color when exposed to acids or bases. This color change was remotely detectable using a drone-mounted color camera. The data thus obtained was processed through specialized software, accurately pinpointing areas of contamination. This method proves the efficacy and scalability of passive smart dust technology for real-time, environmentally sustainable remote sensing of hazardous materials
Use/distribution of cost-effective and biocompatible pieces of cellulose paper as passive optical chemosensors and readout by the camera system.
Modification of the paper surface using various chemically sensitive dyes (indicators) provides an optically detectable reaction and conclusions about the target substances and precise location.
This paper outlines significant advancements in our previously developed aerial gas tomography system, now optimized to reconstruct 2D tomographic slices of gas plumes with enhanced precision in outdoor environments. The core of our system is an aerial robot equipped with a custom-built 3-axis aerial gimbal, a Tunable Diode Laser Absorption Spectroscopy (TDLAS) sensor for CH4 measurements, a laser rangefinder, and a wide-angle camera, combined with a state-of-the-art gas tomography algorithm. In real-world experiments, we sent the aerial robot along gate-shaped flight patterns over a semi-controlled environment with a static-like gas plume, providing a welldefined ground truth for system evaluation. The reconstructed cross-sectional 2D images closely matched the known ground truth concentration, confirming the system’s high accuracy and reliability. The demonstrated system’s capabilities open doors for potential applications in environmental monitoring and industrial safety, though further testing is planned to ascertain the system’s operational boundaries fully.
Gas Distribution Mapping (GDM) is essential in monitoring hazardous environments, where uneven sampling and spatial sparsity of data present significant challenges. Traditional methods for GDM often fall short in accuracy and expressiveness. Modern learning-based approaches employing Convolutional Neural Networks (CNNs) require regular-sized input data, limiting their adaptability to irregular and sparse datasets typically encountered in GDM. This study addresses these shortcomings by showcasing Graph Neural Networks (GNNs) for learningbased GDM on irregular and spatially sparse sensor data. Our Radius-Based, Bi-Directionally connected GNN (RABI-GNN) was trained on a synthetic gas distribution dataset on which it outperforms our previous CNN-based model while overcoming its constraints. We demonstrate the flexibility of RABI-GNN by applying it to real-world data obtained in an industrial steel factory, highlighting promising opportunities for more accurate GDM models.
This paper explores combining a gimbal-mounted tunable diode laser absorption spectroscopy (TDLAS) sensor and a video camera to create a virtual gas camera for methane leak detection. This provides a low-to-zero-cost extension of typical TDLAS gas tomography systems. A prototype setup mounted on a ground robot is evaluated. Results acquired using a simulated methane leak show the feasibility of the virtual gas camera, accurately detecting methane leaks by overlaying concentrations onto a visual image. While the acquisition time is significantly longer than for traditional gas cameras, potential enhancements are discussed. The study concludes that the virtual gas camera is feasible and useful, despite its longer acquisition time. It serves as a valuable software-only addition to typical TDLAS gas tomography systems, offering quickly-available on-site data augmentation for visual leak assessment at low-to-zero cost.
This paper outlines significant advancements in our previously developed aerial gas tomography system, now optimized to reconstruct 2D tomographic slices of gas plumes with enhanced precision in outdoor environments. The core of our system is an aerial robot equipped with a custom-built 3-axis aerial gimbal, a Tunable Diode Laser Absorption Spectroscopy (TDLAS) sensor for CH4 measurements, a laser rangefinder, and a wide-angle camera, combined with a state-of-the-art gas tomography algorithm. In real-world experiments, we sent the aerial robot along gate-shaped flight patterns over a semi-controlled environment with a static-like gas plume, providing a welldefined ground truth for system evaluation. The reconstructed cross-sectional 2D images closely matched the known ground truth concentration, confirming the system’s high accuracy and reliability. The demonstrated system’s capabilities open doors for potential applications in environmental monitoring and industrial safety, though further testing is planned to ascertain the system’s operational boundaries fully.
Gas Distribution Mapping (GDM) is essential in monitoring hazardous environments, where uneven sampling and spatial sparsity of data present significant challenges. Traditional methods for GDM often fall short in accuracy and expressiveness. Modern learning-based approaches employing Convolutional
Neural Networks (CNNs) require regular-sized input data, limiting their adaptability to irregular and sparse datasets typically encountered in GDM. This study addresses these shortcomings by showcasing Graph Neural Networks (GNNs) for learningbased GDM on irregular and spatially sparse sensor data. Our Radius-Based, Bi-Directionally connected GNN (RABI-GNN) was trained on a synthetic gas distribution dataset on which it outperforms our previous CNN-based model while overcoming its constraints. We demonstrate the flexibility of RABI-GNN by applying it to real-world data obtained in an industrial steel factory, highlighting promising opportunities for more accurate GDM models.
Microalloying elements such as Nb and Ti play a decisive function in achieving the desired mechanical strength of quenched and tempered, high-strength fine-grain structural steels with a nominal yield strength ≥ 690 MPa. The current specifications for the chemical composition only provide manufacturers with upper limits. However, even minor deviations in the alloy concept can have a significant impact on the mechanical properties. Consequently, accurate prediction of weldability and the integrity of welded joints becomes difficult or even impossible due to differences in composition and the resulting microstructures. Undesirable consequences include a possible softening of the heat-affected zone (HAZ) or, conversely, hardening effects. In view of these challenges, various microalloying strategies with different Ti and Nb contents are being systematically investigated for the first time using specially developed laboratory casting alloys. Each alloying route is based on the common S690QL, maintaining both the chemical composition and the heat treatment parameters. To analyse the weldability, three-layer welds were performed using gas metal arc welding (GMAW) and critical microstructures, such as areas of the heat-affected zone (HAZ) that exhibit significant softening or hardening, were identified. The effect of the softened HAZ region on failure was evaluated using transverse tensile specimens. Digital image correlation (DIC) is used to image changes in local strains in different HAZ regions in situ. Using a specially developed mirror system, the local strains of the microstructure zones on the top and bottom of the weld are recorded simultaneously. This makes it possible to analyse how the weld seam geometry (e.g., V-seam) influences the strain gradients. In addition, the analysis of the local deformation helps to understand the effects of the softened HAZ on the global strain, the fracture constriction, the fracture position, and the overall fracture behavior.
With the reduction of large oil spills because of stricter regulations and safety measures, the question of how to manage smaller oil spills arises. Few on-site analytical tools are available for first responders or other law enforcement personnel to rapidly test for crude oil in the early management of localized polluted areas. The approach reported here relies on well-described computer-assisted multivariate data analysis of the intrinsic fluorescence fingerprints of crude oils to build a multivariate model for the rapid classification of crude oils and the prediction of their properties. Thanks to a dedicated robust portable reader, the method allowed classification and accurate prediction of various properties of crude oil samples like density (according to API, the American Petroleum Institute and viscosity as well as composition parameters such as volume fractions of paraffins or aromatics. In this way, autonomous operation in on-site or in-the-field applications becomes possible based on the direct (undiluted and untreated) measurement of samples and a rapid, tablet-operated readout system to yield a robust and simple analytical test with superior performance. Testing in real-life scenarios allowed the successful classification and prediction of a number of oil spill samples as well as weathered samples that closely resemble samples collected by first responders.