Analytische Chemie
Filtern
Dokumenttyp
- Vortrag (3)
- Zeitschriftenartikel (2)
- Forschungsdatensatz (1)
Schlagworte
- XCT (6) (entfernen)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (3)
The mixture of ammonium nitrate (AN) prills and fuel oil (FO), usually referred to as ANFO, is extensively used in the mining industry as a bulk explosive. One of the major performance predictors of ANFO mixtures is the fuel oil retention, which is itself governed by the complex pore structure of the AN prills. In this study, we present how X-ray computed tomography (XCT), and the associated advanced data processing workflow, can be used to fully characterise the structure and morphology of AN prills. We show that structural parameters such as volume fraction of the different phases and morphological parameters such as specific surface area and shape factor can be reliably extracted from the XCT data, and that there is a good agreement with the measured oil retention values. Importantly, oil retention measurements (qualifying the efficiency of ANFO as explosives) correlate well with the specific surface area determined by XCT. XCT can therefore be employed non-destructively; it can accurately evaluate and characterise porosity in ammonium nitrate prills, and even predict their efficiency.
It is shown that preconditioning of experimental X‐ray computed tomography (XCT) data is critical to achieve high‐precision segmentation scores. The challenging experimental XCT datasets and deep convolutional neural networks (DCNNs) are used that are trained with low‐resemblance synthetic XCT data. The material used is a 6‐phase Al–Si metal matrix composite‐reinforced with ceramic fibers and particles. To achieve generalization, in our past studies, specific data augmentation techniques were proposed for the synthetic XCT training data. In addition, two toolsets are devised: (1) special 3D DCNN architecture (3D Triple_UNet), slicing the experimental XCT data from multiple views (MultiView Forwarding), the i.S.Sy.Da.T.A. iterative segmentation algorithm, and (2) nonlocal means (NLM) conditioning (filtering) for the experimental XCT data. This results in good segmentation Dice scores across all phases compared to more standard approaches (i.e., standard UNet architecture, single view slicing, standard single training, and NLM conditioning). Herein, the NLM filter is replaced with the deep conditioning framework BAM SynthCOND introduced in a previous publication, which can be trained with synthetic XCT data. This leads to a significant segmentation precision increase for all phases. The proposed methods are potentially applicable to other materials and imaging techniques.
We employ in-house generated synthetic Al-Si matrix composite XCT data for training deep convolutional neural networks for XCT data conditioning and automatic segmentation. We propose an in-house multilevel deep conditioning framework capable of rectifying noise and blur in corrupted XCT data sequentially. Furthermore, for automatic segmentation, we utilize a special in-house network coupled with a novel iterative segmentation algorithm capable of generalized learning from synthetic data. We report a consistent SSIM efficiency of 92%, 99%, and 95% for the combined denoising/deblurring, standalone denoising, and standalone deblurring, respectively. The overall segmentation precision was over 85% according to the Dice coefficient. We used experimental XCT data from various scans of Al-Si matrix composites reinforced with ceramic particles and fibers.
Die Entwicklung von Werkzeugen zur realitätsnahen Nachbildung eines industriellen CT-Systems in einer Simulationssoftware ist derzeit Hauptaufgabe des WIPANO Forschungsprojektes CTSimU2 Realistische Simulation realer Röntgencomputertomografie - Systeme mit basisqualifizierter Simulationssoftware. Als Voraussetzung dienen dabei Simulationssoftwares, die durch das Testframework aus dem Vorprojekt CTSimU1 basisqualifiziert wurden. Das Testframework testet die hinreichende physikalische Korrektheit und Funktionalität einer Simulationssoftware (Basisqualifizierung der Software). Für eine realitätsnahe Nachbildung ist nicht nur die Güte der Simulationssoftware, sondern insbesondere die Güte der Parametrisierung des realen CT-Systems in der Simulationssoftware ausschlaggebend. Dabei kann das Vorgehen der Parametrisierung in vier Schritte unterteilt werden: die Datenaufnahme am realen CT-System (Schritt 1), die Auswertung der aufgenommenen Daten für die Generierung allgemeiner Parameterangaben (Schritt 2), die Übertragung der Parameter in die spezifischen Simulationssoftwares (Schritt 3) und die Validierung der resultierenden Simulationsparameter durch einen geeigneten Test (Schritt 4). Ziel des Projektes ist es daher neben der Erarbeitung eines Werkzeugkastens mit allgemeinen Methoden zur Datenaufnahme und Auswertung der Daten, die Entwicklung eines Tests, auf dessen Basis die ausreichend korrekte Simulation einer realen Anlage beurteilt werden kann. Die erarbeiteten Ergebnisse sollen wie bereits im Vorprojekt CTSimU1 in einen Richtlinienentwurf für die Richtlinienreihe VDI/VDE 2630 übertragen werden. Dieser Beitrag soll einen Überblick über das Projekt und die ersten Ergebnisse geben.