Multi-image super-resolution (MISR) usually outperforms single-image super-resolution (SISR) under a proper inter-image alignment by explicitly exploiting the inter-image correlation. However, the large computational demand encumbers the deployment of MISR in practice. In this work, we propose a distributed optimization framework based on data parallelism for fast large-scale MISR using multi-GPU acceleration named FL-MISR. The scaled conjugate gradient (SCG) algorithm is applied to the distributed subfunctions and the local SCG variables are communicated to synchronize the convergence rate over multi-GPU systems towards a consistent convergence. Furthermore, an inner-outer border exchange scheme is performed to obviate the border effect between neighboring GPUs. The proposed FL-MISR is applied to the computed tomography (CT) system by super-resolving the projections acquired by subpixel detector shift. The SR reconstruction is performed on the fly during the CT acquisition such that no additional computation time is introduced. FL-MISR is extensively evaluated from different aspects and experimental results demonstrate that FL-MISR effectively improves the spatial resolution of CT systems in modulation transfer function (MTF) and visual perception. Comparing to a multi-core CPU implementation, FL-MISR achieves a more than 50$$\times$$ × speedup on an off-the-shelf 4-GPU system.
This paper presents the non-uniform change in cell thickness of cylindrical Lithium (Li)-ion cells due to the change of State of Charge (SoC). Using optical measurement methods, with the aid of a laser light band micrometer, the expansion and contraction are determined over a complete charge and discharge cycle. The cell is rotated around its own axis by an angle of α=10° in each step, so that the different positions can be compared with each other over the circumference. The experimental data show that, contrary to the assumption based on the physical properties of electrode growth due to lithium intercalation in the graphite, the cell does not expand uniformly. Depending on the position and orientation of the cell coil, there are different zones of expansion and contraction. In order to confirm the non-uniform expansion around the circumference of the cell in 3D, X-ray computed tomography (CT) scans of the cells are performed at low and at high SoC. Comparison of the high resolution 3D reconstructed volumes clearly visualizes a sinusoidal pattern for non-uniform expansion. From the 3D volume, it can be confirmed that the thickness variation does not vary significantly over the height of the battery cell due to the observed mechanisms. However, a slight decrease in the volume change towards the poles of the battery cells due to the higher stiffness can be monitored.
Adding steel fibers to concrete improves the post-cracking tensile strength of the composite material due to fibers bridging the cracks. The residual performance of the material is influenced by fiber type, content and orientation with respect to the crack plane. The latter is a main issue in fiber-reinforced concrete elements, since it significantly influences the structural behavior. The aim of this research is to develop a tailor-made composite material and casting method to orient fibers in order to optimize the performance of the material for structural applications. To this aim, a mechanized concreting device that induces such preferred fiber orientation is designed and fabricated. It uses vibration and a series of narrow channels to guide and orient fibers. A composite with oriented fibers is produced using a hybrid system of macro and micro fibers and high-performance concrete. From the same concrete batch, specimens are cast both with and without the fiber orientation device, obtaining different levels of fiber orientation. Three-point bending tests are performed to measure and compare the residual tensile strength capacities with standard specimens cast according to EN 14651. Elements with favorable fiber orientation show a significant increase in residual tensile strength with respect to the standard beams. Finally, computed tomography and an electromagnetic induction method are employed to better assess the orientation and distribution of fibers in the beams. Their results are in good agreement and enable to link the residual tensile strength parameters with fiber orientation.
The description and prediction of the highly transient processes on the molecular scale in gas-liquid Taylor flows in microchannels are a great challenge. Due to a lack of in situ measurement techniques with satisfactory spatial and temporal resolution, existing models and numerical simulations of reactive two-phase flows are poorly developed. The Raman spectroscopic and holographic system presented herein overcomes shortcomings regarding the high temporal resolution and concomitant difficulties with the high energy densities that must be coupled into microstructured devices to give good limits of detection. It gives insight into the concentration profiles in the liquid phase within the microchannel to study the interaction between hydrodynamics, mass transport, and reaction kinetics.
Die industrielle Computertomographie (CT) ermöglicht den Blick „hinter die Kulissen“ von erhärteten Betonbauteilen und liefert hochpräzise Messdaten aus dem Bauteilinneren. Unter der Voraussetzung der Verfügbarkeit eines für das Bauteil geeigneten Computertomographen kann so eine Vielzahl von Informationen erfasst werden, die zuvor meist nur durch eine Bauteilöffnung/‐zerstörung zugänglich war. Diese Informationen, wie z. B. die genaue Bewehrungslage oder die lokale Häufung von Lufteinschlüssen im Beton, sind relevant für die Qualität von Betonfertigteilen und können in einem Qualitätssicherungssystem im Betonfertigteilwerk gewinnbringend eingesetzt werden. Besonders im Rahmen der Automatisierung und Digitalisierung der Betonfertigteilherstellung birgt eine in den Fließfertigungsprozess integrierte CT‐Qualitätsüberwachung ein großes Potenzial. Auf diese Weise können beispielsweise Segmentfertigteilbrücken hergestellt werden, bei denen die einzeln qualitätsgeprüften Fertigteilsegmente später auf der Baustelle durch Vorspannung und mithilfe von verzahnten Kontaktfugen zu Brücken verbunden werden. Mit den durch CT erfassten Informationen ist es möglich, schwankende, bislang durch die Teilsicherheitsbeiwerte abgedeckte Parameter genau zu analysieren und somit das Sicherheitskonzept für Beton neu zu denken. Durch die Kontrolle von streuenden Parametern und die darauffolgende Reduktion von sicherheitsbedingtem Materialverbrauch kann ein wertvoller Beitrag zum Klimaschutz und zur Schonung von natürlichen Ressourcen geleistet werden.
Computational fluid dynamic simulations involve large state data, leading to performance degradation due to data transfer times, while requiring large disk space. To alleviate the situation, an adaptive lossy compression algorithm has been developed, which is based on regions of interest. This algorithm uses prediction-based compression and exploits the temporal coherence between subsequent simulation frames. The difference between the actual value and the predicted value is adaptively quantized and encoded. The adaptation is in line with user requirements, that consist of the acceptable inaccuracy, the regions of interest and the required compression throughput. The data compression algorithm was evaluated with simulation data obtained by the discontinuous Galerkin spectral element method. We analyzed the performance, compression ratio and inaccuracy introduced by the lossy compression algorithm. The post processing analysis shows high compression ratios, with reasonable quantization errors.
Signal integrity (SI) analysis based on state-of-theart measurements can be difficult to perform especially when the structures of interest are on inner layers of multilayer boards or are enclosed by IC packages. To enable an SI analysis in such cases the authors have recently developed a method that is based on the extraction of accurate simulation models from computed tomography (CT) scans. These models can be used in electromagnetic (EM) field simulators for computer-aided SI analyses. Such CT-based models include geometry variations or defects due to the manufacturing process so that computed EM field simulation results have a good correlation with common measurements. In order to identify the potential of the method an analysis of the required voxel resolution for the extraction of single-ended and differential striplines is presented. The analysis is based on the measurement uncertainty of length measurements in CT scans and an analysis of the propagation of uncertainty for the characteristic impedances of single-ended and differential striplines. This analysis shows that the voxel resolution of industrial CT scans is well suited for the extraction of accurate simulation models which can be used for an SI analysis. His research interests include computed-tomography-based analysis of the electrical parameters of RF and high-speed digital circuits. He is also interested in EM field simulation as well as compute-intensive sensor systems like computed tomography systems. Mr. Hillebrand received the graduate price in 2014 from the Real-Time Systems Committee of the Gesellschaft für Informatik e.V. of the ITG -VDI/VDE for his master thesis.Steffen Kieß received the Diploma degree in software engineering from the University of Stuttgart, Stuttgart, Germany in 2010, where he is currently working toward the Ph. D. degree.In 2011, he joined the Parallel Systems Department, Institute of Parallel and Distributed Systems, University of Stuttgart. Topic of his diploma thesis was the parallelization of the Discrete Dipole Approximation for electromagnetic field simulations. After his diploma thesis, he focused on the finite-difference time-domain methods on hardware accelerators. His current research interests include the extraction of models for electromagnetic field simulations based on computed tomography, parallel electromagnetic field simulations and parallel systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.