Real time three-dimensional electrical impedance tomography applied in multiphase flow imaging

In many industrial applications the aim is to obtain information on three-dimensional (3D) material distribution within the process vessels. With standard two-dimensional (2D) techniques only vague cross-sectional information can be obtained. It could be possible to carry out several 2D reconstructi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Measurement science & technology 2006-08, Vol.17 (8), p.2083-2087
Hauptverfasser: Heikkinen, L M, Kourunen, J, Savolainen, T, Vauhkonen, P J, Kaipio, J P, Vauhkonen, M
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In many industrial applications the aim is to obtain information on three-dimensional (3D) material distribution within the process vessels. With standard two-dimensional (2D) techniques only vague cross-sectional information can be obtained. It could be possible to carry out several 2D reconstructions on different layers and in this way to obtain 3D information. However, in this approach errors are induced since no real 3D information is utilized in the image reconstruction. In this paper we describe an approach to measure, reconstruct and visualize three-dimensional electrical impedance tomography images in real time. The reconstruction is based on a difference imaging scheme. An efficient current injection and voltage measurement protocol is used in order to increase the sensitivity and reduce the data collection time. The proposed approach can produce and visualize up to 15 3D EIT images per second when 80 measurement electrodes are used. Imaging results from a stirred vessel and a flow loop will be shown. The reconstructions show, for example, that 3D air/liquid distribution in the stirred vessel can reliably be visualized in real time and material flow can be monitored in a 3D section of the flow loop. Reconstructions can be visualized and analysed in many different ways in order to produce essential information on the behaviour of the processes.
ISSN:0957-0233
1361-6501
DOI:10.1088/0957-0233/17/8/005