Direct flow-pattern identification using electrical capacitance tomography

Non-invasive techniques such as process tomography are beginning to make promising contributions to control systems and are well fitted for flow pattern identification in opaque pipes or conduits. Even though process tomography and electrical capacitance tomography are usually associated with imaging, the image in itself can very seldom be directly used for control purpose, The path to be followed from data collection to flow-pattern identification often involves an image reconstruction phase, followed by some form of image processing and analysis. When online control is concerned, this approach may be too time consuming. This paper proposes a direct approach that discriminates between an annular and stratified flow pattern without the need for imaging. The model relies on capacitance data collected from an experimental rig, and extracts the information content from these raw-data measurements. Flow-pattern identifiers are kept simple and are designed to be rapidly evaluated for online observation purpose. Additional information and monitoring, such as the phase distribution or the detection of a slug building process, can also be obtained from the data by using specific electrode pair combinations. (C) 2002 Elsevier Science Inc. All rights reserved.