Computational and experimental analysis of gas/liquid two-phase flow in rod bundles with mixing-vane spacer grids

• Published in: Nuclear engineering and design Vol. 360; no. C
• Main Authors: Waite, Brian M., Prasser, Horst-Michael, Podowski, Michael Z.
• Format: Journal Article
• Language: English
• Published: United States Elsevier 14.01.2020
• Subjects: Experimental model validation; Gas/liquid flow; Multi-field model; NUCLEAR PHYSICS AND RADIATION PHYSICS; Nuclear Science & Technology; Rod bundle with spacer grids; Void fraction measurements with wire mesh sensor
• ISSN: 0029-5493; 1872-759X

The existing evidence clearly shows that the physical phenomena governing gas/liquid two-phase flows are quite complicated even in the case of smooth conduits and simple geometries. Needless to say, the development of experimental, analytical and computational methods for predicting such flows in complex geometries is an even more complicated and challenging task. A configuration of interest to a broad range of industrial applications, including nuclear reactors, deals with the flow of two-phase mixture along the channels formed between narrow arrays of multiple parallel cylindrical elements (tubes or rods). The alignment of such elements is normally accomplished by installing spacer grids placed at regular distances along the flow. The presence of spacers actually affect flow conditions, including the velocity field, pressure drop, heat transfer and, in the case of two-phase flows, phase distribution. The objective of this article is to present a comparative analysis of the results of a combined experimental, theoretical, and computational study of phase distribution around and downstream from complex-geometry spacer grids with split-vane type mixing devices. The main emphasis has been given to the analysis of the effect of proper interpretation of the experimental data on the modeling consistency. The importance of the understanding of uncertainties and limitations associated with the results of multidimensional computer simulations performed using mechanistic modeling principles based on an average bubble size is also discussed.