Simulation of gas mixing and transport in a multi-compartment geometry using the GOTHIC containment code and relatively coarse meshes

• Published in: Nuclear engineering and design Vol. 240; no. 6; pp. 1506 - 1527
• Main Authors: Andreani, Michele, Paladino, Domenico
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.06.2010; Elsevier
• Subjects: Applied sciences; Computational fluid dynamics; Computer simulation; Containment; Controled nuclear fusion plants; Energy; Energy. Thermal use of fuels; Exact sciences and technology; Finite element method; Fission nuclear power plants; Fuels; Installations for energy generation and conversion: thermal and electrical energy; Mathematical models; Nuclear fuels; Nuclear reactor components; Vessels; Nuclear fusion reactor; Computational fluid dynamics; Superheated steam; Plume; Helium; Water vapor; Nuclear reactor
• ISSN: 0029-5493; 1872-759X
• DOI: 10.1016/j.nucengdes.2010.02.020

The recently concluded OECD SETH project included twenty-four experiments on basic flows and gas transport and mixing driven by jets and plumes in two, large, connected vessels of the PANDA facility. The experiments featured injection of saturated or superheated steam, or a mixture of steam and helium in one vessel and venting from the same vessel or from the connected one. These tests have been especially designed for providing an extensive data base for the assessment of three-dimensional codes, including CFD codes. In particular, one of the goals of the analytical activities associated with the experiments was to evaluate the detail of the model (mesh) necessary for capturing the various phenomena. This work reports an overview of the results obtained for these experimental data using the advanced containment code GOTHIC and relatively coarse meshes, which are coarser than the ones typically used for the simulation with commercial CFD codes, but are still representative of the models which are currently affordable for a full containment analysis. In general, the phenomena were correctly represented in the simulations with GOTHIC, and the agreement of the results with the data was in most cases pretty good, in some cases excellent. Only for a few tests (or particular phenomena occurring in some tests) the simulations showed noticeable discrepancies with the experimental data, which could be referred to either an insufficiently detailed mesh or to lack of specialized models for local effects.