Numerical investigation of flow characteristics through simple support grids in a 1 x 3 rod bundle

• Published in: Nuclear engineering and technology Vol. 51; no. 8; pp. 1905 - 1915
• Main Authors: Karaman, Umut, Kocar, Cemil, Rau, Adam, Kim, Seungjin
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2019; Elsevier; 한국원자력학회
• Subjects: 1x3 rod bundle; CFD; Hybrid mesh; Reynolds stress model; Simple support grid; Turbulence; 원자력공학; CFD; Simple support grid; Turbulence; 1x3 rod bundle; Hybrid mesh; Reynolds stress model
• DOI: 10.1016/j.net.2019.06.020
• ISSN: 1738-5733

This paper investigated the influence of simple support girds on flow, irrespective of having mixing vanes, in a 1 x 3 array rod bundle by using CFD methodology and the most accurate turbulence model which could reflect the actual physics of the flow was determined. In this context, a CFD model was created simulating the experimental studies on a single-phase flow [1] and the results were compared with the experimental data. In the first part of the study, influence of mesh was examined. Tetra, hybrid and poly type meshes were analyzed and convergence study was carried out on each in order to determine the most appropriate type and density. k−ε Standard and RSM LPS turbulence models were used in this section. In the second part of the study, the most appropriate turbulence model that could reflect the physics of the actual flow was investigated. RANS based turbulence models were examined using the mesh that was determined in the first part. Velocity and turbulence intensity results obtained on the upstream and downstream of the spacer grid at –3dh, +3dh and +40dh locations were compared with the experimental data. In the last section of the study, the behavior of flow through the spacer grid was examined and its prominent aspects were highlighted on the most appropriate turbulence model determined in the second part. Results of the study revealed the importance of mesh type. Hybrid mesh having the largest number of structured elements performed remarkably better than the other two on results. While comparisons of numerical and experimental results showed an overall agreement within all turbulence models, RSM LPS presented better results than the others. Lastly, physical appearance of the flow through spacer grids revealed that springs has more influence on flow than dimples and induces transient flow behaviors. As a result, flow through a simple support grid was examined and the most appropriate turbulence model reflecting the actual physics of the flow was determined.