Application of CFD methods to the evaluation of flow characteristics of open thermosyphons
The open thermosyphon has several characteristics which makes it suitable for verification of CFD codes. Prior experimental research has provided data suitable for benchmarking case histories, furthermore, in spite of its simple natural circulation flow geometry, the open thermosyphon exhibits complicated flow behavior. In this study two different CFD codes are employed to generate parameters which can be compared with known experimental values. A quantitative comparison of experimental and calculated results showed that the experimentally measured energy carrying capacity of this strictly axially-symmetric flow geometry could not be reproduced by 2-D computations. Permutations of the computational model showed that this failure was not a function of noding structure, turbulence models, or even of the CFD code employed. Irrespective of the computational options employed, when compared on the basis of Ra{sub m}, (where Ra{sub m} = Ra*R/L), the calculated results remained a factor of five or more below the experimentally measured data. The CFD generated results had in fact a larger disagreement with experimental data then earlier theoretical computations which employed boundary layer integration techniques. The computations were repeated employing 3-D cylindrical geometry while maintaining the same mesh size as for the 2-D computations. The calculated results showed that above values of Log (Ra{sub m}) = 3, the flow field above the thermosyphon tube, does not maintain axial symmetry. The ascending warm stream and the returning cooler flow break the symmetric flow pattern which is imposed in a 2-D computation and flow around and past each other. Mixing takes place at the interfaces of these streams, but for the most part they maintain a separate identity. The computed energy transport rate increases by a factor of two to four, compared to the 2-D calculations. This moves the calculated result closer to experimentally measured values, but is still not entirely satisfactory. The study analyzes the causes of the observed calculated difference between the 2 and 3 dimensional approaches. The evaluated 3 dimensional flow field is used to show that the reasonable theory/experiment agreement achieved by the boundary layer integration method depends on assumptions which have a compensating effect.
- Research Organization:
- Univ. of Maryland, College Park, MD (US)
- OSTI ID:
- 20014428
- Resource Relation:
- Conference: 32nd National Heat Transfer Conference, Baltimore, MD (US), 08/08/1997--08/12/1997; Other Information: PBD: 1997; Related Information: In: ASME proceedings of the 32nd national heat transfer conference (HTD-Vol. 349). Volume 11: Interfacial thermal phenomena in thin films; Heat pipes and thermosyphons; Heat and mass transfer in porous media, by Goodson, K.; Chang, W.S.; Charmchi, M.; Hadim, H. [eds.], 211 pages.
- Country of Publication:
- United States
- Language:
- English
Similar Records
Development of High-Fidelity Numerical Models for Supercritical CO2 Oxy-Combustion: SBIR Phase I Final Technical Report
Particle Image Velocimetry Measurements in a Wire-Wrapped 61-Pin Hexagonal Fuel Bundle