Gas–liquid slug flow in a horizontal concentric annulus, a comparison of numerical simulations and experimental data

•Pressure gradient and sluq frequency in a horizontal concentric annulus was studied.•Slug frequency matches experiments to within 9% for the 3 and 7 m domains.•The best match for pressure gradient results were achieved in the 3 m domain.•Domain length should be near an integer multiple of the slug...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:The International journal of heat and fluid flow 2019-08, Vol.78, p.108437, Article 108437
Hauptverfasser: Friedemann, C., Mortensen, M., Nossen, J.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•Pressure gradient and sluq frequency in a horizontal concentric annulus was studied.•Slug frequency matches experiments to within 9% for the 3 and 7 m domains.•The best match for pressure gradient results were achieved in the 3 m domain.•Domain length should be near an integer multiple of the slug to slug distance. Multiphase flow simulations were run in OpenFOAM utilizing InterFoam, a volume of fluid type solver. A concentric annulus geometry was constructed and periodic boundary conditions were applied to alleviate the requirement for a longer domain. The simulations were run in 3, 5 and 7 m long domains with the annulus dimensioned such that the outer and inner cylinder diameter were 0.1 and 0.05 m respectively. The 4 individual mesh designs were constructed such that the coarsest mesh consists of 102k cells/m while the finest mesh was generated with 400k cells/m. Each mesh was significantly refined within 0.005 m of both the inner and outer cylinder wall. The simulation data was compared with experimental pressure and holdup data collected at Institute for Energy Technology in Norway. The 3 and 7 m domains reproduced slug frequencies to within 9% of the experiment results of 1.43 Hz for all mesh densities. Comparatively, the 5 m domain has larger errors with respect to slugging frequency (22–27%). The 5 m case performs poorly, probably due to an artificial restraint introduced by the limitation of available liquid which is set as αoil=0.53 for all cases. The αoil restriction combined with the domain length determines the amount of liquid in the system. This interaction of factors means that the domain length is an important parameter when preparing the simulation. The pressure data display a stronger dependence on the mesh quality in comparison to the slug frequency analysis. The 3 m domain with a 400k cells/m mesh resulted in a maximum and minimum pressure gradient of 1783.5 and 803.9 Pa/m, compared to the experiment values of 1785 and 822 Pa/m, which are within 3% of the expected results.
ISSN:0142-727X
1879-2278
DOI:10.1016/j.ijheatfluidflow.2019.108437