Faculté des Sciences appliquées
Faculté des Sciences appliquées

Assessment of Numerical Simulations for High Reynolds Flows Around the Pump Tower of a LNG membrane Tank on a Floating Structure

Sajib, Md Kawsar ULiège
Promotor(s) : Rigo, Philippe ULiège
Academic year : 2022-2023 • Permalink :
Title : Assessment of Numerical Simulations for High Reynolds Flows Around the Pump Tower of a LNG membrane Tank on a Floating Structure
Author : Sajib, Md Kawsar ULiège
Advisor(s) : Rigo, Philippe ULiège
Language : English
Number of pages : 169
Keywords : [fr] hydrodynamic loads
[fr] mask effect
[fr] temporal and spatial resolution
[fr] hybrid RANS-LES (IDDES)
[fr] postcritical
[fr] URANS
[fr] turbulence models
[fr] 3D effects
[fr] GTT
[fr] eddy viscosity-based turbulence model
[fr] Kolmogorov turbulence spectrum
Discipline(s) : Engineering, computing & technology > Multidisciplinary, general & others
Commentary : This research project is funded by Gaztransport Technigaz (GTT). The research has been carried out at LHEEA lab in ECN under the supervision of Dr. Emmanuel Guilmineau
Funders : Gaztransport Technigaz (GTT)
Research unit : METHRIC-Modelling of Incompressible Turbulent Flows at High Reynolds numbers and Couplings at LHEEA
Name of the research project : -
Target public : Researchers
Professionals of domain
Institution(s) : Université de Liège, Liège, Belgique
Degree: Master : ingénieur civil mécanicien, à finalité spécialisée en "Advanced Ship Design"
Faculty: Master thesis of the Faculté des Sciences appliquées


[fr] Gaztransport Technigaz (GTT), a French naval engineering company, focuses on producing cryogenic membrane containment systems for storing and transporting LNG (liquefied natural gas). The Morison equation is used to determine hydrodynamic loads on the tubular Tripod mast structure (TMS). The fact that Morison equation cannot exhibit a mask effect, therefore GTT aims to evaluate the design phase using high-fidelity ISIS-CFD simulation. The previous study utilized a k-$\omega$ SST turbulence model to forecast unsteady dynamics, suggesting a better temporal and spatial resolution could improve unsteady dynamics prediction. The first stage of the present study involved simulations of a single cylinder at different Reynolds numbers, spanning from sub-critical (O($10^{4}$)) to postcritical (O($10^{7}$)) flow regimes, to determine optimal temporal and spatial resolution. Two-dimensional simulations of TMS were performed using the resolutions from the first phase. 3D simulations were conducted using URANS k-$\omega$ SST and hybrid RANS-LES (IDDES) turbulence models, simplifying the 3D case to assess the 3D effect in the spanwise direction. The final phase involved 3D simulation with a free surface using the URANS k-$\omega$ SST turbulence model. The present investigation relating to single cylinders showed satisfactory agreement with the existing numerical studies documented in the literature over the whole spectrum of flow regimes. The results indicated that the eddy viscosity-based turbulence model is unable to accurately predict the coefficient of drag in the sub-critical to critical flow regime seen in the experiment, mostly owing to inherent limitations. Nevertheless, the research exhibited a high level of conformity with the experimental findings from the supercritical flow regime. The two-dimensional investigation of TMS showed promising results in comparison to the preceding 3D study, therefore highlighting the impact of enhanced temporal and spatial resolution. The present study conducted an in-depth examination of the 3D effects in the spanwise direction. This investigation revealed the presence of 3D effect which was absent in the two-dimensional (2D) study. The 3D hybrid RANS-LES demonstrated a successful capture of a wide range of the Kolmogorov turbulence spectrum. The presence of the free surface effect was evident when compared to the simpler 3D case. The mask effect was observed across all situations (involving both 2D and 3D cases) and utilizing various turbulence models. The investigation showed that URANS (k-$\omega$ SST) turbulence model has the capability to accurately represent the unsteady dynamic of the flow, as long as appropriate temporal and spatial resolutions are maintained. The computing cost of 2D flow simulations was lower compared to 3D simulations, however, it could provide accurate predictions of hydrodynamic parameters. The hybrid RANS-LES approach should be evaluated using the identified temporal and spatial resolution for each orientation.



Access MKSAJIB_MS_Thesis1.1.pdf
Size: 53 MB
Format: Adobe PDF


  • Sajib, Md Kawsar ULiège Université de Liège > Master ing. civ. méc. (EMSHIP+)


  • Total number of views 30
  • Total number of downloads 29

All documents available on MatheO are protected by copyright and subject to the usual rules for fair use.
The University of Liège does not guarantee the scientific quality of these students' works or the accuracy of all the information they contain.