Simulation of the Flow around a Semi-Submersible using CFD

Typ
Examensarbete för masterexamen
Master Thesis
Program
Publicerad
2017
Författare
Eriksson, Jonathan
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
The application of Computational Fluid Dynamics (CFD) is almost taken for granted within sectors of engineering dealing with fluid flows. However, its usage within the offshore sector is still limited, primarily due to the large computational times necessary for solving complete sea states. But with increasing computational power and creative new solution methods emerging, the threshold of usability is continuously being lowered. Today the industry relies heavily on scaled model tests; tests which are expensive, time consuming, and may face discrepancies due to scaling effects and unreliable measuring equipment. By utilizing CFD instead, one can analyze the physical properties anywhere within the computational domain, as well as customize and adjust the geometrical and numerical settings at any time. The main objective of this thesis is to investigate the use of CFD simulations for the design of semi-submersible offshore structures. In this study the submerged section of an offshore structure is the subject of examination. Data available from a previously conducted wind tunnel experiment are used as means of both verification and calibration of the CFD model. Two different CFD software are used in this work; the open-source software OpenFOAM, as well as the commercial software Fluent. Initially a replica of the wind tunnel is modelled and incompressible steady-state Reynolds Averaged Navier-Stokes (RANS) simulations are performed, as it is of interest to find possible instances where steady-state simulations are feasible. Investigations include full geometry and truncated nearfield steady-state simulations. Non-satisfactory results were obtained. The drag force measured in the steady-state simulations reached only a value of 60% of the experimentally measured one. Unsteady simulations, both RANS and LES showed similar results. Reasons for the unsatisfactory results could be insufficient mesh resolution, along with wrong choice of discretization methods and turbulence model.
Beskrivning
Ämne/nyckelord
Transport, Farkostteknik, Transport, Vehicle Engineering
Citation
Arkitekt (konstruktör)
Geografisk plats
Byggnad (typ)
Byggår
Modelltyp
Skala
Teknik / material