Numerical Simulation of Boundary Layer Flow Around Simplified Subsea Structures

Abstract

Numerical simulations of boundary layer flow over simplified subsea structures on a flat seabed have been performed. A two-dimensional RANS method with a standard high Reynolds number k-ε turbulence model have been applied.

Resulting values of CD are in good agreement with published physical experiments for δ/h=0.73, δ/h=1.70 and δ/h=2.55. Different values of width to height ratio for δ/h=1.70 also prove similar results between numerical simulation and experiments. Reattachment is present when b/h=5 while not present for b/h=3, as observed in published experiments. Local velocity properties for square cross section with δ/h=0.73 from present study is also validated against published experiment. The resulting trend from the present study seem to agree with the experiment.

Results from numerical simulations show excellent agreement of CD when compared to experiments for δ/h ≥ 1.70, hence further study of various width to height ratio b/h when δ/h=1.70 is performed. The resulting value of CD from the numerical simulations indicated a slight over prediction when δ/h ≤ 0.73.

Transient (Unsteady RANS) and steady-state (RANS) simulations are carried out. The steady-state runs did not have fluctuating values of CD and CL in any simulation, indicating that two-dimensional boundary layer flow over simple structures with sharp corners may be treated as a stationary problem.

Resulting hydrodynamic quantities for simplified geometries of GRP covers are found to be similar to equivalent values obtained from square and rectangular cross sections. It is also found that mudmat can be neglected in further analysis of GRP 1.

The numerical method used is capable of producing physically sound hydrodynamic quantities for simplified subsea structures on the seabed. This is relevant for GRP covers which may move due to hydrodynamic forces from extreme currents and the added weight needed to avoid displacement.