394
12. Special Topics
The inlet, outlet, lateral, and bottom boundaries are placed about 1.5
model lengths away from the model. Due to the symmetry of the flow, only
half of the model was considered. The grid resolution is high a t the model
surface and in the proximity of the waterline to better capture the boundary
layer and free surface distortion. It expands continuously in all directions
away from the hull and the waterline. The simple geometry of the model
makes grid generation relatively easy. The coarsest grid is a matching blockstructured grid, in which some cells are prisms; the finer grids were obtained
by locally refining the base grid in the vicinity of the model and free surface.
At the inlet, the velocities of both water and air were set to the hull velocity, and the turbulence parameters were derived by assuming a turbulence
intensity of 1% and a turbulent viscosity equal to the molecular viscosity. The
top, bottom and lateral boundaries were treated as slip walls. At the symmetry plane, the symmetry boundary condition was enforced. At the outlet,
extrapolation in streamwise direction was used, and the hydrodynamic pressure was specified according to a prescribed water level. The standard k-6
eddy-viscosity model was used.
The presence of wave breaking required that the simulation allow unsteady flow; over 12 periods of breaking were computed. Figure 12.12 shows
time-averaged velocity vectors in the symmetry plane. The still waterline and
average free-surface shape are also shown. Some back-flow is observed in the
stern region; here, water flows upwards while air flows downwards, with free
surface as a dividing streamline.
Fig. 12.12. Computed average velocity vectors, the undisturbed waterline, and
the average free-surface profile in the symmetry plane (medium grid)
Figure 12.13 shows instantaneous experimental photographs of breaking
waves around the model taken a t the Ship Research Institute in Tokyo. They
are compared with the instantaneous free surface obtained on the finest grid.
12. Special Topics
The inlet, outlet, lateral, and bottom boundaries are placed about 1.5
model lengths away from the model. Due to the symmetry of the flow, only
half of the model was considered. The grid resolution is high a t the model
surface and in the proximity of the waterline to better capture the boundary
layer and free surface distortion. It expands continuously in all directions
away from the hull and the waterline. The simple geometry of the model
makes grid generation relatively easy. The coarsest grid is a matching blockstructured grid, in which some cells are prisms; the finer grids were obtained
by locally refining the base grid in the vicinity of the model and free surface.
At the inlet, the velocities of both water and air were set to the hull velocity, and the turbulence parameters were derived by assuming a turbulence
intensity of 1% and a turbulent viscosity equal to the molecular viscosity. The
top, bottom and lateral boundaries were treated as slip walls. At the symmetry plane, the symmetry boundary condition was enforced. At the outlet,
extrapolation in streamwise direction was used, and the hydrodynamic pressure was specified according to a prescribed water level. The standard k-6
eddy-viscosity model was used.
The presence of wave breaking required that the simulation allow unsteady flow; over 12 periods of breaking were computed. Figure 12.12 shows
time-averaged velocity vectors in the symmetry plane. The still waterline and
average free-surface shape are also shown. Some back-flow is observed in the
stern region; here, water flows upwards while air flows downwards, with free
surface as a dividing streamline.
Fig. 12.12. Computed average velocity vectors, the undisturbed waterline, and
the average free-surface profile in the symmetry plane (medium grid)
Figure 12.13 shows instantaneous experimental photographs of breaking
waves around the model taken a t the Ship Research Institute in Tokyo. They
are compared with the instantaneous free surface obtained on the finest grid.