Introduction
The objective of this Computational Fluid Dynamics (CFD) study was to predict the impact of the Vessel turning manoeuvre in Newhaven Harbour on the water flow in the adjacent marina.
This work follows on from a previous study, which focussed on the prediction of waves generated by a vessel travelling through the marina. That study concluded that any waves produced by a similar or smaller vessel in the marina would be relatively small. However, the study also highlighted various regions of recirculation in the marina with velocity magnitudes up to around 0.2 m/s. The study concluded that any future work should not try to predict the formation and movement of waves by using a computationally intensive two-phase model. Instead, the air-water interface should be defined at a constant elevation throughout the model and hence steer computational resources to giving a more accurate prediction of the flow profile through the marina. This was the approach adopted during this study.
As well as modelling the Vessel at the three critical angles identified in the previous study (44, 60 and 90), the flow of water around the another vessel was remodeled using the same approach to allow direct comparison between the two vessels. Also, a run without a vessel in the channel was executed to allow the impact of a turning vessel on the marina to be evaluated.
Scope
The supplied drawings were used to create a 3D geometry model of the Vessels. This was used, together with the existing channel and marina geometry model, to construct a new CFD model of the water in the vicinity of the Shorter Vessel and the Newhaven Harbour marina.
The size and complexity of the geometry meant it was not feasible to model the realtime turning of the vessel. Instead, as before, three individual simulations were performed with the vessel at 44, 60 and 90 degree angles to the East Quay.
The supplied on-site flow measurement data was used to define the simulated inlet flow and water level under what was previously considered the worst case condition with respect to the marina (2 hours before Low Water in July).
Also, a model was created without the vessel present so that a detailed assessment of the impact the turning of a vessel has on the flow of water through the channel and marina could be made.
The latest version of the CFD software suite, FLUENT 6.2 was used to solve and post-process each case.
CFD Modelling of Fluid Flow
Geometry Modelling of the MV Dieppe and Newhaven Harbour
The drawings of the Shorter Vessel were scanned and imported into the 3D CAD package, Rhinoceros 3D. The geometry model of the MV Dieppe hull was constructed in a similar fashion to the channel. Vertical polylines were traced over the vertical cross sections described in the 2D drawing and horizontal polylines constructed along the length of the hull to form a grid. This ‘network of curves’ was used to define a surface that accurately represented the form of the hull.
Figure 1 shows a 3D view of the geometry model of the MV Dieppe, together with the hull surface isocurves visible to aid visualisation of the hull form.
Figure 1 – 3D View of Vessel Geometry Model
The Channel geometry was developed in a similar fashion as shown in Figure 2
Figure 2 – 3D View of Channel and Marina Geometry Model
Three unique geometry models were created by importing the vessel into the channel model the 90, 60 and 44 positions (relative to the East Quay), as in the previous study. The model geometry is shown below in Figure 3 with the vessel in the 90° position.
Figure 3 – 3D View of a Vessel at 90° to the East Quay
The vessel position was fixed for the duration of each run. The three geometry models from the previous study were split along the water level and the upper (air) portion discarded ready for meshing.
Construction of the CFD Mesh
In order to describe the computational domain to the CFD software and hence solve the equations of flow at discrete locations throughout the domain, a unique volume mesh was constructed for each vessel and each vessel position. An additional mesh was created without the vessel present to give a total of seven meshes.
As the simulation involved just one phase (water) the fine mesh around the water level was not required as it was for the VOF (Volume Of Fluid) model. This allowed the other features of the model to be resolved to a higher detail than previously.
The meshing tool Gambit was used to construct an unstructured Cartesian mesh throughout the domain.
This enabled the definition of fine cells around the hull of the vessel, the bed of the marina and the bed of the channel. Each mesh consisted of around five million cells (depending on the vessel being modelled and the position of the vessel).
CFD Modelling of the MV Dieppe and Newhaven Harbour
To predict the movement of fluid through the domain the Navier-Stokes equations were solved with the addition of the Volume of Fluid (VOF) model. The additional volume fraction equation introduced by this model was solved for each cell containing 100% air or water using the Euler Explicit Scheme. The Geometric Reconstruction Scheme was applied to the cells that contained the air-water interface. This scheme was considered the most accurate available in FLUENT for predicting the volume of each phase advected along the interface, although it was also the most computationally intensive.
Figure 4 – Example Isopleth Output Map
The model was treated as transient due to moving waves being predicted in the marina. The equations of flow, described above, were solved by FLUENT’s segregated solver for each time step using the Non Iterative Time Advancement (NITA) approach. The PISO algorithm was used for pressure-velocity coupling. The Body-Force-Weighted pressure interpolation scheme was used to predict the pressure at the cell faces, as required by the momentum equation. The Second-Order Upwind discretisation scheme was used to predict the convective terms of the scalar equations.
Each simulation had to run for a long enough time for the water to travel from the inlet past the marina. After this point, the solution was monitored to ensure that it had reached a cyclic behaviour before being halted.
Conclusions
Conclusions were drawn during the execution of the study relating to the:
- Risk of Grounding
- Water Level Disturbance
- Creation of any vortices |
