This study fits within a larger research project that aims to determine the nautical bottom more accurately, when a fluid mud layer is present, by simulating the thixotropic behaviour of the fluid mud using CFD. This thesis tries to model the flow around and the exerted forces on a cylinder towed through a fluid mud layer, without any presence of a water layer, at the current phase of the research project. These towing tests were executed at different drafts and velocities at Flanders Hydraulics Research. The CFD software chosen for this study is OpenFOAM, more specifically the interFoam solver was used. This is an incompressible immiscible two-phase solver that uses a static mesh.
During the implementation of the model, the focus was on maximising flexibility. E.g. the rheologic parameters can easily be changed in the input of the model, as can the geometry of the object that is to be studied. Only one .stl file needs to be replaced and the forces on this new object are calculated automatically.
The drag forces computed during the simulations of the different scenarios (combinations of a towing velocity and a draft) all fit within the interval set by the standard deviation of the experimental results. The computational effort required for the simulation of each of these scenarios differs. At higher velocities or very low drafts the computational effort will increase from 1h23, when the velocity equals 0.1m/s and the draft equals 10cm, over ±3h, when v = 0.3m/s and d = 6cm, to ±9h, when v = 0.5m/s and d = 2cm. This might seem like long simulation times for a two dimensional simulation, as the eventual goal is to simulate the flow around three dimsional objects. However these simulations were performed on a laptop with limited computational power and without the usage of parallel processing.The drag forces resulting from the simulation show to be very depended on the amount of structure (gamma) present, as on the aggregation coefficient a, the break-down coefficient b and their ratio bèta = b/a. When the initial condition of gamma increases, so will the drag force. The adjustment of a and b with a constant ratio bèta will primarily change the rate at which the stucture and therefore the drag force develops. The ratio bèta has a very important impact on the fluid behaviour. When bèta increases, the shear layer will widen and the maximal shear rate in this area will decrease. Finally, if modelling the interaction between a fluid mud and water layer is possible, CFD simulations could predict the forces acting on a ship’s hull. This would facilitate the determination the navigable depth (nautical bottom) for different shapes of hulls and for different maneuvers. |
