This thesis investigates local and global scour processes around four-legged jacket (4LJ) foundations
for offshore wind turbines under current-only conditions using the 3D CFD Large Eddy Simulation (LES) model TUDflow3D. Jacket foundations, while increasingly deployed in deeper offshore environments, remain less studied than monopiles on local scour and scale effects and introduce new scour patterns like global scour. The research addresses this gap by evaluating the predictive performance of TUDflow3D against laboratory experiments, extending simulations to field scale, and assessing the applicability of monopile-based empirical scour relations to jacket configurations.

Laboratory-scale numerical simulations replicate the experimental study of Welzel et al. (2023) for both clear-water and live-bed regimes, validating the model for a 4LJ foundation. Model performance is assessed through morphodynamic comparison and time-averaged observation of the hydrodynamics, supplemented by a sensitivity analysis on numerical parameters: morphological acceleration factor, relaxation factors, numerical domain and grid resolution.

The validated model is then scaled to field conditions using a mobility similarity approach, enabling
investigation of scale effects on scour magnitude, spatial extent, and equilibrium timescales.
After the scaling up, both models are analyzed in order to enhance the knowledge that is known until now from 4LJ foundations.

For local scour, the spatial extent of scour was determined by adapting a formulation for monopiles and quantified throughout the entire scour evolution, revealing a consistent footprint over time. Scaling to field conditions demonstrated that the upstream–downstream scour pattern persists, but with reduced magnitudes in local scour depth due to scale effects, observed by the reduction of time-averaged bed shear stresses around the legs. This reduction in magnitude was approximately 30%–35% when compared in terms of dimensionless scour for both regimes and for both upstream and downstream piles. Timescale analysis showed that scour equilibrium occurs later at locations farther from the piles. When scaling, it was also found that scour on clear-water regimes
require months while live-bed regimes is in the order of magnitude of days to reach local scour equilibrium.

For global scour, live-bed conditions produced a footprint extending up to twice the jacket footprint
radius with rapid initial development, while clear-water conditions yielded slower, more confined scour. The order of magnitude for the global scour is variable depending on the location, but it reached values around 0.9 - 1 D on the center of the jacket and downstream of it.

Finally, empirical monopile scour formulas were evaluated, showing that they still provide reasonable order-of-magnitude estimates for upstream piles and capturing scale effects, as well it was proved that the scaling up from laboratory scale to field scale developed for monopiles can be useful also for 4LJ foundations.

The study demonstrates that CFD LES modeling is a robust and flexible tool for detailed, process based scour prediction around complex foundation geometries. Beyond validation, its ability to simulate field-scale morphodynamics and provide high-resolution temporal and spatial data makes it a valuable complement to laboratory testing,

The Port of Rotterdam experiences large amounts of siltation from the upstream rivers in its basins and canals every year. Furthermore, container vessels increase in size every decade and therefore require a larger navigational depth. These two factors cause the yearly amount of dredged material to increase. Currently, mainly trailing suction hopper dredgers are used to remove the large amounts of sediment, but these vessels come with high costs and operational time. A more time and cost-efficient dredging method is therefore desired.

A potential solution to the increasing problems of costs and operational time by dredgers is the technique of water injection dredging. To make this technique as efficient and effective as possible, this thesis has the objective to measure and analyze flow and sediment properties for different parameter settings of water injection dredging and find the optimal parameter settings of this dredging technique on mud from the Port of Rotterdam. Large-scale experiments have been conducted in the water-soil flume of Deltares, where a jetbar was trailed (or run) over a bed 27 meters long and 0.5 meters deep of port mud while injecting it multiple times with water.
A positive correlation is observed between the production rate and jet momentum. The production is related linearly with the jet momentum using the Vlasblom equation combined with a non-dimensionless empirical fitting parameter per run per traverse velocity.
What is remarkable is that the intrusion depth by the jets increases with ascending runs while the jet parameter settings stay the same. The data shows that the difference between the mass flux by the density current and the mass flux stirred up by the jets when the sediment concentration of an undisturbed bed is assumed, increases between runs when an increased intrusion depth between those runs is observed as well. This indicates that the volume penetrated by the jets contains a smaller amount of sediment than was initially assumed, thus is disturbed by the previous run and is therefore decreased in strength. This decrease in strength results in a larger intrusion depth during the run itself in comparison to the previous run.

The influence of the SOD of the jets is analysed by comparing runs with a SOD to runs without a SOD but with similar remaining jetting parameter settings. When a SOD is applied, the jet pressure applied to the bed is outside the flow development region and therefore the jet pressure decreases with distance from the jet nozzle. The mass of sediment stirred up by the jets, for a SOD of 300 mm, is lower in comparison to a SOD of 0 mm. The density current transports relatively more sediment, however, when a SOD of 300 mm is applied. So, if a large amount of sediment needs to be stirred up, and therefore a large intrusion depth is required, no SOD should be applied and when large horizontal transport by the density current is desired, a SOD outside the flow development region of the jets should be used.