The open source CFD library OpenFOAM is utilized to implement the ghost fluid method (GFM). The implemented solver GFMFoam is compared against the standard volume-of-fluid (VOF) solver interFoam in OpenFOAM.For both VOF solvers, the scalar transport equation is solved similarly but the discretization of the momentum and continuity equations differ between GFMFoam and interFoam. The two methods are compared in four test cases with mild surface deformation. First, the GFM implementation is validated.Second, the results on a 2d interface shear layer highlight the importance of the viscous term treatment in the GFM. Third, the model scale Wiggley hull results indicate that the GFMFoam is more sensitive to the grid aspect ratio and resolution than interFoam but the viscous resistance is better predicted. Also, GFMFoam is shown to be less sensitive to numerical ventilation than interFoam. Fourth, the results from the Hamburg Test Case indicate a total resistance prediction within 1% when evaluated against reference data at model scale and 10% at full scale. Based on the steady state test cases of the present study, GFMFoam has moderate benefits over interFoam but it does not clearly outperform interFoam.

This thesis consists of an overview of the following publications which are referred to in the text by their Roman numerals.

No abstract