Numerical Uncertainty Analysis in Regular Wave Modeling

In recent decades, the use of computational fluid dynamics (CFD) in many areas of engineering as a research and development tool has seen remarkable growth. Recently, an increasing concern with the assessment of the quality of CFD results has been observed. Wave modeling is an important task in many ocean engineering applications. Although numerical modeling studies of waves can be found in the literature for many applications, it is hard to find studies that present the numerical uncertainties of the results. In this study, the numerical uncertainties in mean wave parameters simulated using a viscous model were estimated using a procedure based on grid/time refinement studies and power series expansions. STARCCM+ software was used to simulate wave propagation. The computational domain was discretized using a trimmer mesh. The results obtained for a regular wave with a wave steepness (H/L) equal to 0.025 are presented. The numerical uncertainties in mean wave height and mean wave period were estimated along the computational domain. The results indicate that the convergence properties of the mean wave parameters with the grid refinement depended on both position in the domain and the selected wave parameter.