Statistical Variation Of The 3-Hour Maximum Crest Height in a Survival Sea State

Abstract Due to the random nature of extreme waves, wave impacts follow a highly stochastic pattern. To reduce the statistical uncertainties that are naturally arising in estimates of design loads related to extreme waves, sufficient data must be gathered. The first step in a design load analysis is the realization of a set of realistic (typically 3-hour) waves which is large enough to describe the randomness in the impact in sufficient detail, such that the probability of exceedance of the maximum 3-hour load levels can be predicted accurately. In this paper we will investigate if Computational Fluid Dynamics (CFD) is capable of predicting realistic random 3-hour extreme waves. Since the maximum 3-hour load is driven by the highest and steepest waves we will look into the distribution of the maximum 3-hour crest height and the corresponding wave steepness of these events. A comparison is made with wave flume measurements in which 100 random realizations of an extreme wave (Hs = 16.7m, Tp = 15.9s) with a 10,000-year return period were generated and measured.