Estimating dynamic stability event probabilities from simulation and wave modeling methods

Abstract

Predicting the dynamic stability of ships in severe wave environments is challenging, not only due to the complex non-linear hydrodynamics, but also the need to characterize the rarity of events. The latter involves conducting enough simulations to calculate associated small probabilities or alternate approaches for estimating the rarity of events. This paper presents techniques for calculating probabilities of occurrence of rare dynamic stability events using direct counting, Poisson distribution fitting techniques and estimating the dynamic event probabilities. The latter probability estimate is obtained by defining dangerous wave conditions that produce rare events through hydrodynamic simulations and estimating their probabilities of occurrence through joint probability distributions or simulations of the wave environment. The accuracy of these calculations is discussed. An example application is presented using a U.S. Coast Guard Cutter along with information useful for operator guidance in heavy weather. A recommendation is presented for further work on defining the limiting probabilities one might use for design or operational criteria.