Non-standard stochastic averaging of large-amplitude ship rolling in random seas

Abstract

The current intact stability criteria for ships do not account for their dynamic behaviour in irregular waves sufficiently. We provide analytical criteria for ship and sea state parameters that indicate large roll amplitudes or capsizing and determine the mean time for these events. These results can be used to formulate vulnerability criteria for the occurrence of parametric roll. We use reduced order models in order to perform an analytical analysis to this complicated problem. Therefore, we study the dynamics of a ship in random seas as a perturbation of a Hamiltonian system. The random process across the energy levels of the Hamiltonian system is approximated by a Markov process, which is obtained by stochastic averaging. The noise owing to wave excitation is a non-white stationary stochastic process. This is an extension to previous results, where white noise excitation was used.