Predicting Extreme Parametric Roll in Container Ships

Abstract

Parametric roll resonance is an unstable phenomenon that can lead a ship into dangerous situations such as possibly capsizing or loss of cargo. Container ships are particularly susceptible to parametric roll due to the time variance of the metacentric height. Recent incidents have reinvigorated the conversation concerning parametric roll which in turn has caused many to wonder what types of early stage design tools could be developed to provide parametric roll design guidance. This paper presents several predictive modeling techniques that were used to investigate if parametric roll leading indicators can possibly be identified and then considered in early stage design to avoid parametric roll resonance. This paper will utilize four predictive models that utilize and are trained by the parametric roll data created through the execution of a nonlinear non-homogenous damped Mathieu equation, subject to stochastic stiffness and forcing. The ultimate goal of the execution of these techniques is to expand the understanding the potential for parametric roll in early stage design.