Computational Framework for Nonlinear Seakeeping Prediction under Hydrodynamic Uncertainty

Seakeeping plays a significant role in the areas of ship design, operability, and safety assessment in dynamic open-sea conditions. One of the primary challenges of computational marine hydrodynamics is the accurate prediction of non-linear seakeep-ing behaviour under irregular wave excitation and operational uncertainty. Existing works use threshold-based criteria for roll,acceleration, slamming and deck wetness which are computed independently and under model-specific assumptions, disregarding computational trade-offs, uncertainty and real-time operational requirements. This study proposes a unified benchmarking and uncertainty-aware framework for non-linear seakeeping prediction that integrates hydrodynamic modelling, proba-bilistic benchmarking, and digital twins. A generic mathematical representation is used to define seakeeping criterion in terms of six-degree-of-freedom vessel dynamics and wave-structure interactions. The paper proposes a composite hydrodynamic risk function that represents the joint occurance of slamming loads,green water events, excessive roll motion, and acceleration breach. Modeling uncertainty arising from wave-spectrum variability, sensor noise, and parameter uncertainty is addressed using probabilistic uncertainty propagation and fine-tuning using data-driven learning. The integration of physics-based hydrodynamic solvers with real-time data assimilation enables dynamic updating of operational safety parameters and non-linear performance prediction.The proposed framework bridges the gap between existing seakeeping theories by leveraging data-driven digital twin design, which serves as a foundation for model selection and future research in computational marine hydro-dynamics.