Development of a Reduced-Order Model for Hydrodynamic Analysis of a Wind-Assisted Ship

Artikel i vetenskaplig tidskrift, 2026

Decarbonization is one of the most pressing challenges facing the modern shipping industry. Wind-assisted ship propulsion (WASP) offers a promising pathway by integrating wind technologies with conventional systems, but understanding the hydrodynamic behaviour of such vessels under realistic conditions is essential for design optimisation. This study investigates the nominal wake field of a bare hull across various leeway angles and Froude numbers, along with pressure and wall shear stress distributions of a self-propelled ship under different leeway angles and reduced propeller thrust conditions. Computational Fluid Dynamics (CFD) simulations are performed for leeway angles from −9° to 9° and sail thrust values from 0% to 50% of ship resistance. A virtual disk model represents propeller effects under self-propulsion conditions. The CFD results are used to develop a ReducedOrder Model (ROM) combining Weighted Proper Orthogonal Decomposition (WPOD) and the Kriging method. This hybrid framework enables efficient flow field reconstruction across varying leeway angles, Froude numbers, and sail thrust values. Once trained, the ROM predicts wake fields, pressure, and shear stress within seconds, providing a fast alternative to high-fidelity CFD. Results show the ROM effectively captures the target field quantities while significantly reducing computational cost. The WPOD models achieve high accuracy, with average errors of 2.41% for mean wake fraction,
3.5% for resistance, and 3% for side force compared to reference CFD results