HPC-Based Aerodynamics Surrogate Model For Wind Assisted Ship Propulsion Performance Predictions

The AEROWASP project aims to develop a highly accurate CFD-based aerodynamic surrogate model capable of predicting the performance of wind-assisted ships, including the flow perturbations caused by the vessel’s hull and superstructures. The key challenge is to create a model that is both reliable and cost-effective, enabling its use throughout all stages of a wind-assist project, from initial feasibility studies to post-installation performance assessment. By providing more accurate performance predictions, the solution will help shipowners and charterers select, evaluate, and operate wind propulsion technologies with greater confidence, leading to improved vessel performance, reduced fuel consumption, and shorter payback periods.

The aerodynamic surrogate model will be applicable to a wide range of vessel and wind propulsion system types. It will be trained using an extensive HPC-generated dataset of CFD simulations. A proprietary Active Machine Learning approach will be employed to train a Deep Neural Network, ensuring a high level of predictive accuracy while significantly reducing the need for computationally expensive simulations during operational use.

Organisations involved: 

End-User: Blue Wasp Marine

Technology expert: Bailardi Engineering