A new collaborative optimisation framework combining CFD and neural networks has achieved up to 42.87 per cent aerodynamic drag reduction on containership bow fairings, potentially saving 2-4 per cent fuel annually on 20,000 TEU vessels, reports ScienceDirect.

Developed by researchers in China, the system integrates Optimal Latin Hypercube Sampling, high-fidelity CFD simulations, Elliptical Basis Function neural networks and Covariance Matrix Adaptation Evolution Strategy to optimise three bow fairing geometries.

The polygonal arcuate fairing showed the best performance, reducing drag by 42.87 per cent under 8 Beaufort headwinds. All designs reshaped airflow, suppressed vortex formation and streamlined pressure distribution over stacked containers.

The framework was validated against the KCS benchmark model with simulation error under 1.62 per cent. It offers a scalable solution for aerodynamic resistance, which can account for up to 10 per cent of total drag on ultra-large container ships.

The study highlights aerodynamic optimisation as a practical route to meet IMO decarbonisation targets. Unlike cargo layout changes, fairings do not interfere with container operations and are adaptable to other marine structures.

The ME-GI engine's low methane emissions and compatibility with bio- and synthetic LNG were cited as key enablers for net-zero shipping. Everllence reported over 1,000 cumulative sales of the engine, with 400 orders from the container segment.