Deep Learning Physics for the Hydrodynamics of Trading Vessels

Jonas Verrière, Jocelyn Ahmed Mazari, Antoine Reverberi, Francis Hueber

NAFEMS World Congress 2023, Tampa, Florida, USA


In a context of more and more stringent environmental regulations and of energy savings, ship hull efficiency has become a key design aspect for all trading vessels like tankers or container vessels. To achieve an efficient hull design, the resistance curve of the boat needs to be assessed as early as possible in the initial design to establish the best possible architecture in a very short time before contracting. Such an assessment remains largely made with empirical methods due to the time constraint that prevents an extensive usage of CFD solvers, but those methods remain quite limited in terms of robustness and accuracy. Also it does not provide insight on the 3D physical field like wave patterns or pressure distributions on the hull. In this paper we are presenting how the Deep LearningPhysics (DLP) developed by Extrality can help address this challenge by delivering fast 3D high-fidelity predictions. First a generic DLP-model is built from a dataset of 288 CFD simulations covering a large portion of the design space. A general description of the DLP-model training is provided. Then a validation is carried out against CFD results, at different scales, from global coefficients up to 3D volume. Once validated, the DLP-model is then used to optimise the hull form of a general purpose oil tanker. This leads to a wave height in the bow portion of the new vessel significantly smaller, resulting in a reduced wave resistance. While CFD calculations require significant time and effort without learning ability, DLP can predict instantly fluid behaviours to assess hydrodynamic performances of multiple designs. As a result, 3D high-fidelity predictions can be performed extensively in the initial design phase and allows early stage optimisation.


1/ Extrality, 75002 Paris, France

2/ Caponnetto Hueber, 46024 Valencia, Spain


/ Other published papers

INFINITY: Neural Field Modeling for Reynolds-Averaged Navier-Stokes Equations
Louis Serrano, Leon Migus, Yuan Yin, Jocelyn Ahmed Mazari, Patrick Gallinari
Learn more
Speed up Aerodynamic Optimization Thanks to Deep Learning Physics: Stellantis RAM Truck Example
Learn more
Multi-Objective Hull Form Optimization with CAD Engine-based Deep Learning Physics for 3D Flow Prediction
Jocelyn Ahmed Mazari, Antoine Reverberi, Pierre Yser, Sebastian Sigmund
Learn more
AirfoilRANS: High Fidelity Computational Fluid Dynamics Dataset for Approximating Reynolds-Averaged-Navier–Stokes Solutions
Florent Bonnet, Jocelyn Ahmed Mazari, Paola Cinnella, Patrick Gallinari
Learn more
An extensible Benchmarking Graph-Mesh dataset for studying Steady-State Incompressible Navier-Stokes Equations
Florent Bonnet, Jocelyn Ahmed Mazari, Thibaut Munzer, Pierre Yser, Patrick Gallinari
Learn more
Multi-scale Physical Representations for Approximating PDE Solutions with Graph Neural Operators
Léon Migus, Yuan Yin, Jocelyn Ahmed Mazari, Patrick Gallinari
Learn more
Need to hit production with short go-to-market and better control?
Cookie Consent

By clicking “Accept”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts.