DROPSS - Drag Optimisation of Ship Shapes

Background

International shipping is responsible for the transport of around 90% of the global trade. The dominant role of shipping is attributable to the low fuel consumption per tonne-km of transported cargo. However, the mere magnitude of around 45-50 thousand operating merchant vessels puts environmental and economic aspects of shipping more and more into the focus of optimisation efforts. The seaborne pollution as well as approximately 50% of the direct operating costs for shipping are related to fuel consumption, which in turn is governed by the resistance of the vessel. The latter is largely controlled (approximately 75%) by steady hydrodynamic contributions, i.e. the wave drag in calm water and the friction drag along the wetted surface. Therefore, reducing these drag contributions- even by a few per mille- is highly appreciated from commercial and environmental perspectives.

Aims and Objectives

The present proposal is concerned with the development of a robust, holistical and reliable optimisation framework for the hydrodynamic shape optimisation of free-floating ship hulls exposed to immiscible two-phase flows. Adjoint methods will be used in a parameter-free, gradient-based optimisation environment to capture wave and friction drag at Froude- and Reynolds-numbers of practical interest. Specific interest refers to accurate and robust adjoint two-phase flow models including an adjoint compressive interface treatment inspired by a Cahn-Hilliard Reynolds-Averaged Navier-Stokes approach, a dynamic change in the floating position, a novel reduced adjoint turbulence treatment as well as an improved technique for partially wetted shapes subjected to manufacturing constraints. Results will be transferable to the optimisation of other marine engineering applications, e.g. the design of near-shore operating renewable energy devices harvesting ocean waves in the surge zone close to a free surface. 

Example

Using a steepest descent approach, the shape-derivative (sensitivity) is preferably obtained by adjoint methods due to the independence of the computational costs from the number of design variables. The upcoming figure shows the sensitivity distribution of a drag functional along the controlled wall area (hull), which was computed at the cost of one additional (adjoint) CFD simulation.

 
Funding and Collaborations

The project is funded (100% - 3 years) by the Deutsche Forschungsgemeinschaft (DFG) and also implemented into the Lothar Collatz School (https://www.c3s.uni-hamburg.de).

Personnel    

Niklas Kühl,  Thomas Rung