Hydrodynamic Optimization of Trimaran Workboat Hulls

Abstract

The object of this master thesis was to perform an initial design optimization for a trimaran concept, with the goal of minimizing the hydrodynamic resistance for a speed of 30 knots. The thesis is written in cooperation with VARD Ship Design AS. VARD has already developed a trimaran design operating at a speed of 25 knots, but aim to increase the design speed. With the initial design as a starting point, the focus of attention was set to optimization of the side hulls form and position relative to the main hull. The side hulls can improve the resistance performance of the trimaran as destructive wave cancellation may be achieved with proper arrangement. However, they can also add to the resistance as an increase in form give a larger frictional resistance. Optimization of the trimaran design is hence a complex problem.

Optimizing with respect to both form and location implied testing a vast range of different configurations. In order to do this, a custom hybrid model coupling an optimization algorithm with a flow solver was utilized. Generated by a MATLAB script, the model allowed numerous different configurations to be tested dynamically. Particle swarm optimization was used as the optimization algorithm and the potential theory based software Michlet was used as the flow solver. Optimization was performed within a search space given by the initial design and thus specified by VARD.

Evaluation of the best design found with the hybrid model was performed with Computational Fluid Dynamics (CFD), through the open source software OpenFOAM. The open source 3D geometry software Blender was used to handle the geometry for the CFD simulations.

The results from the hybrid model and CFD showed the same trends, but no significant improvements to the design were obtained during this thesis. A reduction of the total resistance was found to be 3.2\% based on the initial problem statement given by VARD. It was found that the optimal location of the side hulls was located outside the initial area of interest. The resulting side hulls were of insufficient scale as frictional resistance dominated, giving minimal wave cancellation. Further reduction to the total resistance can however be achieved, if the model is allowed to move more freely.