Novel design and hydrodynamic optimisation of a high-speed hull form

Paper in proceeding, 2005

A deep-Vee hull form, renowned for its favourable seakeeping, manoeuvring and competitive resistance qualities at high Froude number (Fn>0.55), is re-designed for lower Froude number (Fn=0.388 & 0.2). Novel features include a bulbous bow and a modified transom stern. The new design, modelled using the FRIENDSHIP CAD system, has a displacement of 16,000 tonnes. The bulbous bow and the modified stern are optimised for resistance using two different techniques, namely an Artificial Neural Network (ANN) approach and the gradient-based Method of Moving Asymptotes (MMA). The techniques are compared for efficiency and performance. The wave pattern resistance response surfaces are computed using the SHIPFLOW CFD software by systematically varying the design parameters for the bulbous bow and stern. These surfaces, which display the cause-and-effect relationships in the design procedure, are stored as a database for training and using the ANN method. In SHIPFLOW, two techniques are used for computing the resistance: pressure integration and transverse wave cut. It turns out that the transverse wave cut technique is the most robust one for this purpose.