Multiobjective parametric optimization of the structural panel of a merchant ship

Vitor Hugo Benevenutti Müller

CHALLENGE - The competitive global shipping market forces ship owners to demand more and more efficient vessels. This encourages engineers to seek new design alternatives and apply optimization techniques that aid decision making. The application of such techniques in the structural design of vessels is highlighted due to the interactivity of this stage. This work is inserted in this context, and seeks to analyze the use of optimization techniques applied to the design of a structural panel of a merchant ship, described through the variables that define it. 

SOLUTION - The reinforcing elements were treated as beams and the optimization utilized the genetic algorithm NSGA-II. The optimization process was run in optimization platform modeFRONTIER. The Pareto frontier analysis obtained for different percentages of DOE - design of experiments - in relation to the total number of individuals was deepened. 

BENEFITS - This work verified that the number of points generating the border is directly proportional to the total of evaluated individuals and that the percentage of 5% of DOE corresponds to the dominant optimal border in relation to the others. The tendency is also seen for the arrangements of heavy reinforcers with spacing of the same order and the multimodality of the structural panel problem is confirmed.