ABSTRACT:

The optimization system is constructed using the genetic algorithm and the PrimeShip-HULL distributed by Nippon Kaiji Kyokai that is the rule calculation software for the CSR (Common Structural Rules). The constructed system is applied to the optimization of midship sectional scantlings of 150KDWT tanker. As a result, the design parameters are obtained so that the hull structural weight and the construction cost are minimized with satisfaction of the scantling requirements of the CSR in a practicable time.

INTRODUCTION

In accordance with the CSR (Nippon Kaiji Kyokai, 2006), the midship sectional scantlings should be determined by the scantling requirements and direct strength analyses. In the scantling requirements of the CSR, there are roughly six categories, the local strength, the hull girder bending strength, the hull girder shear strength, the hull girder buckling strength, the fatigue strength and the hull girder ultimate strength. To minimize the construction cost of a target ship with satisfaction of these requirements, a designer should determine the midship sectional scantlings rationally. However, many design parameters for the midship sectional scantlings should be determined based on the experience and the trial and error method since it is difficult to optimize many design parameters with meeting all the requirements. Such a design problem can be treated as combination of optimization problem of design parameters, and many studies have been performed since before (e.g. Kitamura et al., 2001a, 2001b, 2007; Lee et al., 2001; Nakamori et al., 2010; Okada et al., 1994; Rigo, 2005). In those studies, it is shown that the optimization applying the genetic algorithm is very useful, but the examination that takes all scantling requirements of the CSR into consideration is still difficult. The rule calculation software that the Classification society offers is usually used for evaluating the scantling requirements.

This content is only available via PDF.
You can access this article if you purchase or spend a download.