A Particle Swarm Algorithm-Based Optimization for High-Strength Steel Structures

Abstract

Crash resistant ship structures with high post accidental residual strength are more and more important as a requirement of Goal Based Standards. A first step in the development of structures, which fulfill these requirements, is the rational identification of crashworthy structural concepts. Therefore, efficient optimization algorithms are needed, besides reliable numerical collision simulations. Particle swarm optimization (PSO) algorithms tend to improve the objective fast, which makes them a good choice for lengthy nonlinear finite element-based collision simulations. As an application of PSO and numerical collision simulations a liquefied natural gas tanker is optimized for crashworthiness. Furthermore, one concept using normal shipbuilding steel only and one concept using a combination of normal and high strength steel is investigated. Hence, it will be shown if the benefits from high strength steel outweigh the increase in cost. Therefore, it will be shown if the combination of normal and higher strength steel can result in increased crashworthiness, even though the high strength steel has a lower failure strain when compared with the normal steel.