Abstract
The crashworthiness of hull structures is an important aspect which needs to be considered in the design of hull structures. Serious collision accidents may lead to casualties and economic losses. The raked bow is a common design for many ships, such as supply vessels, merchant vessels, oil ships, and warship. However, many collision studies focused on the structural damage caused by the bulbous bow, and the studies on the collision scene between the raked bow and the hull structure are relatively few. In this paper, the optimal design for the crashworthiness of the hull side structure under the impact of the raked bow is numerically studied. The dynamic collision response between the raked bow and the side structure is simulated by using the explicit finite element method in the commercial software LS_DYNA. Considering the damage and internal energy of the side structure, an objective function is established to describe the hull crashworthiness. The optimization design method for hull structure crashworthiness is established by combining the orthogonal experimental design, BP neural network and genetic algorithm (GA). The samples for training the neural network are determined by using the orthogonal experimental design method. The neural network model for the fast prediction of crashworthiness is established by using BP network and GA-BP network respectively, and it is found the GA-BP network can provide more accurate prediction results. Taking the mass and the scantlings of structural members as constraints, the optimization design of the side structure under the impact of the raked bow is realized.
