Rapid Vulnerability Assessment of Naval Structures subjected to Localised Blast

摘要

The development of modern naval vessels is driven by the optimum balance between operational performance, technology restrictions and the costs of ownership. These factors impose limitations on all features of surface ships, including weaponry, structural materials, radar systems, and propulsors. Strategies must be set to identify design features and materials that can enhance the vessels protection in the event of shock loadings e.g. air blast and underwater explosions. Assessment of design solutions is a complicated task due to the large number of unknowns involved. Appropriate computational models and experimental tests can give insights into the expected mechanical behaviour to support the design process. The authors are developing a framework for vulnerability assessment, which includes experimental tests and appropriate finite element (FE) models of representative structural parts subjected to blast loading. This combined approach provides a comprehensive analysis tool but its complexity prevents the quick assessment of the vessel structural vulnerability when various design features and a range of materials are to be considered. To overcome this hurdle, a machine learning model based on Artificial Neural Networks is proposed to identify patterns in numerical and experimental data, yielding timely conclusions about the structural response.