Understanding cascading failures through a vulnerability analysis of interdependent ship-centric distributed systems using networks

Abstract

System vulnerability is of critical concern when evaluating the operational performance of a naval vessel. The identification of system failures as well as subsequent cascading failures of related systems is required to understand if the vessel's distributed systems can survive a damage scenario through the prevention of large scale failures. In late-stage ship design, high-fidelity system analysis and scenario simulations can identify this type of emergent failure. However, in early-stages when distributed system design information is limited, vulnerability analysis typically focuses on component damage and vessel layout. Without considering the effect of distributed system design, the identification of cascading failures can not be identified early in the design process. To address this problem, the authors describe a distributed system model and corresponding vulnerability analyses to consider how vessel layout and distributed system configuration impact vessel performance under damage with limited information. The approach uses network-based methods to reduce the design detail required for distributed system modeling and vulnerability analysis. This approach is demonstrated on illustrative cases and a naval ship concept example to illustrate the importance of identifying cascading failures in early stage design.