- To design engineering systems that have improved reliability, it is important to understand what kind of system faults they will be susceptible to. Mitigation strategies are important to ensuring the performance of these engineering systems. Understanding how the modularity of complex engineering systems affects the risk of devastating failures such as cascading failures can help enable engineers to implement strategies in the design phase to increase reliability. The extent to which decreased system modularity propagates the spread of a cascading failure is unknown. This study analyzes how modularity in complex engineering systems affects resistance to the spread of cascading failures. In this research, synthetic networks are used to represent component models at differing degrees of modularity. These synthetic networks are then infected through epidemic spreading models that model cascading failures. The loss of functionality is determined by the percent of diseased nodes in the system, and the influence of the initial node is measured by eigenvector centrality. Increased modularity is associated with the improved ability of a system to inhibit the propagation of cascading failures over time through failure isolation within a module, measured by percent infected in the system, in comparison to less modular systems (p < 0.001). This finding indicates that the structural design of complex engineering systems could be crucial to increasing reliability in design with reference to cascading failures.
Key Words: Complex engineering systems, modularity, cascading failures