Complex engineering systems consist of many subsystems. Each of the subsystems is composed of a large number of components. While faults arise at component level, sensing capabilities are limited to subsystem level, and system operations and maintenance practices are scheduled based on system level paremeters. This paper presents a hierarchical architecture to analyze the effects of system level parameters on component level faults of dominant failure modes of a complex system. An aeropropulsion system of turbofan type has been used as the application domain. In most of the cases, engine life is limited due to cracks in high-pressure turbine blades. In this paper, it is assumed that creep is the only active failure mechanism. Based on a finite-element model of the turbine blades available in the open literature, design of experiments (DoE) methodology is used to build a subsystem-level model. A simulation package of a commercial aircraft engine is then used to obtain system-level results.