On a truncated accelerated plan for two component parallel systems under ramp-stress testing using masked data

摘要

Several studies on design of AcceptanceLife Test (ALT) focused on a subsystem (single system) totally ignoring itsinternal design. In most cases, it is not always possible to identify thecomponents that cause the system failure or the cause can only be identified bya subset of its component resulting in a masked observation. This papertherefore investigates into development of ramp-stress accelerated life testingfor a high reliability parallel system that consist two dependent componentsusing masked failure data. This type of testing may be very useful in atwin-engine plane or jet. A ramp-stress results when stress applied on thesystem increases linearly with time. A parallel system with two dependentcomponents is taken with dependency modeled by Gumbel-Hougaard copula. Thestress-life relationship is modeled using inverse power law and cumulativeexposure model is assumed to model the effect of changing stress. The method ofmaximum likelihood is thereafter used for estimating design parameters. Theoptimal plan consists in finding optimal stress rate using D-optimalitycriterion by minimizing the reciprocal of the determinant of Fisher informationmatrix. The projected plan is also explained using a numerical example andsensitivity analysis carried out. This formulated model can help controlengineers to obtain reliability estimates quickly of high reliability productsthat are likely to last for several years.