Application of Cumulative Degradation Model to Acceleration Life Test

Publisher Website
Google Scholar
Abstract
A general degradation model which includes conventional acceleration tests such as fixed, progressive, and step stress experiments is derived from the reaction theory under the assumption of linear degradation accumulation. Its application to the acceleration test is discussed. According to the reaction theory, degradation of the characteristic parameter, μ is connected to reaction rate K and time t by a linear transformation function f(μ) = Kt. The total degradation is determined by the linear accumulation of the Kiti product such as f(μn)= # Kiti. This relationship is also expressed as a generalized Miner's equation #(ti/Li) = 1 in which ti and Li are the actual stressing time and expected life under the ith stress condition, respectively, and # ti is the life expectancy of the component under successive stress conditions from i = 1 to n. Validity of this linear accumulation assumption is investigated under various stress application paths. For monotonic step-up, cyclic, and random stress, the rule almost holds as a whole, but monotonic step-down stress sometimes causes erroneous recovering effects of the parameter. The degradation accumulation principle is effectively applicable to integrate degradation and failure pattern information and also to avoid some shortcomings of conventional life test methods. 1) For example, in order to find efficiently the life vs. stress plot, we could combine the degradation (failure fraction) vs. stress diagram, obtained by one step stress experiment, and the knowledge of the degradation pattern obtained by a constant stress test.

This publication has 2 references indexed in Scilit: