A reliability decision framework for multiple repairable units. (June 2016)

• Record Type: Journal Article
• Authors: Garmabaki, A.H.S.; Ahmadi, Alireza; Block, Jan; Pham, Hoang; Kumar, Uday
• Is Part Of: Reliability engineering & system safety. Volume 150(2016:Jun.)
• Page Start: 78
• Page End: 88
• Publication Date: 2016-06
• Subjects: Multiple repairable units -- Reliability modelling -- Homogeneity -- Heterogeneity -- Trend test; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2016.01.020
• Languages: English
• ISSNs: 0951-8320

Abstract

Abstract: In practice, the analyst is often dealing with multiple repairable units, installed in different positions or functioning under different operating conditions, and maintained by different disciplines. This paper presents a decision framework to identify an appropriate reliability model for massive multiple repairable units. It splits non-homogeneous failure data into homogeneous groups and classifies them based on their failure trends using statistical tests. The framework discusses different scenarios for analysing multiple repairable units, according to trend, intensity, and dependency of the units׳ failure data. The proposed framework has been verified in a fleet of aircraft and in two simulated data sets. The results show a reliability model of multiple repairable units may contain a mixture of different stochastic models. Considering single reliability models for such populations may cause erroneous calculation of the time to failure of a particular unit, which can, in turn, lead to faulty conclusions and decisions. When dealing with massive and non-homogeneous multiple repairable units, the application of the proposed framework can facilitate the selection of an appropriate reliability model. Highlights: The study provides a reliability decision framework for multiple repairable units. It splits failure data into homogeneous groups based on their trend behaviour. The application of framework is demonstrated by fleet of aircraft.