JOURNAL OF THE FACULTY OF ENGINEERING AND ARCHITECTURE OF GAZI UNIVERSITY, cilt.25, sa.4, ss.827-837, 2010 (SCI-Expanded)
Before any reliability analysis, trend tests must be carried out to check whether the assumption of independent and identically distributed (iid) observations is satisfied or not. When trend is detected, Nonhomogenous Poisson Process (NHPP), which assumes that the Time Between Failures (TBF(s)) vary as a function of time, must be used to describe the trend. In this study, the tests for trend showed that times between successive failures for the mechanical systems of a dragline are not independent and identically distributed. Goodness-of-fit tests showed that the Power Law Process (PLP), a certain form of NHPP, provides a good fit to the failure data. Besides, expected failure number, predicted time to next failure and reliability for different time periods have been estimated by using the parameters of PLP.