TABLE OF CONTENTS To obtain information about the life distribution F T( t) of a component, it is necessary to carry out a 'life test' where n identical units of the component are activated and their lifetimes recorded.
The fundamental assumptions that are made are that the lifetimes of the n components are statistically independent and identically distributed according to the continuous distribution function F T( t).
The assumption of identically distributed lifetimes corresponds to the assumption that the components are nominally identical, that is of same type and exposed to approximately the same environmental and operational stresses.
The assumption of independence means that the components are not affected by the operation or failure of any other component in the set.
Any censoring mechanism (see below) must also be 'independent', i.e. censorings occur independent of any information gained from previously failed components in the set.
If the test is allowed to run until all the n components have failed and the lifetimes are recorded, the data set thereby obtained is said to be complete (Figure 9.1).
Often, it is impractical or too expensive to wait until all the components have failed. Hence, censoring is applied to cease the test before all components have failed.
A right-censored data set...
