ABSOLUTE VS COMPARATIVE END-OF-LIFE AGE

Conventional insulation statistics typically use test averages of a no rmal, log normal, or Weibull distribution to compare the quality of di fferent materials or constructions. None of these numbers characterize s an age at end of life with a near 100% reliability, which is necessa ry to predict when failure occurs. This is a major fundamental and mis leading aspect in the manner in which life and age data is reported. T he distributed nature of insulation aging and failure requires taking into account the length of samples to report age at end of life. This implies that probability of failure at any time should be reported as a function of cable length. Simply reporting an average probability of failure can endanger human life and have negative economic consequenc es. This paper shows test data and a method of calculating end of life that provides high reliability operation of aircraft wire that is agi ng from root-cause stresses. The authors believe this paper treats a m ajor refinement which is essential for realistic prediction of insulat ion end of life, for instance in the control of reliable operation of aircraft electrical wiring. This technique is now being used in mainte nance scheduling of active aircraft and is hoped to influence future r evisions of standards.