Meeker and LuValle (1995) give data from an accelerated life test on failure of printed circuit boards. The purpose of the experiment was to study the effect of the stresses on the failure-time distribution and to predict reliability under normal operating conditions. More specifically, the experiment was designed to study a particular failure mode - the formation and growth of conductive anodic filaments between copper-plated through-holes in the printed circuit boards. Actual growth of the filaments could not be monitored, only failure time (defined as a short circuit) could be observed directly. Special test boards were constructed for the experiment.
The data provided here are the number of failures observed in each of a series of 4-hour and 12-hour long intervals over the life-test period. This experiment resulted in interval-censored data because only the interval in which each failure occurred was known. Further, the data are part of the results of a much larger experiment aimed at determining the effects of temperature, relative humidity, and electric field on the reliability of printed circuit boards.
A data.frame
with 140 rows and 5 variables:
[, 1] | lower | Start of an inspection interval (in hours) | Numeric |
[, 2] | upper | End of an inspection interval (in hours) | Numeric |
[, 3] | count | Number of events observed in the interval | Numeric |
[, 4] | event | Event observed in the interval (right-censored/interval-censored) | Categoric |
[, 5] | rh | Relative humidity applied | Numeric |
Meeker, W. Q., and LuValle, M. J. (1995), An accelerated life test model based on reliability kinetics, Technometrics, 37, 133-146.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.