A reliability engineer studies the failure rates of engine windings of turbine assemblies to determine the times at which the windings fail. At high temperatures, the windings might decompose too fast.
The engineer records failure times for the engine windings at 80° C and 100° C. However, some of the units must be removed from the test before they fail. Therefore, the data are right censored. The engineer uses Nonparametric Distribution Analysis (Right Censoring) to determine the following:
The times at which various percentages of the windings fail.
The percentage of windings that will survive past various times.
The survival function for the engine windings (as shown on a survival plot).
Whether the survival curves at the two temperatures are significantly different.
Choose Stat > Reliability/Survival > Distribution Analysis (Right Censoring) > Nonparametric Distribution Analysis.
In Variables, enter Temp80Temp100.
Click Censor. Under Use censoring columns, enter Cens80Cens100.
In Censoring value, type 0. Click OK.
Click Graphs. Select Survival plot.
Click OK in each dialog box.
Interpret the results
The estimated median failure time for Temp80 is 55 hours and the estimated median failure time for Temp100 is 38 hours. Therefore, the increase in temperature decreases the median failure time by approximately 17 hours.
Minitab displays the survival estimates in the Kaplan-Meier Estimates table. At 80° C, 0.9000 (90%) of the windings survive past 31 hours. At 100° C, 0.9000 (90%) of the windings survive past 14 hours.
In the Test Statistics table, a p-value < α (usually, α = 0.05) indicates that the survival curves are significantly different. In this case, the both p-values (0.005 and 0.000) are less than α, which suggests that a change of 20° C has an effect on the breakdown of engine windings.