Modeling frailty in manufacturing processes.
Wassell-JT; Kulczycki-GW; Moyer-ES
Lifetime Data: Models in Reliability and Survival Analysis 1996:353-361
The purpose of this study was to estimate the distribution of breakthrough times and the effects of alternative challenge agents in the laboratory testing of organic vapor respirator cartridges. This report concerned multivariate frailty models based on both the gamma and positive stable frailty distributions. Frailty models were used to account for unobserved differences in manufacturing processes and raw materials used to produce respirator safety devices. Survival distribution estimates were obtained when the baseline hazard was assumed to be Weibull. Using laboratory test data, the failure times for 104 respirator cartridges produced by ten different manufacturers were compared. Breakthrough times were determined using carbon-tetrachloride for 30 cartridges, pentane for 36 cartridges, and hexane for 38 cartridges. Both frailty models provided a significant improvement over a Weibull model assuming independence as indicated by likelihood ratio tests. Results indicated that the positive frailty model might provide a better fit than the gamma frailty model for this data set.
Personal-protective-equipment; Respiratory-protection; Safety-equipment; Respirators; Lung-function; Equipment-reliability; Mathematical-models
Book or book chapter
Respirator Research; Respirators
Lifetime Data: Models in Reliability and Survival Analysis