An Accelerated Life Test Model Based on Reliability Kinetics

William Q. Meeker, Michael J. LuValle

Technometrics · 1995 · 95 citations · 14 references

Concepts

TL;DR

Accelerated life tests rely on stress extrapolation, but conventional AFT models are only suitable for simple chemical failure mechanisms, underscoring the need for physics‑based approaches. The study presents humidity‑accelerated life test results for conductive anodic filament failures on printed circuit boards. The authors derive a chemically‑based accelerated life test model from an approximate kinetic failure model and estimate its parameters using likelihood‑based methods. Standard accelerated life test models are inadequate for these data, whereas the new chemically‑based models fit better and yield extrapolations more consistent with field data.

Abstract

Inferences from accelerated life tests typically involve extrapolation in stress. For this reason, it is important to use statistical models that have their basis in the physics and chemistry of the important failure mechanism(s). The standard accelerated failure time regression models (based for example on Weibull or lognormal approximations to the failure-time distribution at a given stress and linear scaling of time) are adequate for modeling some simple chemical processes that lead to failure. In this article we present the results from a humidity-accelerated life test of conductive anodic filament failures on printed circuit boards. Standard accelerated life test models are clearly inadequate for these data. Using an approximate chemical kinetic model of the failure process as a basis, we derive an alternative, more general, class of accelerated life test models. We illustrate the use of likelihood-based methods to estimate the model parameters. The new models fit the data better than the traditional accelerated life test models and provide extrapolations that are more consistent with actual field data.

References

14