Reliability growth curves are modeled using the results of Weibull distribution regression analysis. The Weibull Learning Curve (WLC) model is useful because it provides an early statistical indication of wear out issues. The vibration, thermal, humidity and corrosion environments of the F-22 are severe. It is prudent to be concerned about wearout problems in severe aircraft environments. The power of Weibull is the ability to predict the characteristic life and distribution shape with a small amount of data. The proposed model is compared to the traditional Duane Learning Curve approach. Statistical model validity is evaluated using numeric and graphical methods, including coefficients of determination (r2) and residuals. The WLC model provides insight into interdisciplinary issues, including supplier quality control, qualification testing, physics of failure, and reliability engineering. In addition, the WLC model quantifies the number of items with removal statistics consistent with wearout; therefore, this study helps to quantify successes and shortfalls of durability analysis and testing approaches.

This content is only available via PDF.
You do not currently have access to this content.