Abstract
In this paper, more than thirty years of local operational field data were used to predict and validate the failure rate of the Lockheed C-130 (T-56) Turbine with respect to turbine life data in hours (t) using both Weibull and lognormal regression models. By using Weibull Analysis on (MS Excel), the data was fitted into the model using two parameters, a good straight line fit to the transformed data is supporting the validity of the Weibull model. Furthermore, a validation of our MS Excel spreadsheet format of Weibull analysis was compared with the result obtained from Weibull and lognormal regression using Weibull++7.
To compare Weibull and lognormal results, we consider which distribution more closely matches the data line, by comparing the respective coefficients of determination R-2 of the regression lines. If the coefficient of determination is close to 1, then we have supporting evidence that the data points from linear relationship and hence the distribution is good model. For the Weibull line, R-2 = 0.999796 and for the lognormal line R-2 = 1, from the comparisons of both R-2 results, the Lognormal distribution provides slightly better results than the Weibull distribution. The correlation coefficient for the Weibull distribution is rho = 0.988875, and for the lognormal distribution rho = 0.941053. So, both distributions are consistent for the reliability analysis. The resulting parameters indicate that the engine turbine has an increasing failure rate over time which makes a planned replacement policy worthwhile. The most common causes of failures in this range are corrosion, erosion, fatigue, and cracking. Since the component exhibits wear out failure pattern, a hard time maintenance action which involves planned replacement and overhaul program is required.
Failure modes and effects analysis FMEA, was used to analyze the failure modes, causes and effects of the c-130 T-56 engine turbine. Once FMEA data is produced, it should be ranked in component order by risk priority number RPN giving a clear picture of the unreliability of components. Also, graphical charts and matrices could be complementary tools to reach the needed more effective actions.