Prediction of failure in time (FIT) of electrical connectors with short term tests

Abstract

Failure in time (FIT) is an important measure for the reliability of electrical connectors. Due to the very long lifetime of connectors, the tests for the determination of FIT rate are time and labour intensive. In this paper a data driven method using a statistical process to estimate the FIT rate of electrical connectors with data of electrical contact resistance development in short term tests is proposed. The results of prediction are then compared with the results from long term tests. The study shows a strong correlation between contact resistance development in short term tests and the development of the number of failures in later stages of tests. In order to predict the development of degradation precisely, the distribution of resistance data in many different tests with different connectors is investigated. The Generalized Extreme Value Distribution, which reveals an ideal fitting, has been implemented for the prediction of the failure rates of connectors, thereby enabling a remarkable time-lapse of lifetime tests. This method can also be employed in the prognosis and management of system health through the forecast of health of connectors in different systems in operation.