Search Results

A method is described for predicting the reliability and useful life of an automotive powertrain system using a warranty database or from warranty records. The database requires failure corrections for misdiagnosis from duplicate data, trouble-not-identified records and multiple failure modes. Compensations not included in the database for high-mileage drop-out and warranty repairs less than the deductible amount, are also necessary. As an example, the cumulative hazard function of the Bathtub Hazard Rate distribution is fitted to the converted removal data of a typical automotive powertrain, to determine the product life characteristics. An algorithm written in Basic language is used to obtain the analytical results.

A Key Life Test is an accelerated test designed to detect a major component failure mode. The Key Life Test process will be outlined, as applied to predicting the life and reliability of an automotive synchronous timing belt. The main objective of this type of testing is the development of a model to improve quality by designing reliability into the product. A math model will be described to assess severe customer usage, determine the test conditions and acceleration factor and is used to find the timing belt life and reliability. The application of Probabilistic Methods will be explained to predict the belt life and reliability, after the model has been correlated against the test results obtained. It will be shown that the design can be optimized to improve the timing belt reliability using the model.