Design stage confirmation of lifetime improvement for newly modified products through accelerated life testing

摘要

After a modification to the original version of a product and before mass production, the expected improvement in the product lifetime or reliability needs to be validated. This paper presents three approaches based on accelerated life testing to verify, estimate and confirm the lifetime or reliability of a newly modified product at design stage: ALT comparative approach, reliability estimation approach, and reliability validation test. Test samples of the original and modified versions are expected to fail during the tests in order to obtain their failure time data. In ALT comparative approach, the statistical comparison between failure time data of the original and modified versions is used to verify the required improvement in lifetime. In reliability estimation approach, the relationship made between available lifetime and failure time data of the original version is used to extrapolate lifetime data of the modified version from its failure time data. Since modified versions are usually highly reliable, all test samples might survive the tests (without any failures) that results in the lack of failure time data for statistical analysis. To confirm a level of service reliability with confidence, reliability validation test is presented to make an estimate of the number of samples required to survive the tests. To fulfill the same level of confidence for fewer number of prototypes (as test samples), the test time must be extended. On the other hand, more prototypes are needed to pass a shorter test time if there are any time constraints.