Normally, this is a one-time process through which a product is stressed at a higher level than it would experience during normal use. Accelerating the use conditions in a short time span will expose latent defects that may otherwise occur much later at a customer site. ALT drives design changes needed to eliminate these causes of failure. In order to derive an accurate failure distribution based on the failure data, the UUT should be tested well beyond the infant mortality stage of the product's life distribution. This applies to mechanical and electromechanical products. For electronic products, it is acceptable to test to the edge of the infant mortality region. This region is considered to be from 3000 to 10000 actual hours. To reach this level at normal operating temperature would therefore take up to 10000 actual hours. Our policy is to design a test to run for 720 actual hours at 70°C which approximates 7000 effective hours of life. At the conclusion of this test, the derived MTBF, the probability (reliability) of operating for x amount of hours, and the percent failure rate per year will all be derived from the test results. The important point to keep in mind is that the longer the test duration the greater the confidence limits around the calculated (advertised) results.
Result: Reduction in field returns and a greatly improved MTBF and Availability predictions.