ABSTRACT
This paper introduces a small scale experiment, which combines performance and reliability measures yielding in component reliability estimation. A sample of LED is tested with constant power. Power values are defined within a given interval. Results are MTTF assigned to given power levels. All MTTF values compose a curve quantifying the relation between performance and reliability. Calculating MTTF values, goodness of fit tests are conducted. As a secondary effect, interpretations of physical effects occurring during the tests can be stated. Validating the MTTF curve, step-stress tests are conducted. The results are embedded in a Linear Damage Hypothesis (LDH). The values of the LDH measure D (see Section 7) give hints on the applicability of the approach. In the given case, validation leads to a power interval re-definition.
