Extensive investigation of Calibrated Accelerated Life Testing (CALT) in comparison with classical Accelerated Life Testing (ALT)

Various accelerated reliability test methodologies have been recently proposed with an aim of reducing sample sizes. Dramatic decrease in failure rates, for electronic products targeted in this study, is the main motivation for these methodologies; performing conventional Accelerated Life Tests (ALT) is getting extremely time consuming and costly. Calibrated Accelerated Life Testing (CALT) is one of the well accepted methodologies that aims to use fewer samples than those used in ALT. We investigate ALT and CALT with different failure rates. We show that as opposed to the conventional view, considering required test time CALT does not always overwhelm ALT; it highly depends on failure rates, acceleration factors, and stress levels. We also compare these methodologies by considering accuracy that is the main drawback of using CALT. In this study, we define an accurate test method choice between CALT and ALT by failure rates, acceleration factor, warranty times and mean-time-to-failures.