Step-Stress Accelerated Life Testing Designs

The term "Accelerated Life Test (ALT)" applies to the type of study where failure times can be accelerated by applying higher "Stress" to the component. This implies that the failure time is a function of the so-called "Stress Factor" and higher stress may bring early failure. Models and methods of accelerated life testing are useful when technical systems under test have a very long lifetime. This thesis consists of different designs of Accelerated Life Tests such as Step Stress Accelerated Life Test (SSALT), Step Stress Partially Accelerated Life Test (SSPALT) with progressive data and constant stress accelerated life test using Geometric Process. The method of maximum likelihood (ML) estimation is used to obtain the estimates of model parameters because it provides estimators that have both a reasonable intuitive basis and many desirable statistical properties. In order to get the asymptotic variance of the ML estimator, the Fisher Information Matrix is constructed. In particular, this thesis focused on the derivation of different ALT plans and designs for the lifetimes of units that are assumed to follow Frechet life distribution at use stress with different types of schemes.