Kittaneh, OmarMajid, MohamedHelal, Sara2023-06-042023-06-0420231099-1638https://doi.org/10.1002/qre.3378http://hdl.handle.net/20.500.14131/881The paper introduces a new model for accelerated life testing analysis with nonthermal accelerating stress, which combines the inverse power law (IPL) with a normal distribution. Typically, lifetime probability distributions with a shape parameter are used in this context, but many distributions without this parameter are excellent fits to experimental lifetime data and have not been considered. For the first time, the paper demonstrates that the coefficient of variation can replace the shape parameter, allowing the use of normal distributions. The IPL-normal model is introduced in a mathematical setup that leads to the least squares estimating equations and maximum likelihood estimates of the parameters. This model uses accelerated experimental data to predict the lifetime of organic light-emitting diodes (OLEDs) at use conditions and is benchmarked against prior works validated by market studies.In accelerated life testing analysis with nonthermal accelerating stress, the inverse power law (IPL) is often solely merged with a particular lifetime probability distribution with a shape parameter. Although many fundamental lifetime distributions, such as the normal distribution, are excellent fits to the experimental lifetime data, they have not been considered as they lack the shape parameter. As such, this paper, for the first time, demonstrates that the shape parameter can be replaced by the coefficient of variation, allowing the use of normal distributions in this context. The work further introduces the IPL-normal model in a rigorous mathematical setup that precisely leads to the least squares estimating equations and maximum likelihood estimates of the IPL-normal accelerating parameters and the general coefficient of variation. The proposed model uses accelerated experimental data to successfully predict the lifetime of organic light-emitting diodes (OLEDs) at use conditions. Based on these fundamentals, the predictions are benchmarked with prior works that were validated by market studies.en-USOLED; ALT; coefficient of variation; inverse power law model; regression analysis; MLE.ENERGY