Temperature-humidity oriented reliability prediction for electronic equipments

This paper presents a practical method of reliability prediction for electronic devices with operating temperature and relative humidity that will not always be in steady state. Since operating temperature (T_u) and relative humidity (RH_u) are various in different regions, the authors use temperature probability distribution function f(T_u) to indicate the change of T_u, and humidity probability distribution function g(RH_u) is used to indicate the variation of RH_u. In order to estimate equipment reliability for the entire area of product usage, this paper uses these two functions to modify prediction model. The proposed method offers a way to predict the average value of failure rate for the entire area of product usage with accelerated test data and climate information.