Fatigue failures caused by cyclic stresses are commonly modeled by Birnbaum-Saunders (B-S) and Weibull distributions. Sometimes, materials with high cycle fatigue exhibit bimodal failure rates which are difficult to model with Weibull distribution. The objective of this paper is to investigate a general Birnbaum-Saunders (GB-S) distribution which covers diverse hazard rates such as increasing, upside down, multimodal and others. This paper also utilizes GB-S distribution to model fatigue failure data and validate its performance under different conditions. GB-S-based-ALT model is also developed to provide accurate reliability and lifetime prediction at operating conditions using data from accelerated life testing experiments.
