Prognostics and Health Management using Nonlinear Cumulative Damage Model for Electronic Devices under Variable Loading

This paper explores nonlinear cumulative damage models applied to Prognostics and Health Management in electronic systems, with a focus on solder joint reliability under variable vibrations. Several cumulative damage models, including the total strain energy density model, are investigated to address the limitation of Miner's rule. The State-of-Damage (SoD) concept is introduced as a pivotal metric for evaluating Prognostics and Health Management performance within each model. A novel SoD definition with a scale factor is proposed to balance linearity and exponential behavior, meeting the demands of real-time monitoring. Furthermore, by calculating Remaining Useful Life (RUL) using the SoD metric, the RUL predictions from Miner's rule and total strain energy density model with the scale factor are compared. The results reveal that the total strain energy density model, enhanced with the scale factor, outperforms Miner's rule in accurately predicting SoD and RUL. This study advances Prognostics and Health Management methodologies by introducing an innovative SoD concept with a scale factor, resulting in enhanced accuracy for both SoD and RUL predictions. This innovation holds the potential for more reliable electronic systems under variable loading conditions.