High-Rate Damage Classification and Lifecycle Prediction via Deep Learning

Abstract

While the lack of high-fidelity physics in high-rate dynamic events has posed as a challenge for many years, machine learning (ML) is making significant breakthroughs in numerous practical engineering applications. The ability to classify precise levels of damage and predict component-level failure becomes increasingly important with high-rate applications. Structural health monitoring (SHM) benefits from ML algorithms, which extract features containing a component’s damage information, thereby reflecting a system’s overall health status. This research presents a deep convolutional variational auto-encoder (CVAE) that classifies damage levels of electronic assemblies subjected to high-acceleration mechanical shock. These electronic assemblies were shocked using a drop tower, damaging individual components, and causing overall system failure. Wavelet transformations are used to maintain adequate time-frequency information of the shock response and augment the presentation of its high-rate dynamics. Given only six training examples containing sparse data, the CVAE used in this paper classifies the severity of an electronic package’s damage, which could extend to predicting useful remaining lifetime.