Defect detection for solder joints with spectrum kurtosis and empirical mode decomposition

Abstract

The solder joints easily acquire defects due to the vibration loadings in service. So defect detection is of great importance for prognostics and health management (PHM) of electronics. In this paper, a defect detection method with spectrum kurtosis and empirical mode decomposition (EMD) is proposed for solder joints subjected to vibration loadings. First, the global and local strain data of electronic components is decomposed into several intrinsic mode functions (IMF) based on empirical mode decomposition (EMD). Then the maximum spectrum kurtosis method is used to construct band-pass filters to filter the high frequency IMFs. Finally, the fault symptom vector is developed by computing and reconstructing the envelope spectrum, which contains incipient fault information. The experimental results demonstrated that the method can detect the latent failure of board level package effectively.