Surface defect detection from additive manufacturing components at elevated temperatures using laser-generated Rayleigh waves

Abstract

Online monitoring is a critical issue for additive manufacturing components layer by layer to ensure the integrity of the structure. However, the rough surface and high temperature make it difficult to use conventional non-destructive testing methods to carry out. In this work, the laser ultrasonic technique, as a non-contact nondestructive testing method, is applied to detect defects in a 316L stainless steel specimen manufactured using selective laser melting at elevated temperatures. The unreported phenomenon of phase evolution of Rayleigh waves for different excitation locations and temperatures is observed and systematically explored using the finite element method and experiments.Furthermore, an improved variational mode decomposition strategy based on particle swarm optimization is proposed to enhance the signal-to-noise ratio. A fitness function based on permutation entropy and mutual information is developed to determine the optimized parameters in variational mode decomposition. The experimental results indicate that, for the different excitation locations, the amplitude of the Rayleigh wave generated by laser irradiation on the defect is larger than on the small area around the defect, and the defect can be effectively detected using B-scans at elevated temperatures according to the abnormal evolution of Rayleigh waves; for the fixed excitation locations, the amplitude and velocity of Rayleigh waves gradually decreases at 100–500 °C, which will provide a scientific basis for online monitoring and evaluation of additive manufacturing components.