Abstract: In this paper, a method to detect bolt crack defects using vibroacoustics technology is proposed. The feasibility of this method to identify bolt crack damage is demonstrated through percussion test, modal simulation and theoretical derivation. The test proves that the vibroacoustics method can effectively identify the location and depth of bolt crack defects. In addition, the minimum discernible defect depth of M30×200 mm bolt is 2 mm, and the defects located in 1/2 and 1/4 positions of bolts are most easily identified. Combined with the common fatigue fracture forms of bolts in engineering field, this method can effectively identify the fatigue cracking of bolts, which has certain engineering application significance. In addition, this paper proposes the research direction of the fusion of vibroacoustics technology and machine learning technology, providing intelligent research ideas for bolt damage detection.