Barber, GaryHuang, XiaobaoGu, RandyWang, XiaLatcha, MichaelQu, HongweiZhou, Jun2024-10-022024-10-022024-01-01https://hdl.handle.net/10323/18262A numerical method was developed to detect debonding (crack initiation) condition during fatigue testing of an adhesive joint with thin steel sheets in standard configurations of coach peel and lap shear specimens. The goal is to establish the relationship between the sample's vibration behavior and initial failure condition, this relationship can be used to develop fatigue S-N curves for the adhesive under geometrical configurations of either coach peel or lap shear specimens. This method focuses on the continued monitoring of the specimen's vibration status by transferring the wave form data (in time domain) to a Fast Fourier Transform (FFT) process and obtaining frequency and amplitude data (in the frequency domain), these data are then processed to identify the critical events such as debonding (crack initiation), crack propagation and total failure with associated cycle counts for each event. The amplitude output files were used to analyze the progression of different stages of failure for the samples under cyclic loads. Video analysis of failure modes was utilized to identify the different types of damage matching the cycle count based on time scale (frames) for each stage: ANSYS models were generated to provide supporting evidence in terms of change in natural frequencies and amplitude due to change of internal stiffness of the specimens. Certain mode shapes were analyzed to gain confirmation of the relationship between mode shape, amplitude, and initial failure of the samplesAdhesiveAmplitudeBondingCrack initiationCrack propagationVibration