Analysis of sound and vibration interaction on a crack and its use in high-frequency parameter selection for vibro-acoustic modulation testing

Vibro-acoustic modulation (VAM) is a sensitive nonlinear method used to nondestructively detect cracks and other contact-type defects in materials and structures. However, the modulation between two excitation signals in a structure has not been investigated well. In this study, the interactions between the ultrasound (probing), vibration (pumping), and crack are investigated. Using aluminum beam samples, a sweep signal is used as the probing excitation, and synchronic demodulation and a short-time Fourier transform are used to extract the modulation from the output signal. The results demonstrate that the modulation can be affected by the crack position when it is located at the longitudinal resonance node point x(0) = L min (where L is the beam length, n is the mode number, m is an integer, and 1 < m < n), and both the modulation at the multiples of the number n-order resonance frequencies and peaks on the side of these resonances are much lower. It is also demonstrated that the modulation peak values occur on both sides of the high-frequency structure response resonances. The modulation peak frequencies change with the pumping amplitude; however, the variation is too small to affect the probing frequency selection in our experimental setup. Further analysis suggested that the nonlinear spring model is appropriate for crack simulations to predict modulation distribution and select parameters. The VAM technique exhibited the best crack sensitivity when the probing frequency was selected as the sum or difference of the resonance and pumping frequency. To lock this optimum probing frequency, the minimum sweep range should be twice the beam longitudinal fundamental frequency. Further, the energy dissipation in the crack interface also plays an important role in the modulation and has the potential to be used for crack size estimation with proper excitation. All of the obtained results will aid in understanding the modulation mechanism and making further improvements to the reliability and efficiency of the VAM technique. (C) 2020 Elsevier Ltd. All rights reserved.