Detection of freeze-crack using digital signal processing

An acoustic method to characterize and quantify freeze-crack was developed. Mozzarella cheese, turkey-ham, cheesecake, and margarine, 38 × 25 × 25 mm (1.5 × 1 × 1 in.), at room temperature were directly immersed in liquid nitrogen and the event was tape recorded. Analog to digital conversion of the data was performed using a sound board and signal processing software (WaveSE). Transformation of the data from the time domain to the frequency domain was performed using Fast Fourier Transformation. Frequency analysis revealed the vaporization of liquid nitrogen generates signals in the 500- to 2,500-Hz frequency range. All four food products evaluated generated signals during freezing in the 0- to 3,500-Hz frequency range with different intensities than liquid nitrogen alone. Sounds associated with a freeze-crack also occurred in this range, but with, higher amplitude. In addition, cracking sounds had frequency components extending beyond 3,500 Hz. Severe cracks had sound signals over the entire range of 0 to 10,000 Hz. The ratio of the area under the frequency distribution (from 3,000 to 10,000 Hz) for a crack compared to that of background noise was at least 2 for small cracks and greater than 30 for severe cracks. This indicates that an on-line, freeze-crack detection method can be developed by targeting signals in the 3,000- to 10,000-Hz frequency range.