ASA 124th Meeting New Orleans 1992 October

4pSA10. Acoustic monitoring of beams by wavelet analysis.

Peeter M. Akerberg

Ben H. Jansen

Pierre Y. Robin

Dept. of Elec. Eng., Univ. of Houston, 4800 Calhoun Rd., Houston, TX 77204-4793

Robert D. Finch

University of Houston, Houston, TX 77204-4793

This talk proposes to use the discrete wavelet transform (DWT) of the acoustic signal produced by a hammer impact, to examine the integrity of metallic beams. The DWT is naturally adapted to transient signals and provides better frequency resolution for low frequencies than Fourier transform-based methods. Time-scale representations, such as the discrete wavelet transform, are related to time-frequency representations, which provide a framework to study variations of the signal frequency content over time. It has been found that the rate at which the amplitudes of the resonant frequencies decay is an important indicator of the integrity of metallic beams [Robin et al., J. Acoust. Soc. Am. 91, 2359(A) (1992)]. The frequency-dependent decays have been investigated more closely with the discrete wavelet transform. The decay appears to ondulate, and the frequency of this ondulation is related to the integrity of the beam. DWTs of actual vibration signals will be presented to show that the ondulation in the decay rates of the resonant frequencies can be used to differentiate between intact and slotted beams. [Work supported by NSF Grant No. MSS-9024224.]