Academic literature on the topic 'Zero velocity curves'

The properties of ultrasonic Lamb waves, such as relatively small attenuation and high sensitivity to structural changes of the object being investigated, allow performing of non-destructive testing of various elongated structures like pipes, cables, etc. Due to the dispersion effect of Lamb waves, a waveform of the received informative signal is usually distorted, elongated and overlapping in the time domain. Therefore, in order to investigate objects using the ultrasonic Lamb waves and to reconstruct the dispersion curves, it is necessary to know the relationship between frequency, phase and group velocities and thickness of the plate. The zero-crossing technique for measurement of phase velocity of Lamb waves (the A0 and S0 modes) has been investigated using modelled dispersed signals and experimental signals obtained for an aluminium plate having thickness of 2 mm. A comparison between two reconstruction methods of Lamb wave phase velocity dispersion curves, namely, the two-dimensional fast Fourier transform (2D-FFT) and zero-crossing technique, along with the theoretical (analytical) dispersion curves is presented. The results indicate that the proposed zero-crossing method is suitable for use in reconstruction of dispersion curves in the regions affected by strong dispersion, especially for the A0 mode.

Common-midpoint (CMP) sorting of pure-mode data in arbitrarily complex isotropic or anisotropic media leads to moveout curves that are symmetric around zero offset. This greatly simplifies velocity determination of pure-mode data. Common-asymptotic-conversion-point (CACP) sorting of converted-wave data, on the other hand, only centers the apexes of all traveltimes around zero offset in arbitrarily complex but isotropic media with a constant P-wave/S-wave velocity ratio everywhere. A depth-varying CACP sorting may therefore be required to position all traveltimes properly around zero offset in structurally complex areas. Moreover, converted-wave moveout is nearly always asymmetric and nonhyperbolic. Thus, positive and negative offsets need to be processed independently in a 2D line, and 3D data volumes are to be divided in common azimuth gathers. All of these factors tend to complicate converted-wave velocity analysis significantly.