System and method for mode identification and extraction in guided wave signals and nondestructive methods for automatically sensing, locating, and/or characterizing problems in structures
View Patent ↗A system and method are provided herein for determining a characteristic of a guided wave traveling in a structure. The method includes comparing theoretical dispersion curves for the structure and a wavelet transform image of the guided wave to identify a match, which in an embodiment is a cross-correlation. In an exemplary embodiment, the wavelet transform image is a three-dimensional image, and the theoretical dispersion curves are calculated using Lamb mode theory or finite element analysis.
1 . A method comprising: detecting a signal propagating in a structure using a sensor positioned on the structure, the signal emanating from and/or interacting with a defect in the structure; converting a waveform of the signal to a wavelet transform image;
accessing a plurality of theoretical dispersion curve images associated with the structure;
automatically comparing the wavelet transform image with the theoretical dispersion curve images; and wherein the comparing comprises applying an artificial intelligence model trained with the theoretical dispersion curves.
2 . A method comprising: detecting a signal propagating in a structure using a sensor positioned on the structure, the signal emanating from and/or interacting with a defect in the structure; converting a waveform of the signal to a wavelet transform image;
accessing a plurality of theoretical dispersion curve images associated with the structure;
automatically comparing the wavelet transform image with the theoretical dispersion curve images;
wherein the comparing comprises cross-correlation;
wherein the cross-correlation comprises, for each of a plurality of the theoretical dispersion curves, multiplying matching points of the theoretical dispersion curve and the wavelet transform to calculate products, and adding the products together to provide a sum associated with the theoretical dispersion curve; and
said method further comprising analyzing the respective sums of the plurality of theoretical dispersion curves as a function of time to identify a wave arrival time relative to a sensor.
3 . A method comprising: detecting a signal propagating in a structure using a sensor positioned on the structure, the signal emanating from and/or interacting with a defect in the structure; converting a waveform of the signal to a wavelet transform image;
accessing a plurality of theoretical dispersion curve images associated with the structure;
automatically comparing the wavelet transform image with the theoretical dispersion curve images;
wherein the comparing comprises cross-correlation;
wherein the cross-correlation comprises, for each of a plurality of the theoretical dispersion curves, multiplying matching points of the theoretical dispersion curve and the wavelet transform to calculate products, and adding the products together to provide a sum associated with the theoretical dispersion curve; and
said method further comprising analyzing the respective sums of the plurality of theoretical dispersion curves as a function of time to identify a wave arrival time relative to a sensor; and
further comprising graphically representing the results of the cross-correlation and selecting an initial peak as a wave arrival time relative to a sensor.
4 . A method comprising: detecting a signal propagating in a structure using a sensor positioned on the structure, the signal emanating from and/or interacting with a defect in the structure; converting a waveform of the signal to a wavelet transform image;
accessing a plurality of theoretical dispersion curve images associated with the structure;
automatically comparing the wavelet transform image with the theoretical dispersion curve images; and including use of individual mode correlations to sort out direct arrivals from reflections and overcome reflections.
5 . A method comprising: detecting a signal propagating in a structure using a sensor positioned on the structure, the signal emanating from and/or interacting with a defect in the structure; converting a waveform of the signal to a wavelet transform image;
accessing a plurality of theoretical dispersion curve images associated with the structure;
automatically comparing the wavelet transform image with the theoretical dispersion curve images; and
using group velocity inherent in the wavelet transform image analysis for the arrival time difference calculations in a sensor array analysis to overcome the effects of phase shifts due to the dispersive wave propagation in plate-like structures.
6 . A method comprising: detecting a signal propagating in a structure using a sensor positioned on the structure, the signal emanating from and/or interacting with a defect in the structure; converting a waveform of the signal to a wavelet transform image;
accessing a plurality of theoretical dispersion curve images associated with the structure;
automatically comparing the wavelet transform image with the theoretical dispersion curve images; and
determining flaws in plates by correlating different sets of dispersion curves calculated to include various flaws, such as plate thinning and cracking, and comparing the theoretical dispersion curves to the wavelet transform of the unknown signal to determine the flaw type and criticality.