Development Of Guided Wave Detection Analysis Algorithm&Software And Its Application In Typical Structure

Ultrasonic guided wave has advantage of on-line detection as a kind ofnondestructive testing technology. Guided wave has concentrated energy and smallattenuation characteristic along the propagation path, can carry the detectioninformation of structure completely. The issue uses guided wave technique to detecttypical structure and do research on the characterization of characteristic of defectswithin structure, and the main research works are given as follows.(1) Guided wave testing and analysis software based on guided wave excitationand receiving instrument is developed. The software can transmit and storage theacquisited guided wave datas quickly using USB interface, wavelet noise reductions、FFT、signal envelope and short-time fourier transform functions integrated into thesoftware can effectively improve guided wave signal to noise ratio, analysis frequencycomponents of guided wave, and accurately extract wave packet spikes of guidedwave. Guided wave testing analysis software realizes on-line detection and analysisbased on guided wave inspection technique.(2) Typical structure such as a specific size steel rods and pipes are detectedusing guided wave excitation and receiving instrument and the guided wave testingand analysis software, the frequency sweep combined with STFT is used to analysisthe modal of detected signals and discriminat guided wave mode that can propagate intypical structure such as steel rods and pipes. The results laid the foundation for thefurther analysis of defect identification and guided wave imaging.(3) A algorithm used to recognize the size and other parameters of crack withinthe steel road is proposed, which is based on BP neural network, the software ofAbaqus is used to build two different steel models with circumferential crack andoblique crack, L(0,1) mode guided wave echo signals are used to validat defectrecognition algorithm, which are collected under different crack size parameters. Theresults show that, the algorithm can distinguish the crack size and other parameters ofcircumferential crack and oblique crack in simulated steel rods very well.(4) Axisymmetric longitudinal mode wave guide is excited at235kHz to detect11mm radius steel rods with circumferential cracks for experimental study. Theresults show that, the mode of guided wave that propogates in steel rods is mainlyL(0,2) mode at235kHz. Defect recognition algorithm for guided wave signal is usedto identify different size and position of circumferential cracks in11mm radius steelrods, the recognition correct rate of algorithm for circumferential cracks is87%，thatis the algorithm can identify circumferential cracks in steel rods very well in practicalapplication; Axisymmetric longitudinal mode guided wave is excited at235kHz to detect11mm radius steel rods with oblique cracks, the same algorithm is used toidentify different size parameters of oblique cracks. Results show that, the recognitioncorrect rate of algorithm for oblique cracks is94.4%, that is the algorithm can identifyoblique cracks in steel rods very well in practical application. Defect recognitionalgorithm provides a new method for the use of guided wave technique to analysisdefect size.(5) Common source guided wave imaging technique is used to locate the axialposition and circumferential position of penetrating circumferential crack in steel pipeof specific size, and reconstruct images for different angles of penetratingcircumferential crack in steel pipe. The results show that, the guided wave imagingused by the issue can accurately locate the circumferential crack in steel pipe, and thelength of imaging display area of transparent circumferential crack in reconstructedimage has small difference with the actual circumferential length of crack. Guidedwave imaging provides a new train of thought for locating circumferential crack anddetecting circumferential length of crack in pipe structure.