Reconstruction Of The Defect Images In Pipes By Common-source Synthetic Focusing Of Guided Waves

Pipelines are widely used in long distance transportation, infrastructures and industrial production. Large quantities of pipelines in service are operating under severe environment, and any leakage could cause great damages. Among the existing non-destructive testing technologies for pipe inspection, the guided-wave-based technologies characterize in large-range inspection,full-coverage and high-efficiency. But the processing of guided waves signals is difficult because of the dispersive characteristic of multiple guided wave modes. The common-source synthetic focusing method separates the guided wave modes by circumferential order, and compensates the dispersion of each mode individually, which is an efficient guided wave testing method for pipe inspection. In real testing of a pipeline system, the size range and specification of pipes are various, and the amount of pipes to be inspected is huge. Under the circumstances,the applicability, resolving ability and usability of common-source synthetic focusing method remain to be improved.Firstly,the existing common-source synthetic focusing method is not applicable to the detection of oblique defects, because the reflected guided wave signals are interfered by cross-mode-family mode conversion. Accordingly, this thesis proposes a solution, which is to separate the guided wave echo signals by mode-family in advance. Secondly, the axial resolving ability of existing common-source synthetic focusing method is limited by the time-domain width of the guided wave signal. Accordingly, the deconvolution method is introduced to improve the time-domain resolving ability. Thirdly,existing guided wave testing technologies relies on the prior knowledge of pipe parameters, which hinders the reliability and usability in massive testing work. Accordingly,this thesis proposes a inversing method to obtain the pipe parameters from the guided wave signals. By combining the pipe parameter inversing method,mode-family separation technology and deconvolution method together, a highly sensitive,efficient and widely usable guided wave imaging technology, which does not rely on the prior knowledge of pipe parameters, is developed. A2-D scanning system is designed and constructed for the collection of guided wave signals in a cylindrical area, and is used for experimental verification of the above mentioned technology.This thesis is composed of six chapters:Chapter I: The major non-destructive technologies for pipe inspection are introduced and compared, which gives the conclusion that guided-wave-based pipe testing technology excels conventional non-destructive testing technologies in coverage and efficiency. The history and status of pipe guided wave theory development are briefly summarized. The technologies for the stimulation and receiving of guided waves in pipes are introduced. The merits and demerits of guided wave tomography, active focusing and synthetic focusing for pipe inspection are discussed. After the above introduction, the significance and content of the research in this thesis are demonstrated.Chapter II: A mode-family separation method is proposed to eliminate the interference from cross-mode-family mode conversion. The guided wave array signals are transformed into the 3-D domain of circumferential order, time frequency and axial space frequency. Based on that, a filter that works in the frequency-wavenumber domain is designed for mode separation and unidirectional extraction. The mode conversion in the reflection from oblique interfaces is quantitatively studied based on numerical simulation results. The effectiveness of mode separation and unidirectional extraction is verified by the imaging of oblique defects and semi-circle defects. The filtering process purifies the signal components, and the fake defect patterns that are caused by cross-mode-family mode conversion and secondary reflection are eliminated. The proposed technique improves the applicability of common-source synthetic focusing method to the inspection of pipe defects of various orientations.Chapter III: A revised synthetic focusing method is developed to enhance the axial resolving ability by introducing the deconvolution method. By using the circumferential-order separation and integration after phase shifting, the incident wave and the reflected wave at the defect region are calculated. With the calculated incident wave taken as the reference, a deconvolution process is applied to the calculated reflected wave, which extracts the reflection coefficient and generates defect image. The signal-noise ratio of the reconstructed image is improved by introducing a coherent noise inhibitor. A tiny defect is successfully imaged by experiment, and circumferential linear defects are quantitatively detected. Additionally, the detectability of the developed technique to area defects is researched by numerical simulation and experiment. The proposed deconvolution-based imaging method excels conventional integral method in both axial resolution and SNR.Chapter IV: The pipe parameters are inversely identified from the spectrum peaks of the guided wave array signals. In order to improve the efficiency and convergence of the inversing,the analogy between the axisymmetric pipe guided wave modes and the plate guided wave modes is utilized to simplify the optimal function. Additionally, the hypothesis of thin pipe wall is applied to reduce the dimension obtain the analytical solution of the problem. Based on the numerical simulation results, the feasibility of the proposed inversing technique is verified. The effects of axial range of the scanned region, axial scanning step and the positioning error to the inversely determined results are analyzed.Chapter V: By combining the pipe parameter inversing method, mode-family separation technology and deconvolution method together, a highly sensitive, efficient and widely usable guided wave imaging technology, which does not rely on the prior knowledge of pipe parameters, is developed. To verify the feasibility of this technique by experiment, a 2-D scanning system is designed and constructed for the collection of guided wave signals in pipes.The collected guided wave data are utilized for the inverse identification of the pipe parameters,and the imaging of defects with oblique interfaces. The redundancy of the collected guided wave signal is analyzed. The minimal axial scanning range and the lowest axial scanning density are discussed to provide guidance for real testing.Chapter VI: The research contents of this thesis are summarized, the innovation points are highlighted, and some future works are discussed.