Research On Inner Wall Defect Detection Technology For Marine Pipelines Based On ACFM

With the implementation of the marine power strategy,the development of offshore oil and gas has been improved continuously,and various marine engineering equipment has been developed.The offshore oil and gas industry is flourishing,and it has been promoted from the shallow sea to the far sea gradually.Marine pipelines have many advantages,such as tensile resistance,pressure resistance,fatigue resistance and so on.Marine pipelines are used to transport oil and gas,and it plays a key role in deep-water oil and gas exploration.Marine pipelines have served in the harsh seawater environment for a long time.Under the comprehensive function of various loads,the marine pipelines can be easily introduced cracks and corrosion defects due to the corrosion and impact.If defects in marine pipelines are not detected in time,false detections and missed detections can bring safety risks and even lead to safety accidents.In this dissertation,the identification,size inversion,lift effect compensation for cracks,and visualization for volume defects in the inner surface of marine pipeline are studied based on alternating current field measurement(ACFM)technique.The detection methods are determined,and the detection system is developed,providing the techniques and foundations for the intelligent identification and safety assessment of inner wall defects in the marine pipelines.The main research contents of this dissertation are as follows:(1)Based on the conventional theoretical model of ACFM,the ACFM equivalent model for defect detection is proposed,the width parameter of defect is introduced into the equivalent model,the theoretical basis of magnetic imaging for defect is provided,the ACFM theoretical model for defect detection is further supplemented.The finite element model of inner wall defect in marine pipeline for ACFM is established.The distribution law of the alternating magnetic field around the defect is studied.The optimal excitation frequency of ACFM is determined.The effect of medium and concrete layer on the ACFM signal of defect is analyzed.These models and simulations lay the theoretical foundation for the detection of inner wall defects in marine pipeline using ACFM technique.(2)The effect of crack size on the biaxial ACFM signal is studied and the characteristics of uniaxial ACFM signal for crack characterization is determined.The crack size inversion methods based on the uniaxial ACFM signal are proposed.Aiming at the shortcoming of poor robustness for the interference in traditional ACFM defect inversion method,an extreme learning machine(ELM)neural network is proposed to obtain the crack dimensions,which provides the technical support for the size evaluation of crack.(3)The effect of lift-off distance on size inversion for defect detection is analyzed.According to Biot-Savart law,a mathematical model of lift-off effect compensation is abstracted,and the lift-off effect compensation algorithm is proposed based on the minimum value searching method,which can evaluate the inner crack size in marine pipeline when the lift-off distance is varied.(4)The effect of defect width on ACFM signal is studied,the characteristics of uniaxial ACFM signal are determined for the width representation of the defect,the quantification method for the width of defect is developed.A magnetic imaging algorithm for visualization of the inner volume defect of marine pipeline is proposed based on the relationship between the defect size and the ACFM axial signal characteristics.(5)The ACFM detection probe is developed based on a pair of tunneling magnetoresistance(TMR)sensors.The multi-parameter coupling coefficient is used to evaluate the denoising effect of EMD combined with wavelet threshold denoising method on the ACFM signal of defect,the denoising scheme for the ACFM signal of defect is determined.An inner defect detection system for marine pipeline including ACFM detection algorithm is constructed.The artificial defects in the marine pipeline are detected using the proposed defect detection system.Experimental results show that the crack size can be qualitatively characterized using the uniaxial ACFM method and the ELM neural network.The lift-off effect can be compensated by the proposed algorithm.Finally,the visualization for the volume defect is achieved using the magnetic imaging algorithm.