石油天然气管道等重要机械结构在长期服役过程中，由于材料原始缺陷、加工过程以及服役环境对结构的影响，使得准确检测缺陷的位置和尺寸对结构完整性评价及寿命预测具有重要的研究意义。直流电位降法在裂纹扩展速率监测方面具有广泛的应用，鉴于其能够监测到微小变化，因此将该方法应用于缺陷检测。本文以重要机械结构常用的304奥氏体不锈钢为研究对象，在室温环境中进行表面缺陷位置的检测和尺寸预测。对缺陷检测仪的测量方法进行了改进，并分析缺陷周围的电压以确定表面缺陷检测方法，同时结合神经网络算法，对缺陷的尺寸进行预测。主要的研究内容如下：

在现有仪器的基础上进行改进，通过添加ADS1256信号采集模块和迪文屏显示界面，实现了静态缺陷的检测并将电压数据显示在界面上；设计点阵式探针结构，实现试样上任意位置的电压信号采集。

利用ABAQUS软件建立了二维平板试样有限元模型，分析了缺陷存在对原电场的影响以及含不同长度和位置的缺陷对电压的影响趋势；根据有限元数据电压结果之间的变化规律，提出合理的缺陷检测准则，并采用有限元的结果对该准则进行初步验证。

建立平板试样的三维有限元模型，将计算结果的多个电压作为神经网络算法的输入特征值，分析试样表面不同位置的电压大小以选择合适的电压测量位置；通过经验公式确定隐含层节点数目，以缺陷的半长和深度作为输出特征值，实现缺陷尺寸的预测。

以含缺陷的304奥氏体不锈钢试样为研究对象，搭建了DCPD缺陷检测试验平台，进行缺陷区域检测和缺陷尺寸预测的试验验证，从真实缺陷区域和识别缺陷区域、缺陷真实尺寸和预测尺寸两个方面判断检测方法的准确性。

The accurate detection of defect location and size in important mechanical pipeline structures such as petroleum and natural gas, which have been in service for a long time, is of significant research significance for the evaluation of structural integrity and life prediction due to the effects of material original defects, processing, and service environment on the structure. The direct current potential drop method has been widely used in monitoring crack propagation rates, and therefore, this method is applied to defect detection due to its ability to detect small changes. This study focuses on the detection of surface defect location and size prediction of commonly used 304 stainless steel in important mechanical structures in room temperature environment. The measurement method of the defect detection device was improved, and the voltage around the defect was analyzed to determine the surface defect detection method. Additionally, the defect size was predicted by combining the neural network algorithm. The main research contents are as follows:

Improvements were made to the existing instrument by adding an ADS1256 signal acquisition module and a Divin screen display interface to achieve the detection of static defects and display voltage data on the interface. A dot matrix probe structure was designed to achieve voltage signal acquisition at any position on the specimen.

A two-dimensional finite element model of the plate specimen was established using ABAQUS software to analyze the influence of defects on the original electric field and the trend of the effect of defects with different lengths and positions on the voltage. Based on the variation law of the finite element data voltage results, a reasonable defect detection criterion was proposed, and the criterion was preliminarily verified using finite element results.

A three-dimensional finite element model of the plate specimen was established. Multiple voltage results of the calculated results were used as the input feature values of the neural network algorithm to analyze the voltage magnitude at different positions on the specimen surface to select appropriate voltage measurement positions. The hidden layer node number was determined by an empirical formula, and the defect size was used as the output feature value to achieve defect size prediction.

The DCPD defect detection test platform was built using 304 austenitic stainless steel specimens with defects as the research object. The experimental verification of defect area detection and defect size prediction was performed. The accuracy of the detection method was judged from the perspectives of real defect area and recognized defect area, as well as real defect size and predicted size.

[1]Yuan J, Mao W J, Hu C, Zheng J F, Zheng D Z, Yang Y B. Leak detection and localization techniques in oil and gas pipeline: A bibliometric and systematic review[J]. Engineering Failure Analysis, 2023, 107060.

[2]Ege Y, Coramik M. A new measurement system using magnetic flux leakage method in pipeline inspection[J]. Measurement, 2018, 123: 163-174.

[3]董绍华, 袁士义, 张来斌, 胡瑾秋, 陈怡玥.长输油气管道安全与完整性管理技术发展战略研究[J]. 石油科学通报, 2022, 7(03): 435-446.

[4]丁建林, 西昕, 张对红. 能源安全战略下中国管道输送技术发展与展望[J].油气储运, 2022, 41(06): 632-639.

[5]黄敏, 周德红, 樊旭岩, 陈锦辉, 李晓柯, 李嘉琦. 基于两种组合赋权法的天然气长输管道安全风险评估比较[J].工业安全与环保, 2023, 49(02): 42-47.

[6]董旭霞, 张东华. 以国脉建设带动产业高质量发展——中俄东线南段打造“发展管道”纪实[J]. 中国石油企业, 2022, No.445(05): 50-53.

[7]梁海波, 王怡. 基于深度学习的天然气钢制管道缺陷检测方法研究[J]. 电子测量与仪器学报, 2022, 36(09): 148-158.

[8]石明江, 陈瑞, 冯林.基于磁记忆的金属管道缺陷检测方法[J]. 电子测量与仪器学报, 2022, 36(01): 44-53.

[9]Vishnuvardhan S., Murthy A. R, Choudhary A. A review on pipeline failures, defects in pipelines and their assessment and fatigue life prediction methods[J]. International Journal of Pressure Vessels and Piping, 2023, 104853.

[10]Zhang M, Guo Y B, Xie Q J, Zhang Y S, Wang D G, Chen J Z. Defect identification for oil and gas pipeline safety based on autonomous deep learning network[J]. Computer Communications, 2022, 195.

[11]Taheri H, Jones C, Taheri M. Assessment and detection of stress corrosion cracking by advanced eddy current array nondestructive testing and material characterization[J]. Journal of Natural Gas Science and Engineering, 2022, 102: 104568.

[12]Xu Y, Jiang X. Nondestructive testing and imaging of corrosion thickness of steel plates using THz-TDS[J]. Infrared Physics & Technology, 2022, 127: 104467.

[13]McIntyre P, Priest A. Measurement of sub-critical flaw growth in stress corrosion, cyclic loading and high temperature creep by the DC electrical resistance technique[R]. 1971.

[14]Nakai Y, Wei R P. Measurement of short crack lengths by an ac potential method[J]. Engineering fracture mechanics, 1989, 32(4): 581-589.

[15]Yan X L, Dong S Y, Xu B S, Cao Y. Progress and challenges of ultrasonic testing for stress in remanufacturing laser cladding coating[J]. Materials, 2018, 11(2): 293.

[16]Hosseini S M, Azadi M, Ghasemi-Ghalebahman A, Jafari S M. Fatigue crack initiation detection in ductile cast iron crankshaft under rotating bending fatigue test using the acoustic emission entropy method[J]. Engineering Failure Analysis, 2023, 144: 106981.

[17]Zu R L, Yang Y, Huang X F, Jiao D C, Zhao J Y, Liu Z W. A stress detection method for metal components based on eddy current thermography[J]. NDT and E International, 2023, 133: 102762.

[18]Zhao J M, Li W, Yuan X-A, Yin X K, Li X, Ding J X, Zhao J C. A novel fatigue crack angle quantitative monitoring method based on rotating alternating current field measurement[J]. Measurement,2022,195: 111101.

[19]Salazar A, Mendioroz A, Oleaga A. Flying spot thermography: Quantitative assessment of thermal diffusivity and crack width[J]. Journal of Applied Physics, 2020, 127(13): 131101.

[20]黄松岭, 彭丽莎, 孙洪宇, 李世松. 航空发动机叶片缺陷无损检测与在线监测技术综述[J/OL]. 测控技术: 1-11[2023-04-24].

[21]Singh Y P, Presby M J, Kannan M, Morscher G N. Multi-lead direct current potential drop method for in situ health monitoring of ceramic matrix composites[J]. Journal of Engineering for Gas Turbines and Power, 2019, 141(3).

[22]Si Y, Rouse J P, Hyde C J. Potential difference methods for measuring crack growth: A review[J]. International Journal of Fatigue, 2020, 136: 105624.

[23]Yoon J Y, Kim Y J, Hwang I S, Ayers L, Short M, Bromberg L, Ballinger R G. On-Line monitoring of cracks in piping weldment using array probed direct current potential drop[C]//Pressure Vessels and Piping Conference. American Society of Mechanical Engineers, 2013, 55690: V005T10A011.

[24]Yoon J Y, Kim Y, Choi S, Nam W C, Hwang I S, Bromberg L, Stahle P W, Ballinger R G. On-line monitoring of environment-assisted cracking in nuclear piping using array probe direct current potential drop[J]. Journal of Nondestructive Evaluation, 2016, 35: 1-11.

[25]Na S, Yoon D, Kim J, Kim H, Kim D. An evaluation of the fatigue crack propagation rate for powder metallurgical nickel-based superalloys using the DCPD method at elevated temperatures[J]. International Journal of Fatigue,2017,101: 27-35.

[26]Corcoran J, Nagy P B, Cawley P. Monitoring creep damage at a weld using a potential drop technique[J]. International Journal of Pressure Vessels and Piping, 2017, 153: 15-25.

[27]Campagnolo A, Bär J, Meneghetti G. Analysis of crack geometry and location in notched bars by means of a three-probe potential drop technique[J]. International Journal of Fatigue, 2019, 124: 167-187.

[28]Bär J. Crack detection and crack length measurement with the DC potential drop method–possibilities, challenges and new developments[J]. Applied Sciences, 2020, 10(23): 8559.

[29]Meneghetti G., Vecchiato L., Campagnolo A., Rech P., Cova M.. Numerical calibration of the direct current potential drop (DCPD) method in fracture mechanics fatigue tests[J]. Procedia Structural Integrity,2020,28.

[30]Xue H, Xu P C, Ni C Q, Wei Q S, Cui Y H, Zhang L, Bashir R. A Monitoring System for Crack Growth Rate Based on Direct Current Potential Drop[C]//2018 International Conference on Sensor Networks and Signal Processing (SNSP). IEEE, 2018: 317-321.

[31]薛河, 苟思育, 倪陈强, 张亮, 赵有俊. 基于直流电位降法的裂纹监测仪器系统设计[J]. 传感器与微系统, 2020, 39(11): 85-88.

[32]苟思育. 基于DCPD法的裂纹扩展监测仪系统开发与性能优化研究[D]. 西安：西安科技大学, 2021.

[33]薛河, 王双, 赵有俊, 苟思育, 倪陈强. 基于DCPD的CT试样裂纹扩展数值模拟标定方法[J]. 中国测试, 2023, 49(01): 36-42.

[34]薛河, 赵有俊, 王双, 倪陈强, 孙裕满, 王欣玥. 裂纹监测仪恒流源系统设计[J]. 传感器与微系统, 2023, 42(04): 91-94.

[35]Harrington D S, Bell R, Tan C L. Calibration of the localized DCPD method for crack shape measurement using the boundary element method[J]. Fatigue & Fracture of Engineering Materials & Structures, 1995, 18(7-8): 875-884.

[36]解社娟, 陈振茂, 张东利, 耿强. 基于直流电位法的泡沫金属定量无损检测方法[J].无损检测, 2008(08): 536-539.

[37]Tada N, Okada M, Iwamoto J. Three-dimensional identification of semi-elliptical surface crack by means of direct-current electrical potential difference method with multiple-probe sensor[J]. 2007, 129(3).

[38]Tada N, Funakoshi A. Three-dimensional identification of semi-elliptical crack on the back surface by means of direct-current electrical potential difference method with multiple-probe sensor (PVP2006-ICPVT-11-93359)[J]. Journal of pressure vessel technology, 2009, 131(2).

[39]Tada N, Uchida M. Electric field analysis of simultaneous evaluation of crack on an inner pipe surface and pipe wall thickness using direct-current potential difference method of multiple-probe type[J]. Journal of pressure vessel technology, 2012, 134(4).

[40]张静, 解社娟, 陈振茂. 泡沫金属直流电位法检测技术[J]. 无损检测, 2010, 32(08): 612-615.

[41]Wang X J, Xie S J, Wang X W, Chen Z M. Sizing of Cavity Defect in Metallic Foam from DC Potential Drop Signals with Stochastic Inversion Methods[C]//.Proceedings of 2013 Far East Forum on Nondestructive Evaluation/Testing: New Technology & Application. Institute of Electrical and Electronics Engineers,2013:70-73.

[42]Wang X J, Cai W L, Chen Z M. NDT of Cavity Defect in Metallic Plate Using dc Potential Drop Method[J]. Journal of Testing and Evaluation,2018,46(2).

[43]代占鑫, 胡博, 余业山, 刘怡. 直流电位法对Q235钢材表面裂纹的检测仿真[J]. 表面技术, 2018, 47(01): 236-241.

[44]Funk M, Bär J. DCPD based detection of the transition from short to long crack propagation in fatigue experiments on the aluminum alloy 7475 T761[J]. Procedia Structural Integrity, 2019, 17: 183-189.

[45]Hicks M A, Pickard A C. A comparison of theoretical and experimental methods of calibrating the electrical potential drop technique for crack length determination[J]. International Journal of Fracture, 1982, 20(2): 91-101.

[46]路明旭, 肖欣然, 姜伟之. 一种开裂紧凑试样的电位分析[J]. 北京航空学院学报, 1985(01): 105-112.

[47]Ikeda K, Yoshimi M, Miki C. Electrical potential drop method for evaluating crack depth[J]. International Journal of Fracture, 1991, 47: 25-38.

[48]Merah N. Detecting and measuring flaws using electric potential techniques[J]. Journal of Quality in Maintenance Engineering, 2003, 9(2): 160-175.

[49]王亮, 丁传富. 高温下直流电位法测量裂纹长度的应用[J]. 理化检验：物理分册, 2006, 42(9): 3.

[50]丁传富, 王亮, 刘建中. 直流电位法自动检测高温疲劳裂纹长度的研究及应用[J]. 实验室研究与探索, 2007(10): 275-277.

[51]Spitas V, Spitas C, Michelis P. A three-point electrical potential difference method for in situ monitoring of propagating mixed-mode cracks at high temperature[J]. Measurement, 2010, 43(7): 950-959.

[52]Zhang D L, Chen Z M, Xu M L, Lu T J. Quantitative NDE of welding defects in lattice sandwich metallicplate by using DC potential drop method[J]. International Journal of Applied Electromagnetics andMechanics,2010,33(3-4): 1109-1117.

[53]徐蒙, 李培源. 直流电位法在高温合金疲劳裂纹测量中的应用[C]// 北京力学会第二十三届学术年会会议论文集. 2017.

[54]Hayashi M. Determination of Surface Crack Configuration by DC Potential Drop Method Assisted by Three-Dimensional Boundary Fitted Coordinates Transformation Technique[J]. Journal of the Society of Materials Science, Japan, 2019, 68(10): 791-797.

[55]Meneghetti G., Vecchiato L., Campagnolo A., Rech P., Cova M. Numerical calibration of the direct current potential drop (DCPD) method in fracture mechanics fatigue tests[J]. Procedia Structural Integrity, 2020, 28: 1536-1550.

[56]Vecchiato L, Campagnolo A, Meneghetti G. Numerical calibration and experimental validation of the direct current potential drop (DCPD) method for fracture mechanics fatigue testing of single-edge-crack round bars[J]. International Journal of Fatigue, 2021, 150: 106316.

[57]Sun D X, Li H, Feng H, Li Y, Li Q. Calibrating Johnson's formula for applying DCPD method to an axial through-wall crack in a pipe[J]. Engineering Fracture Mechanics, 2021, 242: 107461.

[58]Cai W L, Xie S J, Jomdecha C, Wang X J, Pei C X, Li Y, Chen Z M, Yusa N. Assessment of local conductivity distribution in stress corrosion crack region using direct current potential drop method[J]. Corrosion Science, 2017, 123: 197-208.

[59]孙沁梅, 卢益民. 高精度模数转换器ADS1256的原理和应用[J]. 计算机与数字工程, 2005(03): 94-97.

[60]周超, 李春茂. 新型8通道24位Δ-Σ型模数转换器ADS1256的原理及应用[J]. 国外电子元器件, 2005(06): 35-38.

[61]贾畅鹏. 固态继电器的特性及应用[J]. 煤炭技术, 2007, (01): 33-35.

[62]周海涛, 董全林, 周卫宁. ADS1256在多路高精度加速度计数据采集中的应用[J]. 航空电子技术, 2009, 40(04): 15-18.

[63]薛河, 赵有俊, 王双, 孙裕满, 王欣玥, 张建龙. 疲劳裂纹扩展电位信号的最优光滑降噪算法[J]. 电子测量与仪器学报, 2022, 36(07): 115-124.

[64]薛河, 苟思育, 倪陈强. 紧凑拉伸试样参比电位差接线点布局的优化方法[J]. 电子学报, 2020, 48(07): 1396-1402.

[65]倪陈强, 薛河, 崔英浩. 直流电位降法裂纹监测试验中接线点位置分析[J]. 热加工工艺, 2016, 45(24): 122-125.

[66]Chen W H, Chen J S, Fang H L. A theoretical procedure for detection of simulated cracks in a pipe by the direct current–potential drop technique[J]. Nuclear Engineering and Design, 2002, 216(1-3): 203-211.

[67]林启权, 卜根, 王镇柱, 董文正, 王凯. 基于BP神经网络的DP980钢板本构关系[J]. 塑性工程学报, 2022, 29(08): 139-144.

[68]苟建军, 王森,王健,范利锋,张文清, 田晓东. 基于BP神经网络的金属薄板包辛格效应预测[J]. 塑性工程学报, 2022, 29(08): 152-157.

[69]苑希民, 黄玉啟, 田福昌, 曹鲁赣. 基于LSTM-GM神经网络模型的风暴潮增水预报方法研究[J/OL]. 水资源保护: 1-12[2023-03-10]

[70]徐元中, 曹翰林, 吴铁洲. 基于SA-BP神经网络算法的电池SOH预测[J]. 电源技术, 2020, 44(03): 341-345.

[71]师洪涛, 杨静玲, 丁茂生, 王金梅. 基于小波—BP神经网络的短期风电功率预测方法[J]. 电力系统自动化, 2011, 35(16): 44-48.

[72]马健, 滕召胜, 邱伟, 马丽莎, 刘颉. 基于IGA-BP神经网络的智能电能计量设备退化趋势研究[J]. 仪器仪表学报, 2021, 42(11): 123-131.

[73]Wang S, Wang B, Janin Y J, Bourga R, Xue H. Effects of the surface crack shape on J values along the front of an elliptical crack[J]. Fatigue & Fracture of Engineering Materials & Structures, 2021, 44(11): 2944-2961.

[74]张立明. 人工神经网络的模型及其应用[M]. 上海: 复旦大学出版社, 1994.

[75]高大启. 有教师的线性基本函数前向三层神经网络结构研究[J]. 计算机学报, 1998(01): 80-86.