摘要
无损检测技术对现代工业、制造业都非常有意义,各种无损检测技术都具有各自的优缺点,多种检测技术复合可以扬长避短。论文以“基于超声和涡流的复合式检测技术”为题,系统地研究超声与涡流复合机理、超声与涡流协同作用、超声涡流集成探头设计、基于D-S证据理论实现数据融合技术,对促进仪器科学技术、制造科学技术具有重要意义。研究工作得到广东省自然科学基金(200805611085)和广东省工业攻关项目(2006B12401001)的资助。
论文阐述了复合式无损检测及数据融合技术在复合式无损检测技术中应用的国内外研究进展,对超声与涡流复合式检测机理、复合式无损检测数据融合理论等关键技术进行研究,提出对铁磁性产品进行全面检测的基于超声和涡流的复合式无损检测技术,设计集超声与涡流检测技术一体的复合式集成探头,结合数据融合手段,实现了复合式无损检测真正意义的复合检测技术。主要工作包括:
①系统研究超声与涡流复合检测机理理论体系,所研究结论对于解决超声与涡流集成问题具有通用指导价值。打破现有单个无损检测技术不能完成对检测试件全面检测的局面,超声涡流两种普遍检测技术的结合,满足了大多数工业生产需求,具有普遍意义;
②提出将平底超声直探头与涡流探头集成于一体的新型探头,打破了传统的简单物理层面上的技术复合,实现了超声和涡流真正意义上的技术复合,节约了制作成本、使用成本、检测成本,同时提高了检测速度、精度、准确度;
③探索性地将超声与涡流检测参数进行协同分析,达到参数设置最优化,保证检测要求情况下,降低系统技术要求,实现成本降低的目的;
④细分检测区域,分析各区域特征,针对不同的区域采用相应信号处理方法和缺陷模式识别法,在实现全面检测同时,提高检测速度、降低信号处理难度、节约成本;并将Manhattan距离法与D-S证据理论结合应用于构造超声与涡流置信度分配函数;
⑤创造性地在Manhattan距离法中引入修正系数k。解决了当_(ij)小于1时,求出的置信度将大于1的矛盾。因此,本作者首次引入了修正系数k ,当发生_(ij)小于1的情况时,依然能得到p_(ij)的归一化出来结果。实验数据分析结果显示该方法的有效性。
选用45#钢的焊缝作为检测对象,研究本课题基于超声与涡流复合式检测技术的有效性和可行性。焊缝板的超声涡流复合式检测实验表明,本课题提出的超声与涡流独立模式识别方法对超声与涡流独立检测区域效果很好,D-S证据理论在重叠区域的应用,一是可以降低{不明}的置信度;二是提高超声检测对于裂纹与未焊合的识别率。
With the title Technique of Complex NDT Based on Ultrasonic & Eddy Current Testing, the thesis systematically studies fundamental theory of ultrasonic and eddy current complex testing. It include synergy of ultrasonic and eddy current testing, integrated probe designing which include ultrasonic and eddy current sensor, and the fundamental and key technology of data fusion based on D-S theory, et al. Complex NDT technique is developed, which has important academic value and practical significance in improving the technological development and application of NDT technology. This work is funded by the National Natural Science Foundation of China (No. 50805053), The Natural Science Foundation of Guangdong province (200805611085), Guangdong province key project (2006B12401001).
This thesis firstly analyzes the domestic and international research on technique of complex NDT and the applications of data fusion technology in complex NDT field. It also researches the key theory of data fusion and complex NDT technology based on ultrasonic and eddy current testing. Put forward the complex NDT technique based on ultrasonic and eddy current for ferromagnetic products. A integrated probe which include ultrasonic and eddy current sensor, combine with data fusion technique are promoted to a real sense of complex NDT technique. The main work includes the following parts:
①Study theory of complex NDT technique based on ultrasonic and eddy current systematically. The conclusions have general guiding value for solving the problem of integrating ultrasonic and eddy current testing. It solves the problem of existing NDT techniques can not provide a comprehensive detection to test specimen. Combining ultrasonic and eddy current the two popular NDT techniques meet the needs of most industrial production which has universal significance.
②It creatively develops integrated probe of ultrasonic and eddy current probe, breaking the traditional complex techniques which simply compound several testing techniques on physical aspects. It also reduces the production costs, using costs and testing cost, while increasing the detection speed and accuracy.
③Firstly synergy analyzing ultrasonic and eddy current testing parameters to get optimization parameters to promote to reduce technical requirements which will reduce the cost meanwhile meet the detecting requirements.
④It creatively divides detection area into three parts. According to each region character, it uses corresponding signal processing methods and pattern recognition methods to promote to extended test for workpiece meanwhile increasing the detection speed, easier the signal processing, saving cost.
⑤It creatively combines Manhattan distance method with D-S evidence theory applying to construct ultrasonic and eddy current sensors confidence distribution function. Whit this method, we successfully obtained a variety of defects confidence distribution with ultrasonic and eddy current complex technique. It creatively introduce correction factor to Manhattan distance method, which solve the problem of when the value of d_(ij) less than 1. So, this thesis firstly using this correction factor k , when d_(ij) less than 1, we can still get the reasonable p_(ij) value. The experimental results show the effectiveness of this method.
Using 45# steel weld as test object to study the effective and feasible of this complex NDT technique based on ultrasonic and eddy current testing.
The experiments show that ultrasonic and eddy current pattern recognition methods using in each independent detecting regions well and the D-S evidence theory using in overlapping region, first can reduce confidence value of {unknown}, then, can improve the accuracy rating of ultrasonic testing with crack and lack of penetration defects.
引文
[1]谢宝忠.材料复合式无损检测及其信号处理系统[D].广州:华南理工大学, 2007
[2]亢彦军.超声TOFD法焊缝探伤系统的研究及开发[D].天津:天津大学, 2007
[3]戴光智.基于超声成像无损检测系统的超分别率重建技术研究[D].广州.华南理工大学, 2008
[4]金长善.超声工程[M].哈尔滨:哈尔滨工业大学出版社, 1989, 105-125
[5] Lee H, Wade G. Modern Acousticial Imaging[M]. New York:IEEE Press, 1986:232-234
[6]苏剑彬.内窥涡流集成化无损检测系统研究与实现[D].长沙:国防科技大学. 2008
[7]Lamarre, A., Dupuis, O., Moles, M. Complete inspection of friction stir welds in aluminium usingultrasonic and eddy current arrays [J]. CINDE Journal, 2006, Volume: 27 Issue: 4 Pages: 14-22
[8] R.S. Edwards, A. Sophian, S. Dixon, et al. EMAT and Eddy Current Dual Probe for DetectingSurface and Near-surface Defects. Review of Quantitative Nondestructive Evaluation Vol. 25,2006 American Institute of Physics: 1515～1522.
[9]邓勇.不确定性信息的融合及其应用研究.上海:上海交通大学博士学位论文, 2003, pp.158
[10]陈积懋,张颖.复合材料及其构件的新型综合无损检测技术[J].无损检测, 2002, 24(6):253-271
[11]陈秋南,张永兴.桥梁桩基础缺陷复合检测及其加固新方法[J].岩石力学与工程学报, 2004,23(20): 3518-3522
[12]张丽琴,钟志民,李劲松等.燃料组件控制棒包壳在役超声涡流自动检测[J].无损检测, 2005,27(1): 38-40
[13]孙金立,肖兴明,袁英民等.基于虚拟仪器的综合式无损检测系统[J].无损检测, 2005, 27(11):580-582
[14]罗兴盛,谷林涛,钱觉时.“超声法”与“拔出法”综合检测结构混泥土的质量[J].国外建材科技,2007, 28(5):56-59
[15]丁训慎.核电站蒸汽发生器传热管的降质及其无损检测技术[J].无损检测, 2008, 30(1):30-33
[16]范智勇.超声-涡流一体化NDE成象系统的研究[D].北京:北方交通大学,1998
[17]周正干,黄凤英,于芳芳.电磁超声和涡流组合检测方法[J].无损检测. 2009, 31(9)：706-710
[18] Renken,C.J., selner,I.H. Refractory metal tubing inspection using ultrasonic and pulsed eddy-currentmethods. Materials Evaluation, 1966, 5(4):257
[19]Song YW, Udpa SS. A new morphological algorithm for fusing ultrasonic and eddy current images[C]. IEEE, ultrasonics symposium, 1996, Pages: 649-652
[20] Degtyar AD, Pearson LH. Ultrasonic and eddy current characterization of impact damage ingraphite/epoxy rocket motor cases [C]. AIP CONFERENCE PROCEEDINGS, 2000, Volume:509 Pages: 1255-1262
[21] Moysan J, Corneloup G, Sollier T. A general thresholding method for NDT images: the case of eddycurrent and ultrasonic images [J]. MATERIALS EVALUATION，2001, Volume: 59 Issue:8Pages: 955-960
[22] Roge, B. Giguere, J.S.R., McRae, K.I., Fahr, A. Nondestructive evaluation of a cermet coating using ultrasonic and eddy current techniques [J]. AIP Conference Proceedings, 2002, Volume: no.615B Issue: 615B Pages: 1201-8
[23] Thomas H-M , Heckel, T., Hanspach, G. Advantage of a combined ultrasonic and eddy current examination for railway inspection trains. INSIGHT Volume: 49 Issue: 6 Pages: 341-344 Published: JUN 2007
[24] S.J.Peri. "Approaches to multisensor data fusion,"Johns Hopkins APL Technical Digest (Applied Physics Laboratory), vol.22,pp. 624-633,2001
[25]许军,罗飞路,张耀辉.多传感器信息融合技术在无损检测中的应用研究[J].无损检测, 2000: 22(8): 342～344
[26]张立东,吴建军.信息融合技术在无损检测中的应用(J).兵工自动化. 2006.7(25):73,81
[27]何友,王国宏,陆大金.多传感器信息融合及应用[M].北京:电子工业出版社, 2000
[28]潘建国.海底管道超声信号检测与数据存储系统的研制[D].上海交通大学, 2009
[29]《超声波探伤》编写组.超声波探伤.电力工业出版社.1980.11:43~150
[30]罗雄彪.基于小波理论和神经网络的超声检测信号分析与识别[D].广州:华南理工大学, 2005.5
[31]魏志刚.基于超声检测的轮箍缺陷模糊模式识别研究[D].中国科学院研究生院, 2006
[32]李春芾.基于加速度计的汽车制动与转向性能监测方法研究[D].吉林长春:吉林大学, 2007.
[33] Li J, Chen P. Study of instrument for vehicle braking performance test based on multi dimensional accelerometer[J]. Instrument Techniques and Sensor. 2006, (10): 11-44.
[34]杨宾峰.脉冲涡流无损检测若干关键技术研究[D].长沙:国防科技大学. 2006
[35] Baek K S, Kyogoku K, Nakahara T. Influence of parameters on the transient traction of wheel/rail contact under low-speed and wet conditions[J]. Toraibarojisuto/Journal of Japanese Society of Tribologists. 2007, 52(5): 359-366.
[36] Kang L, Misener J A, Hedrick K. (Draft) on-board road condition monitoring system using slip-based tire-road friction estimation and wheel speed signal analysis[C]. Chicago, IL, United States: American Society of Mechanical Engineers, New York, NY 10016-5990, United States, 2006.
[37] Hoseinnezhad R, Bab-hadiashar A, Rocco T. Real-time clamp force measurement in electromechanical brake calipers[J]. IEEE Transations on Vehicular Techonlogy. 2008, 57(2): 770-777.
[38]王小旭,薛红香,夏全喜,孙明.无陀螺惯性测量单元设计及仿真分析[J].中国惯性技术学报, 2008,(02):154-158.
[39] Bernieri A., Betta G., Ferrigno L. An Automated Self-Calibrated Instrument for Nondestructive Testing on Conductive Materials [J]. IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2004, 53(4): 955-962
[40]徐平.多层金属结构中腐蚀缺陷的脉冲涡流检测技术研究[D].长沙:国防科技大学, 2005
[41]陈重,崔正勤.电磁场理论基础[M].北京:北京理工大学出版社, 2003
[42]孙向阳.涡流探伤在铁路接触线裂纹检测中的应用[D].西南交通大学, 2005
[43]薛永盛.超声波(UT)探伤中影响缺陷检出的因素[J].云南水力发电, 2008, 24(4): 79-81
[44]陆唯一.海底管道检测超声探头及探头阵列的设计[D].上海:上海交通大学, 2007
[45]关卫和,王乐勤,阎长周等.压力容器与管道高温超声检测中的衰减[J].压力容器,2004.21(5):12~53
[46] JIANG Y Q, HE Y G. New digital algorithm for measuring electric frequency based on adaptive short time fourier transform[J]. Journal ofElectronicMeasurement and Instrument, 2006, 20(2): 10-15.
[47] WU ZY, LIY X, LIUH ZH, et al Application of short time Fourier transform in the vibration analysis of the large hydroelectricmachine[J]. Journal of Hydroelectric Engineering, 2005, 24(6): 115-120.
[48]张贤达,保铮.非平稳信号分析与处理[M].北京:国防工业出版社,1998.
[49]张佩瑶,马孝江.小波包信号提取算法及其在故障诊断中的应用[J].大连理工大学学报, 1997, 37(1): 67-72.
[50]中国机械工程学会无损检测分会.超声波检测[M].机械工业出版社. 2000: 11~36,56~109
[51]竺科仪.水浸超声探头频率、声场特性的分析及测定研究[D].浙江:浙江大学, 2006
[52]冯若.超声诊断设备原理与设计[M].中国医药科技出版社1993
[53]徐晨曦.脉冲涡流检测实验系统的研制[D].上海:上海交通大学, 2009
[54]孙传友.测控电路及装置[M].北京航空航天大学出版社. 2002
[55] Freescale Semiconductor, Inc.MMA1200D Datasheet[M]. www.freescale.com,2008.
[56]李家伟,陈积懋.无损检测手册[M].北京:机械工业出版社, 2002
[57] Analog Devices, Inc. ADXL323Datasheet[M]. Allwww.analog.com,2007.
[58] Freescale Semiconductor, Inc.MMA7455L Datasheet[M]. www.freescale.com,2008.
[59]胡天明.超声波探伤[M].电力出版社. 1992
[60]毛佳,刘子良.嵌入式超声探伤仪的信息采集及处理[J].仪器仪表学报, 2003, 24(4): 424-426
[61]孙洁娣,温江涛,靳世久.基于D-S证据理论的管道安全检测数据融合技术的研究[J].传感技术学报, 2007.20(2): 404～407
[62]侯俊.证据理论几个关键问题研究[D].西安:西北工业大学, 2003
[63] P.Smets. The combination of evidence in the transferable belief model. IEEE Transactions on Pattern Analysis and Machine Intelligence,Vol.12.NO.5. MAY 1990: 447-458
[64]杨建平.证据理论及其在早期故障诊断信息融合中的应用研究[D].成都:电子科技大学, 2008
[65]徐维超.超声红外热波无损检测技术应用于裂纹检测的研究[D].北京:首都师范大学, 2008
[66]任江毅.铜-钢无熔深焊接接头超声检测缺陷识别及规律研究[D].南京:南京理工大学, 2007
[67]陈建忠,史耀武.超声检测过程的数值模拟[J].无损检测. 2001, 23(5): 198~202
[68]高克全.水轮机不锈钢叶片超声波探伤实验研究[J].无损检测, 2003, 25(12): 628~630
[69]丁学龙.铜钢熔敷焊的超声自动检测[D].南京:南京理工, 2007
[70]贾宇平,杨威,付耀文等.一种基于白化权函数的基本信任分配构造方法[J].信号处理, 2009,25(7): 1038~1043
[71]胡晓明.数据融合和Manhattan距离在液压泵故障诊断中的应用[J].机床与液压, 2009, 37(10):266~268
[72]杨静,田亮,赵爱军等.基于典型样本的证据理论信度函数分配构造方法[J].华北电力大学学报, 2008, 35(5): 70~77