Experimental Investigation of Influence on Non-destructive Testing by Form of Eddy Current Sensor Probe

Transcription

1 Expeimental Investigation of Influence on Non-destuctive Testing by Fom of Eddy Cuent Senso Pobe Fengyun Xie * and Jihui Zhou School of Mechanical and Electonical Engineeing, East China Jiaotong Univesity, Nanchang 3313, China xiefyun@163com Abstact Eddy cuent testing is a kind of non-destuctive testing (NDT) method based on the pinciple of electomagnetic induction The pobe is the key of eddy cuent non-destuctive inspection and its design o combination of fom will influence detectability A complete set of eddy cuent testing system is designed fo standad testing coppe and defect coppe Unde the same liftoff distance and the same high fequency excitation signal, designed thee types of pobe, which ae single pobe single coil, single pobe double coil and double pobe double coil, ae used to cay out NDT expeiment Expeimental esults show that sensitivity and esolution of detection system ae obvious diffeence among the diffeent fom of eddy cuent senso pobe Pobe shape and coil winding ae impoved accoding to the expeimental esults Coesponding impoved pobe is adopted to cay out NDT expeiment at the same condition Expeimental esults show that detectability is enhanced significantly Keywods: Eddy cuent testing, Pobe, Non-destuctive testing (NDT), Sensitivity 1 Intoduction Eddy cuent testing based on the pinciple of electomagnetic induction is a kind of NDT method fo testing metal semi-finished and metal components It has been moe and moe widely used fo its no need touch, high-speed of testing, easy automation, suitable fo online testing and high detection sensitivity of suface defect Fo example, YANG 1 et al, consideed identification of coosion finge in pulsed eddy cuent non-destuctive testing, He et al, consideed pulsed eddy cuent technique fo defect detection in aicaft iveted stuctues Howeve, the fom of senso pobe has a geat influence on the detection pefomance In ecent yeas, it is a eseach focus of many scholas how to impove the sensitivity of eddy cuent pobes But the discussion only is in the influence of one aspect, athe than the system Fo example, Ren 3 et al, poposed to incease the pobe diamete o adapt big pie-type pobe to impove the detection sensitivity, but it will educe detection sensitivity of small defects Kim 4 et al, consideed a dual-electomagnetic senso * Coesponding autho Tel: ; Fax: D Li, Y Liu, and Y Chen (Eds): CCTA 1, Pat IV, IFIP AICT 347, pp , 11 IFIP Intenational Fedeation fo Infomation Pocessing 11

2 444 F Xie and J Zhou system fo weld seam tacking of I-butt joints Li 5 et al, analyzed detection pinciple and influence factos of eddy cuent Yu 6 et al, designed a novel pulsed eddy cuent testing pobe based on 3D magnetic field measuement Gao 7 et al, studied on stuctue optimization of eddy cuent pobe, and obtained thee influence factos on eddy cuent sensitivity: exciting coil of eddy cuent is stonge than slot width, and slot width is stonge than detection coil of eddy cuent In this pape, we summaize the pevious discussion, full unscamble measuement pinciple of eddy cuent The pocedue is: to design a placed device of eddy cuent testing; to use diffeent foms of pobe to cay out compaative test to the standad coppe and defect coppe The specific objectives ae: to get the elations between pobe fom and sensitivity of NDT; to put fowad ecommendations fo impovement of the pobe Testing Pinciple and Methods of Eddy Cuent Eddy cuent testing is a kind of NDT method based on the pinciple of electomagnetic induction, which is applied to conductive mateials Testing pinciple is the following: the signal geneato povides high fequency altenating cuent to the pobe in the detection coil, detection coil poduces an altenating magnetic field, and the specimen poduces eddy cuent Eddy cuent is affected by the popeties of the specimen, and in tun it changes impedance (voltage) of coil, then oscilloscope output impedance (voltage) of coil Testing pocess includes picked up signal, signal amplification, signal pocessing, eliminated intefeence signal, and display and ecod test esults Thee ae two kinds of detection methods One is a double-coil high fequency eflection, whose schematic diagam is shown in Fig 1it has a high detection sensitivity of suface defect, but it is lowe defect sensitivity with deepe below the suface; The othe is double-coil low fequency tansmission, whose schematic diagam is shown in Fig Fig 1 Schematic diagam of low fequency eflection Fig Schematic diagam of high fequency tansmission Two methods have the same testing pocess Signal geneato poduced altenating cuent flows though the coil, when the pobe moves to the specimen, the eddy cuent will decease Impedance of the coil is displayed and ecoded by the mete

3 Expeimental Investigation of Influence on Non-destuctive Testing 445 Using auxiliay equipment to do compaative analysis of esults fom Standad specimens and defect specimens can obtain NDT esults 3 Theoetical Basis fo Expeimental Study The wokpiece signal of detected is fom impedance of the detection coil o the induced voltage changes of seconday coil in eddy cuent testing We use long staight solenoid of cylindical conducto as the object of eseach, suppose the adius of cylindical conducto as a less than the solenoid inne adius b, and suppose tuns of pe unit length as n In the cylindical conducto ( < <b), the magnetic field stength is H z ( ), magnetic field stength is equal to the excitation magnetic field H in the gap (a < <b)using the concept of Foste ective pemeability can obtain the magnetic flux though the coil coss-section[3] Φ = μ μ μ H π a + μ H π ( b a ) (1) Coil phaso of induced electomotive foce(emf) pe unit length is witten as ε = jω n Φ = jω n μ μ μ H πa jω n μ H π ( b a ) () Coil no-load induced EMF pe unit length is ε = jϖ n Φ = jϖ n μ H π b (3) Thus, coil of nomalized induced EMF pe unit length is ε ε = 1 η + ημ (4) a whee η = the coil filling facto Consideing elationship between the im- b pedance pe unit length and the induced voltage ae Z =, nomalized impedance is Z Z μ U I = 1 η + ημ μ (5) Fom eqs (4) and (5), we can obtain the same expession by nomalized induced EMF of the solenoid detection coil and nomalized impedance, which can instead each othe in calculating sensitivity Moeove, if the detection coil and the excitation coils have the same flux and tuns N, the induced EMF of detection coil is π d U = U (1 η + ημ μ ) = π fn μ H (1 η + ημ μ ) (6) 4 whee the induced EMF of detection coil is elation to excitation fequency f, detection coil diamete d, detection coil tuns N, excitation magnetic field fo H, fill facto η,the ective pemeability μ

4 446 F Xie and J Zhou 4 Eddy Cuent Testing Scheme Signal Geneato Detection pobe Peamplifie, Phase sensitive detection Balance filte Digital phase otation Adjustable gain amplifie A/D Compute system Display Fig 3 Eddy cuent testing scheme Accoding the pinciple of eddy cuent testing, the scheme is shown in Fig 3 Signal Geneato issues high fequency altenating signal to the detection pobe, taking into account impedance of the exciting coil and detection coil may not be equal We use a balanced cicuit to eliminate the voltage diffeence between the two coils in pe-test In the detection, the detection coil to detect defective pats will poduce a slight and imbalance signal Afte amplification, phase sensitive detection, filte, and emoved the intefeence signal will tun into the DC signal which contains phase and amplitude chaacteistics of impedance (voltage) of coil, at last displayed and ecoded by display device Accoding to ecods of the esults, we daw the nomalized sensitivity cuves with the liftoff change, put fowad ecommendations fo impovement of pobe the fom and cay out impovement by analyzing the cuves At the same time, we daw the nomalized sensitivity cuves with the lift-off change, and daw final test conclusions 5 NDT Expeimental Study 51 Test Pepaation In accodance with the testing pogam, pepaing expeimental appaatus, the main instuments include a high fequency signal geneato, an oscilloscope, a standad coppe, a defect coppe whose depth of defect size is mm, and length of defect size is 1mm, multiple pobes which have a single pobe self-inductance coil (1 tuns), a single pobe mutual inductance coil (5 tuns each), two double pobe tansmission coils (5 tuns)to make the coil diamete of pobe is 8mm and 1mm, thickness is 4mm and 6mm 5 Test Pocedue Test pocedue is to build test equipment in accodance with testing pogam We egulate output fequency of high-fequency signal geneato as 4KHz, adjust lift-off distance of excitation coil and detection coil espectively as mm, 1mm, mm, 3mm, 4mm, 5mm Using diffeent foms pobe test standad coppe and defect coppe, output voltage peak peak value by oscilloscope, we daw the nomalized sensitivity cuves of the diffeent pobe foms with the lift-off change, Abscissa fo the lift-off distance and odinate fo the nomalized signal amplitude of specimens Cuves ae the esults of diffeent pobe foms including single pobe self-inductance, single

5 Expeimental Investigation of Influence on Non-destuctive Testing 447 pobe mutual inductance, double pobe tansmission, double pobe eflection, double pobe in seies and in paallel Output voltage peak-peak value of can be obtained fom the display device 53 Influence of Detection Sensitivity by Diffeent Pobe Foms Fig 4 and 5 ae the nomalized sensitivity cuves with the lift-off change to initial test standad coppe and defect coppe Cuve 1 is the esult of oscilloscope by diectly connecting to signal geneato, cuve is the detective esult of the single pobe, cuve 3 is the esult of the double pobe in seies, cuve 4 is the no-load esult of the double-pobe, cuve 5 is the esult of double pobe eflection, cuve 6 is the detective esult of the double pobe tansmission Fig 4 Standad coppe sensitivity cuves with the lift-off Fig 5 Defect coppe sensitivity cuves with the lift-off 54 Influence of Impedance (Voltage) by Pobe Diamete Unde the 4KHz high-fequency altenating signal geneato, we change the pobe diamete Fig 6 shows the influence of pobe diamete, with the pobe diamete inceasing, the impedance moves down along the cuves, as the coil diamete inceases, the detection sensitivity deceases Nomalized impedance Lift-off 8 D mm Nomalized impedance Fig 6 Influence of impedance by pobe diamete

6 448 F Xie and J Zhou 55 Impoved Pobe Test Analyzing initial test esults can obtain that pobe size, coil thickness and coil tuns closely elated test esults Impoved pobe is the following : aanged the coppe wie of pobe as fa as possible odely, enhanced the magnetic flux density of the suface, inceased the pobe tuns by 1 tuns, changed the pobe diamete fo 8mm and thickness fo 4mm Using the impoved appopiate pobe include single pobe selfinductance, single pobe mutual inductance, double pobe tansmission, double-pobe eflection, double-pobe in seies and in paallel to cay out NDT and expeimental analysis, and daw sensitivity cuves of Fig 7 and 8 Fig 7 Standad coppe sensitivity cuves of impoved pobe with the lift-off Fig 8 Defect coppe sensitivity cuves of impoved pobe with the lift-off 56 Results and Discussion We contast sensitivity cuve of befoe and afte Pobe impoved 1) Sensitivity cuve 1 to 3 almost has no change unde small lift-off distance Cuve 1 is the esult of oscilloscope by diectly connecting to signal geneato, cuve is the detective esult of the single pobe, cuve 1 and cuve almost coincide, it shows little ect on the output changes by single pobe self-inductance Cuve 3 is the esult of the double pobe in seies, cuve 1 and cuve 3 almost coincide, it shows the pobe in seies with diffeent the numbe of tuns, if equal of tuns add, the esults have no change ) Sensitivity cuve 4 to 6 changes significantly, sensitivity of double-pobe tansmission no specimen is highe than the existing specimen, minimum sensitivity is double-pobe eflection 3) Detective sensitivity of standad coppe is highe than defect coppe With liftoff distance inceasing, sensitivity deceases apidly, fo example, Fig 5 shows the defect signal amplitude down to the 1/ 4 of the suface when cuve 6 lift-off fo the 1mm 4) Compaing Fig 4 and Fig 7, Fig 5 and Fig 8, the sensitivity of impoved pobe is significantly highe than befoe impovement 5) Pobe diamete significant has influence on the detection sensitivity, with the coil diamete inceasing, the detection sensitivity deceases As diamete of the pobe coil inceases, the magnetic flux density of wokpiece inceases, the eddy cuent value inceases, then the equivalent esistance deceases

7 Expeimental Investigation of Influence on Non-destuctive Testing Conclusion Eddy cuent NDT is high sensitivity fo detecting defects of conducto suface and nea suface, by compaison of the test, the following ae concluding summaies: 1) The diffeent pobe foms have a significant influence on the ability of eddy cuent testing Detectability of double pobe tansmission is geate than detectability of double pobe eflection, detectability of double-pobe is geate than detectability of single pobe The coil in seies, detectability will be no change if equal of tuns add ) Pobe diamete has influence on the detection sensitivity, with the coil diamete inceasing, the detection sensitivity deceases significantly To incease detection sensitivity, the pobe coil diamete must be equal to o less than the length of defects 3) Pobe placed has a geat influence on the detection sensitivity, with lift-off distance inceasing, detection sensitivity deceases apidly 4) Excitation coil has moe influence than detection coils on detection sensitivity, inceased excitation coils tuns can help to impove the detection sensitivity, but inceased detection coils tuns will decease the detection sensitivity Acknowledgment This wok is suppoted by the caying tools and equipment of Ministy of education key laboatoy (9JD3) Refeences 1 Yang, B, Luo, F, PenN, M: Identification of Coosion Finge in Pulsed Eddy Cuent Nondestuctive testing Chinese Jounal of Mechanical Engineeing 44(1), (8) He, Y, Luo, F, et al: Pulsed Eddy Cuent Technique fo Defect Detection in Aicaft Riveted Stuctues NDT & E Intenational 43(), (1) 3 Ren, J, Lin, J, Gao, C: Electomagnetic Detection Mechanical Industy Pess () 4 Kim, J-W, Shin, J-H: A Study of a Dual-electomagnetic Senso System fo Weld Seam Tacking of I-butt Joints Poc Instn Mech Engs Pat B: J Engineeing Manufactue 17, (3) 5 Li, G, Ma, H, Shen, J, He, F: Analysis of Detection Pinciple and Influential Factos of Eddy Cuent Chinese Jounal of Sensos and Actuatos (11), (9) 6 Zhang, Y, Sun, H, Luo, F, Yang, B: Design of a Novel Pulsed Eddy Cuent Testing Pobe Based on 3D Magnetic Field Measuement Chinese Jounal of Mechanical Engineeing 45(8), 49 5 (9) 7 Gao, X, Zhu, Z, Zhang, W: Study on Stuctue Optimization of Eddy Cuent Pobe Chinese Jounal of NDT 8(6), 8 3 (4)