Optmzaton Frequency Desgn of Eddy Current Testng NAONG MUNGKUNG 1, KOMKIT CHOMSUWAN 1, NAONG PIMPU 2 AND TOSHIFUMI YUJI 3 1 Department of Electrcal Technology Educaton Kng Mongkut s Unversty of Technology Thonbur,Bangkok 2 Suranaree Unversty of Technology, Nakornratchasma THAIAND 3 Ota Natonal College of Technology, Ota JAPAN Emal:narong_kmutt@yahoo.com Abstract: The purposes of ths research were to construct the nondestructve metal testng ntrument by usng eddy current method and to fnd an optmal frequency for the metal testng ntrument. The testng ntrument consst of a sne wave oscllator crcut whch can adjust the frequency between 20 90 khz, and a 50 ohms sensor crcut. There are three knds of testng ntrument. The frst was the nondestructve mperfecton testng by usng eddy current method. The sample rons are constructed wth dfferent mperfecton on surface. The output sgnals of testng from the sensor crcut are compared. The second was the nondestructve categorzaton metal testng by usng eddy current method. Many knds of metals are taken to testng. The last one was the nondestructve testng for fndng the thckness of flms on ron by usng eddy current method. In ths testng the thckness of flms vared between 100 700 mcrons. In all testng, dfferences of the sgnal testng were compared to analyze the optmal frequency for the testng ntrument. The results of research showed that the nondestructve metal testng ntrument by usng eddy current method can be used to fnd a dfferent mperfectve ron, categorze the metal and fnd thckness of flms. In addton, the range of an optmal frequency s 30 khz to 70 khz for testng ntrument. Keywords Nondestructve, eddy current, thckness of flms, optmal frequency 1. Introducton Many research actvty on nondestructve testng has been motvated by the need for precse evaluaton of crackes and flaws for the assessment of the expected lfe of the metal components. Dfferent knds of ECT technques are performed for the detecton of defects n conductve materals. Eddy current are nduced n spemen as a result of the applcaton of an alternng magnetc feld. Eddy current could be made by hgh frequency magnetc feld. The magnetc feld happens when hgh frequency AC current enters prmary col. In case there s contnuous space nsde the work materal, the eddy current wll be hgher. In case there s no contnuous space nsde the work materal, the eddy current wll be lower. Ths dfference could be used to measure the contnuty of the work materal by usng eddy current [1-3]. Ths research was to desgn and develop nondestructve metal testng ntrument by usng eddy current method, whch conssted of oscllator crcut whch can adjust frequency and a 50 ohms sensor crcut, and to examne the mpact of used frequency n testng as well. 2. Analyss and Desgn When a exctng col s conducted close to a metal, an eddy current s nduced n metal. If there s acrack n the sheet metal, the eddy current wll avold that area, whch causes varaton n the manetc feld generated by the eddy current. In ths study, desgn of metal testng ntrument was to be used n fndng mperfecton of ron, categorze many knds of metal, and to fnd the thckness of flms on ron surface by nondestructve testng. The researchers appled the prncples of eddy current to desgn sne wave oscllator whch could adjust frequency between 20 90 khz, wth a 50 ohms sensor crcut. The crcut s used to fnd the optmal frequency for the eddy current testng. Detals related to the analyss and the desgn could be shown n forms of blocks controllng each part of for nondestructve metal testng ntrument by eddy current method as follows: 213
Sens Ampl Osclos f Powe Ampl Meter 470Ω f Oscllat 1kΩ Fgure 1. Components of the Nondestructve Metal Testng System by Eddy Current Method 5kΩ 5kΩ 2.1 Desgn of Wave Oscllator Crcut[4] Sne wave oscllator crcut could be smply made by one IC n order to reduce the complexty of desgn and materal. IC number X2206 whch could be found at a reasonable prce and easy to desgn was used. The crcut structure was as shown n Fgure 2 and the frequency s 1 f = C 2.2 Desgn of Crcut for Electrcal Voltage Amplfcaton Sgnal from oscllator crcut must be amplfed so that t could be used n a practcal way. Therefore, crcut for electrcal voltage amplfcaton must be desgned by usng op-amp number CA3130. Ths crcut would work faster and could be used wth dfferent range of frequency. Prncple n desgn and practce was based on nvertng type sgnal amplfer crcut. Gan rate depends on and f and could be calculated by amplfer rate A C of crcut n Fgure 2 as shown n the followng equaton: f AC = (2) Therefore, output could be calculated by the equaton: f VO = xe (3) 2.3 Desgn of Crcut for Electrcal Power Amplfer Sgnal from oscllator crcut had to be amplfed by nvertng type amplfer crcut. The output sgnal was hgher but that sgnal could not drve 50 ohms; therefore, ths sgnal had to be amplfed by transstor amplfer crcut. Transstor number H1061 was electrcal power amplfer so that t could dstrbute voltage to 50 ohms sensor crcut (1) Fgure 2 Wave Oscllator Crcut vo β' Av = vn rπ o Zn A = Av V = (4) = (5) n G = A A (6) v o = A V = V (7) v 2.4 Desgn of Sensor Crcut Inductve col was an electrcal load. Electrcal current whch runs through col would nduct because magnetc lnes of force took place nsde nductve col. Voltage drop for nductve col from crcut could be calculated by the followng equaton: V (8) d = dt n = ( ωi cosωt) = ωi m s m cosωt When sensor crcut was used to test mperfecton of metal by usng nondestructve eddy current method and sensor crcut got closer to metal. The nductance value of the col would change. Ths change was due to varous reasons lke metal type, sze of mperfecton, dstance and oscllator frequency. Thus, nductance value of sensor changed dfferently. 3. Experment Ths experment was to construct 3 knds of nondestructve metal testng by eddy current method as follows[5]: 214
1. The frst was to do a nondestructve test of metal mperfecton by usng eddy current method. The test was a smulaton by makng dfferent mperfecton sze on metal and then sensor crcut was used to test by adjustng frequency between 20 90 khz durng the experment. most obvous results of sgnal change. Ths frequency was then was the optmal frequency. Fgure 5. A Nondestructve Test to Fnd the Thckness of Flms on Iron by Usng Eddy Current Method Fgure 3 A Nondestructve Test of Metal Imperfecton by Usng Eddy Current Method 2. The second was to do a nondestructve categorzaton of metal by usng eddy current method. The test was a smulaton by takng many knds of metal (ron, copper, brass, alumnum and stanless steel) to the test n order to categorze metal. The test was done by sensor crcut wth a 50 ohms crcut and the frequency used was around 45 khz durng the test. 4. esults 4.1 esults for Nondestructve Test of Metal Imperfecton Nondestructve test of metal mperfecton by usng eddy current method was the test to fnd out the dfferences of mperfecton n ron. Sample ron had been drlled wth dfferent wdth and depth on surface. The frequency used was between 20 90 khz and then the output sgnals were plotted n a graph to compare the dfferences of mperfecton. 6.10 6.05 6.00 1 mm hole length 2 mm hole length 3 mm hole length 5.95 Voltage (v) 5.90 5.85 5.80 Fgure 4. A Nondestructve Test of Metal Categorzaton by Usng Eddy Current Method 3. The thrd was to do a nondestructve test for fndng the thckness of flms on ron by usng eddy current method. The thckness of flms n the test vared between 100 700 mcrons. The test would ncrease 100 mcrons of flm thckness on each tme and the frequency used was between 20 90 khz n order to fnd out the optmal frequency. It was found that the frequency between 30 50 khz showed the 5.75 5.70 5.65 5.60 0.00 1.00 2.00 3.00 4.00 5.00 Hole dameter (mm) Fgure 6. Iron mperfecton at 40 khz 215
4.2 esults for Nondestructve Test of Metal Categorzaton Nondestructve test of metal categorzaton by usng eddy current method was the test to fnd out the dfferences n knd of metal (ron, copper, brass, alumnum and stanless steel). The frequency used was around 45 khz. The output sgnals were plotted n a graph as shown n Fgure 7. 4.3 esults for Test of Flm Thckness on Irons Nondestructve test of flm thckness on ron by usng eddy current method was the test to fnd out dfferences n flm thckness. Flms wth 7 dfferent knds of thckness were used for the test wth dfferent frequency. The output sgnals were plotted n graph to show the relatonshp towards to flm thckness as Fgure 8-9. Fgure 8 Typcal Test at Fgure 9 Typcal Test at 40 khz 60 khz Magnet0 Magnet Usua Starlet Bras Alumnu Coppe Fgure 7 Graph for Electrcal Attrbute of Metal 5. Conclusons Ths research was to propose how to desgn and construct nondestructve metal testng ntrument by usng eddy current methods. There were 3 knds as follows: 1. Nondestructve test of metal mperfecton by usng eddy current method was a smulaton test by makng dfferent mperfecton sze and then 50 ohms sensor crcut was used to test. It was found that the frequency between 50 70 khz showed the most obvous dfferences and t was sutable for ths knd of test. 2. Nondestructve test of metal categorzaton by usng eddy current method was a smulaton test by takng many knds of metal to the test and then 50 ohms sensor crcut was used to test. It was found that ron showed the obvous dfferences and t was sutable for ths knd of test. 3. Nondestructve test of flm thckness on ron by usng eddy current method was done by testng flms wth thckness of 100 700 mcrons. The test would ncrease 100 mcrons of flm thckness on each tme and the frequency used was between 20 90 khz n order to fnd out the optmal frequency. It was found that the frequency between 30 50 khz showed the most obvous dfferences and t was sutable for ths knd of test. eferences [1]. T. Takag, M. Hashmoto, H. Fukutom, M. Kurokawa, and K. Mya et al., Benchmark model of eddy current testng for steam generator tube: experment and numercal analyss, Internatonal Journal of Appled Electromagnetcs n Materals, vol. 5, no. 2, pp. 149 162, 1993. [2] H. Fukutom, H. Huang, T. Takag, and J. Tan, 216
Identfcaton of crack depths from eddy current testng sgnal, IEEE Transactons on Magnetcs, vol. 34, no. 5, pp. 2893 2896, Sept. 1998. [3] A. Kamear, Soluton of asymmetrc conductor wth a whole by FEM usng edge-element, COMPE, vol. 9, pp. 230 232, 1990. [4] Wrote Asavarangsee, Chatchawal Temrtwong, Kornchulee Chasatht, 1988, Usage of Op-Amp and near IC, 1 st Impresson, Se-Educaton, Bangkok, pp. 197-201 [6] V.S.Cecco, G.Van Drunen,and F..Sharp, 1986, Advanced Manual For: Eddy Current Test Method, Canadan General Standards Board, Canada, pp 5-24 [5] Toshyuk Takag, 1998,Numercal Evaluaton of Correlaton between Crack Sze and Eddy Current Testng Sgnal by a Very Fast Smulator,IEEE Transactons on Magnetcs, Vol. 34, No. 5, September 1998, pp 2581-2587 [6] Xao-Me Pe, 2002, A Frequency Spectrum [5] Analyss Method for Eddy Current Nondestructve Testng, Proceedngs of the Frst Internatonal Conference on Machne earnng and Cybernetcs, Bejng, 4-5 November 2002, pp. 1194-1197 217