DEVELOPMENT OF EDDY CURRENT PROBES BASED ON MAGNETORESISTIVE ARRAY SENSORS N. Sergeeva-Chollet, C.Fermon, J.-M. Decitre, M. Pelkner, V.Reimund, M. Kreutzbruck QNDE, July, 25, 2013 CEA 10 AVRIL 2012
OUTLINE Introduction EC probe based on MR sensors simulation EC probe based on MR receivers Conclusions Baltimore July,25h 2013 PAGE 2
Sensitivity Sensibilité (db) [db] MAGNETORESISTIVE SENSORS USING INTEREST High density arrays GMR = very small size (length ~ 100µm) 100 µm 8 m m GMR Sensitivity at low frequency Better sensitivity than winding coils when f decreases 10 4 GMR 10 3 bobine coil 10 2 10 1 Detection of small surface defects - large array fast inspection - high density of the sensors high spatial resolution 10 0 10 1 10 2 10 3 10 4 10 5 Frequency Fréquence (Hz) [Hz] Same size of coil and commercial GMR sensor Detection of deep buried defects Skin depth effect implies that low frequency is required for the detection of deep buried defects: 1 m f Baltimore July, 25th 2013 PAGE 3
IMAGIC COLLABORATIVE PROJECT Integrated magnetic imagery based on spintronics components Aim of the project: development of spin electronics based systems compatible with NDT constraints in terms of sensitivity, dynamics of measurements and packaging Partners: INESC MN Non-destructive testing (EC) Magnetic sensors Buried flaws detection Surface cracks detection Probes with magnetic sensors as receivers and smart electronics GMR sensors array TMR sensors array Project main tasks: Applications definition, probe geometry definition Development and fabrication of MR sensing elements with high sensitivity and high resolution ASIC integration of close sensor electronics Magnetic sensors and probes testing Baltimore July, 25th 2013 PAGE 4
OUTLINE Introduction EC probe based on MR sensors simulation EC probe based on MR receivers Conclusions Baltimore July,25h 2013 PAGE 5
EC PROBE SIMULATION WITH CIVA Different possible geometries Combined defects, elliptic, parallelepipedic defects Emitter type coils with or without ferrite rectangular type uniform inducer GMR Reception coils or MR sensors magnetic field (at axes x, y and z) simulation reaction on the defect presence with magnetic sensor Current version: CIVA 10 Baltimore July, 25th 2013 PAGE 6
Imaginary part Imaginary part Real part Real part MR PROBE SIMULATION Configuration: - Inconel plate (1 MS/m), thickness = 1.55mm - Buried defect: L=10mm, W=0.1mm, H=0.9mm, Ligament = 0.62mm - Frequency: 100 khz Modeling of the EC probe: - Emitter: rectangular coil - Receiver: GMR along the X axis (component Bx of the magnetic field) defect defect scan Case n 1 : scan along the X axis (// to the axis of the defect) Case n 2 : scan along the Y axis ( to the axis of the defect) CIVA 10.0 EXP CIVA 10.0 EXP scan Impedance plane Impedance plane Good agreement between experiment and simulation Baltimore July,25th 2013 PAGE 7
J0 (A/mm 2 ) J0 (A/mm 2 ) J0 (A/mm 2 ) EMITTER OPTIMIZATION OPERA 0,1 J0100Hz J0500Hz J01000Hz J02000Hz J04000Hz J06000Hz J08000Hz 0,1 J0100Hz J0500Hz J01000Hz J02000Hz J04000Hz J06000Hz J08000Hz 0,1 J0100Hz J0500Hz J01000Hz J02000Hz J04000Hz J06000Hz J08000Hz 0,0 0-5 -10 z (mm) 0,0 0-5 -10 z (mm) 0,0 0-5 -10 y (mm) Double coil: larger penetration depth buried flaws Single coil: localized EC excitation surface cracks Baltimore July, 25th 2013 PAGE 8
J0 (A/mm 2 ) J0 (A/mm 2 ) COIL DISTANCE DEFINITION OPERA Influence of the coil distance 0,01 1E-3 d=1mm d=2mm d=4mm d=6mm d=10mm 0,1 0,01 d=1mm d=2mm d=4mm d=6mm d=10mm 1E-3 100Hz 1E-4 0-5 -10-15 z (mm) 1000Hz 1E-4 0-5 -10-15 z (mm) Optimal distance: d=4mm Baltimore July,25th 2013 PAGE 9
COMPARISON BETWEEN OPERA AND CIVA simulation of the coils with OPERA simulation of the coils with CIVA e w L dcoil lift-off h Material Aluminum (24.5 MS/m) Thickness: 20mm Coils current density = 1A/mm² each (opposite current direction) L = 20mm; w = 4mm; e = 1mm; h = 1mm; lift-off = 1mm Investigated parameters : - current density J - magnetic induction B Baltimore July, 25th 2013 PAGE 10
J0 (KA/m 2 ) J0 (ka/m²) J0 (ka/m 2 ) J0 (ka/m²) COMPARISON BETWEEN OPERA AND CIVA Investigated parameter: current density J Amplitude of the current density in the center of the coils as a function of z. x = 0 (black line cut). The distance of the coils dcoil is 4 mm The width of the coils varies between [2 10] mm OPERA f = 100Hz CIVA OPERA f = 10kHz CIVA 8 6 4 2 100Hz w coil 2mm 4mm 6mm 8mm 10mm 8 100Hz 6 4 2 wcoil 2mm 4mm 6mm 8mm 10mm 300 250 200 150 100 50 10kHz w coil 2mm 4mm 6mm 8mm 10mm 30 0 20 0 10 0 10kHz 0 wcoil 2mm 4mm 6mm 8mm 10mm 0 0 2 4 6 8 10 12 14 16 18 20 z (mm) 0 0 2 4 6 8 10 1 z (mm) 2 1 4 1 6 1 8 20 0 0 2 4 6 8 10 y (mm) 0 0 2 4 6 8 10 z (mm) Same shapes, similar values obtained in 2 cases Baltimore July, 25th 2013 PAGE 11
EMITTER / RECEIVER COUPLING Parameters: Aluminum plate, σ = 20 MS/m Buried flaw: 5 4 0.2 mm 3 at a depth of 5 mm Coil: w = 12 mm, L = 20 mm, H = 2 mm, e = 2 mm, f = 400 Hz Normal field component (Bz) is preferable Baltimore July, 25th, 2013 PAGE 12
CARTOGRAPHY EXAMPLE Parameters: Aluminum plate, σ = 20 MS/m Buried flaw: 5 4 0.2 mm 3 at a depth of 5 mm Coil: w = 12 mm, L = 20 mm, H = 2 mm, e = 2 mm, 1 turn, NI = 1A, f = 400 Hz GMR: s = 1 V/T, normal field component detection LO coil = 100 µm LO sens = 140 µm Scan direction Baltimore July,25th 2013 PAGE 13
OUTLINE Introduction EC probe based on MR sensors simulation EC probe based on MR sensor receivers Conclusions Baltimore July,25h 2013 PAGE 14
Signal (V) FIRST GMR PROBE PROTOTYPE First probe prototype GMR sensor R=200 S=25V/T First experimental results with GMR probe 80 khz scan Field (Oe) Surface defect L=3mm,w=0,1mm, d=0,8mm detection in Aluminum mock-up Baltimore July,25th 2013 PAGE 15
EXPERIMENTAL RESULTS WITH GMR PROBE Inconel mock-up cscan at 240kHz Vertical cut real imaginary R=200 S=25V/T L=6,5mm, w=0,1mm, H=0,9mm, ligament=0.6mm cscan at 240kHz Vertical cut real imaginary L=4,5mm, w=0,1mm, H=0,6mm, ligament=0.9mm Baltimore July,25th 2013 PAGE 16
EXPERIMENTAL RESULTS WITH GMR PROBE Aluminium Mock-Up x scan area defect 1: 0.3 ligament [mm] y 2: 0.68 3: 1.68 4: 2.72 5: 3.84 6: 4.73 7: 5.72 8: 6.83 9: 7.83 10: 8.88 Baltimore July,25th 2013 PAGE 17
U (mv) EXPERIMENTAL RESULTS 9 mm 15 mm 4 mm coil length = 20 mm NI=2 coils 40 windings 0.2 ma f = 500 Hz I = 200 ma Line scan @ y = 20.4 mm 500 250 defect 1 ligament 0.3 mm defect 7: 5.72 mm 0-250 defect 6 ligament 4.73 mm defect 6: 4.73 mm -500 0 20 40 60 80 x (mm) Baltimore July,25th 2013 PAGE 18
Sensitivity (mv/oe) FIRST TMR PROBE PROTOTYPE INESC MN TMR sensor 70x70µm² 42 TMR in series cscan at 240kHz 1.8 1.5 1.2 0.9 0.6 0.3 0.0-0.3-0.6-0.9 R min = 214.2 +/- 1.8 I = 0.3 ma Sensor 1 Sensor 2 Sensor 3 Sensor 4 Sensor 5-150 -100-50 0 50 100 150 Field (Oe) First experimental result with TMR probe Vertical cut First probe prototype real imaginary L=6,5mm, w=0,1mm, H=0,9mm, ligament=0.6mm, inconel mock-up PAGE 19
CONCLUSIONS Eddy current testing with the use of magnetoresistive array sensors is very promising for cracks and flaws detection in conductive materials Recently developed tools in CIVA and OPERA make possible optimization of mono- or multi- elements magnetoresistive sensors based EC probes. Comparison of these models gives a good accordance between them. New very sensitive probes for buried flaws based on GMR and TMR array sensors and ASIC integration are under development in the frames of a partnership European project IMAGIC FP7 288381 Baltimore July, 25th 2013 PAGE 20