67 Eddy Current Nondestructive Evaluation Based on Fluxgate Magnetometry Umberto Principio Sponsored by: INFM Introduction Eddy current (EC) nondestructive evaluation (NDE) consists in the use of electromagnetic techniques through which the inspection of a part is carried out by detecting magnetic anomalies in the material under examination [1]. An innovative EC NDE method consisting in an advanced magnetometry technique, known as Fluxgate magnetometry, was used in this work. A Fluxgate sensor, known as a core saturation magnetometer, is a solid state device to measure intensity and direction of a magnetic field vector in a sensitivity range, between 10-10 and 10-4 T [2]. The basic configuration of a Fluxgate sensor is shown in Fig. 1. The sensor core is excited by an a.c. current I exc in the excitation winding so that the core permeability µ (t) is modulated with twice the excitation frequency. B o is the measured d.c. magnetic field and B(t) the corresponding field in the sensor core. V ind is the voltage induced in the pick-up (excitation) winding with N turns. The following description of the Fluxgate sensor will be based on the second harmonic principle, by means of which a sensitivity as low as 1 m γ (1 γ = 1 nt) can be reached; under these conditions the sensor is called Super Fluxgate [3]. To convert the magnetic flux and measure it through an electrical signal, the Fluxgate uses a reference signal consisting in a sinusoidal wave applied by a premagnetization or excitation coil (Fig. 2) to a high permeability magnetic core which is periodically brought to its saturation level (Fig. 3a, b). When there is no external field, the magnetic induction, B, in the core detected by a second coil, called pick-up, has an almost sinusoidal trend (Fig. 3c), with a slight flattening at the extreme points (beginning of saturation). The pick-up output signal is proportional to the derivative of the flux in the core and is obviously a periodic function (Fig. 3d). Through a Fourier analysis, it is possible to demonstrate that the latter function includes only odd harmonics (Fig. 3e, g). The application of an external magnetic field in the core axis direction produces, if positive, a further flattening of the magnetic induction signal at maximum points, whereas it approaches a sinusoidal trend at minimum points (Fig. 3c). In such a situation, the second harmonic in the signal detected by the pick-up appears (Fig. 3f) and its amplitude is proportional to the applied field. In the present nondestructive evaluation (NDE) context, the external magnetic field to be measured is produced by the eddy currents (EC) induced by the magnetic induction field generated by an excitation coil of type 2D (Fig. 4) in the sample under examination. The 2D coil is composed of 60 windings so as to form a double D of 25 mm diameter, with a distance between the two Ds of about 1 mm and a thickness of 2 mm. The currents in the central segments of the double D are equiverse so that the magnetic field at the center of the coil is zero. This configuration allows, by moving the coil, to zero the magnetic field sensed by the magnetometers in order to obtain a basic magnetic field that coincides with the measuring sensitivity of the device. The main characteristics of Fluxgate sensors are: high linearity and stability, high sensitivity, ease of realization, robustness and low cost. Moreover, it offers a high sensitivity in a wide frequency range going from DC to 10 khz [3]. Fluxgate EC NDE testing Experimental EC NDE tests were conducted on several samples of Al alloy (7075 and 6061-T6) plates with artificial defects in the form of slots and holes. The EC NDE tests consisted in plane scans carried out on the samples from both the defect side and the side opposite to the defect. Description of the testing apparatus A block scheme of the Fluxgate magnetometer is illustrated in Fig. 5 [5]. A generator produces a sinusoidal current which flows through the premagnetization coil. The second harmonic of the pick-up coil output signal is filtered through a band pass filter tuned to a frequency double than that of the generator. The signal obtained is amplified and then straightened by means of a device sensitive to the phase, controlled by a signal derived from the generator with a frequency doubler. Finally, an integrator produces a display signal proportional to the magnetic field to be measured. The scheme of the EC NDE system with Fluxgate sensors is illustrated in Fig. 6. The main components are:
68 1. Scanning system allowing for scans in the X-Y plane piloted by a PC. 2. Two Fluxgate magnetometers in gradiometric configuration. 3. Excitation coil of type 2D located orthogonal to the sensors. 4. Control electronics and signal conditioning. 5. PC to control the measurements and data acquisition. The sensors are mounted in a configuration called gradiometric ; this configuration has the advantage of considerably reducing the noise generated by sources distant from the device, thus improving the signal-to-noise ratio (SNR). The first sensor detects both the signal and the noise, whereas the second sensor detects only the noise which is eliminated by means of a vector amplifier (lock-in) [6]. The control electronics is basically that of the magnetometers. It allows to select the set point of the single sensors, to amplify the output signal and to relocate the set point of the magnetometers when they are saturated. An operational amplifier with unitary gain and supplied with a stage dephaser between 0 and 180 degrees allows for the phase correction between the two sensors so that amplitude subtraction is feasible. The signals are subtracted and demodulated by the lock-in which uses, as a reference signal, the EC excitation signal. The band width of the demodulated signal varies between 3 and 30 Hz. The demodulated signal is digitized by an A.D.C. (Analog to Digital Converter) at 16 bits and stored on disk for post processing. The measurement is controlled via software by a PC through a GPIB and/or RS232 interface. The software allows to control the movements of the XY table and to acquire the data when the device reaches the required position (x i, y i ). Maps of the magnetic field amplitude and phase generated by the ECs flowing in the sample are obtained. Materials and samples The first sample was an 7075 Al alloy plate with a thickness of 15 mm and 6 slots of depths between 1 mm and 6 mm manufactured by Electro Discharge Machining (EDM). The geometrical characteristics of the sample are illustrated in Fig. 7. The second sample analyzed was a multilayer obtained by the superimposition of two 6061-T6 Al alloy plates. One plate had no artificial defects and was 200 mm x 100 mm x 3 mm in size; the other plate of the same size had a slot of depth 1 mm, width 2 mm, and length 20 mm. The two plates were superimposed according to two different multilayer configurations illustrated in Fig. 8. The aim of the superimposition was that of visualizing the behavior of the magnetic field as it passed from the upper to the lower plate, and of realizing a sample configuration with a slot at the interface. Results and discussion Al alloy sample with slots obtained by EDM (sample 1) The scan was carried out from the defect side over slot 4. The vertical component B z of the magnetic field was detected at each material interrogation point during the X-Y scan. The scan parameters are: Sample: slot 4 of sample 1 Sensor: Fluxgate (A eff 30 mm 2 ) Step: x = 1 mm, y = 1 mm Area: x = 40 mm, y = 40 mm Generator: I = 5 ma, f = 277 Hz Lock-in Amplifier: Sensitivity = 1 mv, Filter = 30 ms. The image was obtained from a matrix of points, the elements of which are normalized values (between 0 and 1) of the vertical component of the measured magnetic field. A 128 gray tone image was obtained (Fig. 9). The defect is revealed by the presence of two characteristic lobes which depend on the particular geometry of the excitation coil. Considering the direction of the coil movement, the presence of the two lobes is explained bearing in mind that the induction coil passes from a balanced condition of the magnetic field 1 over the virgin material to an unbalanced condition caused by the slot when the first of the two Ds enters the defective zone. Once the two Ds are both over the defect, a condition of balance is established, which however does not reproduce the initial condition because the lift-off between coil and sample is modified by to the depth of the defect. The unbalanced condition is restored when the first of the two Ds abandons the defect while the second is still on it. The positivity of the two lobes is due to the fact that the processed signal is representative of the module of the magnetic field vector. 1 A slight initial unbalance is however always present.
69 It is possible to represent the signal in a colored scale both in two or three dimensions, as can be seen in Figs. 10 and 11. Multilayer (sample 2) Scans were carried out on the multilayer in configuration 1 and 2. As in the previous case, the vertical component B z of the magnetic field is detected at each material interrogation point during the X-Y scan. The parameters of the scans are the following: Scan 1: Sample: multilayer in configuration 1 Sensor: Fluxgate (A eff 30 mm 2 ) Step: 1 mm Area: x = 36 mm, y = 64 mm Generator: I = 100 ma, f = 377 Hz Lock-in amplifier: sensitivity = 30 mv, filter = 100 ms Scan 2: Sample: multilayer in configuration 2 Sensor: Fluxgate (A eff 30 mm 2 ) Step: 1 mm Area: x = 36 mm, y = 64 mm Generator: I = 100 ma, f = 377 Hz Lock-in amplifier: sensitivity = 10 mv, filter = 100 ms The 128 gray tone image obtained is reported in Fig. 12. The only difference with regards to the previous case is the presence of one positive and one negative lobe. This is explained bearing in mind that the magnetic field induced by one D is equal and opposite as to the one induced by the other D; therefore, the fields detected under an unbalanced condition are equal and opposite. Fig. 13 illustrates alternative 2D and 3D representations of the results. Future work The future development of the work will deal with the comparison of Fluxgate EC NDE results with SQUID sensor based EC NDE measurements. These sensors are made of superconductive material and are capable of detecting magnetic fields with a higher sensitivity in comparison with conventional magnetometers, including the Fluxgate magnetometer. References [1] Foerster, F., 1955, A Method for the Measurement of d.c. Field Differences and its Application to Nondestructive Testing, Non Destr. Test., Guilford. Engl., 13: 31-41 [2] Primdahl, F., 1970, The Fluxgate Mechanism, Part I, IEEE Trans. Magn., MAG-6: 376-383 [3] Gordon, D.I., Brown, R.E., 1972, Recent Advances in Fluxgate Magnetometry, IEEE Trans. Magn., Vol. MAG-8/1: 76-82 [4] Ripka, P., 1992, Review of Fluxgate Sensors, Sensors and Actuators, A. 33, Elsevier Sequoia: 129-141 [5] Primdahl, F., 1979, The Fluxgate Magnetometer, J. Phys. E: Sci. Instrum., Vol. 12: 241-253 [6] Snare, R.C., McPherron, R.L., 1973, Measurement of Instrument Noise Spectra at Frequencies below 1 Hertz, IEEE Trans. Magn., MAG-9: 232-235
70 Fig. 1 Fluxgate sensor basic configuration. I exc = excitation a.c. current; µ (t) = core permeability; B o = d.c. magnetic field; B(t) = magnetic field in the sensor core; V ind = voltage induced in the excitation winding; N = turns in the excitation winding. Fig. 2 (a) Principle scheme and (b) sensor element of a saturation magnetometer. 2D Induc tion Co il Z X Y Fig. 3 Principle of operation of a Fluxgate magnetometer based on the second harmonic. Fig. 4 Double D induction coil. Genera tor Sensor Band Pass Filter Amplifier Controlled Straightener Fre quency Doubler Display Fig. 5 Block scheme of the fluxgate magnetometer.
71 PreAmp. A PreAmp. B A - B Amp. ADC-DCA Converter Induction Coil Sample Sensors flux-gate PC Amp. Motors Controller X - Y Signal Generator GPIB IEEE488 Fig. 6 Scheme of the measuring apparatus with Fluxgate sensors. F E D C B A z x y 473 150 15 Slot A B C D E F Depth 1 mm 2 mm 3 mm 4 mm 5 mm 6 mm Fig. 7 Al 7075 sample with slots of various depth. A' A Multilayer sample Config. 1 sect. A-A Config. 2 sect. A-A Fig. 8 Multilayer obtained by the superimposition of two 6061-T6 Al alloy plates.
72 Fig. 9 2D grey tone image from the scan of sample 1 with indication of the excitation coil path. Fig. 10 2D pseudocolor image from the scan of sample 1. (a) Fig. 11 (a) 3D wirenet representation combined with the 2D grey tone image; (b) 3D pseudocolor graded surface. (b) Fig. 12 2D grey tone images from the scans on the multilayer sample in (a) configuration 1 and (b) configuration 2, with indication of the excitation coil path. Fig. 13 Contrast enhanced 2D images.
73 Fig. 14 3D wirenet representations combined with 2D gray tone images. Fig. 15 3D representations with pseudocolor graded surface.