·Table of Contents
·Materials Characterization and testing
Absolute Electrical Property Measurements Using Conformable MWM Eddy-Current Sensors for Quantitative Materials CharacterizationAndrew Washabaugh, Vladimir Zilberstein, Darrell Schlicker, Neil Goldfine
JENTEK Sensors, Inc.
200 Dexter Ave., Watertown, MA 02472, USA
Tel: (617) 926-8422
Fax: (617) 926-8744
The conformable eddy current sensors called Meandering Winding Magnetometers (MWMä Sensors and MWM-Arrays) have been designed to match the response of computer models over a wide range of properties expected for materials under test. The MWM sensor responses for a material under test (e.g. a coating on a turbine blade) can be accurately modeled for many practical applications from basic principles using homogeneous layer representations. This permits accurate measurement of layer thickness, and absolute electrical properties (electrical conductivity and magnetic permeability) that can then be correlated with dependent properties such as stress, porosity, crack length, cold work, and microstructural changes due to overheating, aging, and other processing or service effects. This paper presents specific examples including:
Recent advances in electronic instrumentation, computational techniques, and sensor technology have yielded significant new capabilities for non-destructive evaluation applications, including coating characterization, flaw detection, and quality control. These advances include the development of conformable sensors with unique electrode geometries that can be accurately modeled so that quantitative real-time measurements can be performed in the field with minimal calibration requirements. As an example, MWM sensors and MWM-Arrays use unique geometries for the drive and sense windings that can be accurately modeled, can operate over a wide range of frequencies, and are amenable to fabrication by flexible circuit manufacturing techniques. The thin, flexible sensors can be used to inspect complex shaped components. This can be done in a scanning mode or with MWM sensors permanently mounted, similar to strain gages, in fatigue critical locations, particularly in difficult-to-access locations. When combined with grid measurement methods, the sensor response is converted into layer thickness or absolute material properties (such as the electrical conductivity or magnetic permeability) in real time. The results can be used either for manufacturing/rework quality assessment or for flaw detection/monitoring. This paper provides an overview of this sensing technology and some of its applications.
||Fig 1: (a) A schematic representation of an MWM sensor,
and (b) a conductivity-liftoff measurement grid.
The MWM structure can be produced using micro-fabrication techniques typically employed in integrated circuit and flexible circuit manufacture. This results in highly reliable and highly repeatable (i.e., essentially identical) sensors, which have inherent advantages over the coils used in conventional eddy current sensors. As indicated by Auld and Moulder (1999), for conventional eddy current sensors "nominally identical probes have been found to give signals that differ by as much as 35%, even though the probe terminal inductances were identical to better than 2%." This lack of reproducibility with conventional coils introduces severe requirements for calibration of the sensors (e.g., match sensor/calibration block sets). In contrast, duplicate MWM sensor tips have nearly identical magnetic field distributions around the windings as standard micro-fabrication (etching) techniques have both high spatial reproducibility and resolution. As the sensor was also designed to produce a spatially periodic magnetic field in the material under test, the sensor response can be accurately modeled which dramatically reduces calibration requirements. For example, in some situations an "air calibration" can be used to measure an absolute electrical conductivity without calibration standards.
An efficient method for converting the response of the MWM sensor into material or geometric properties is to use grid measurement methods. These methods map the magnitude and phase of the sensor impedance into the properties to be determined and provide for a real-time measurement capability. The measurement grids are two-dimensional databases of the sensor responses or lookup tables that can be visualized as "grids" that relate two measured parameters to two unknowns, such as the conductivity and lift-off (where lift-off is defined as the proximity of the material under test to the plane of the MWM windings). An example grid is shown in Figure 1(b). Since the database of the sensor responses can be generated prior to the data acquisition, the conversion using the table lookup operation is performed in real time as measurements are performed on a coupon, component, or structure. These grid methods also reduce the need for extensive calibration standards. In contrast, conventional eddy-current methods use empirical correlation tables that relate the amplitude and phase of a lift-off compensated signal (rather than an absolute electrical property) to parameters or properties of interest, such as crack size or hardness, and require extensive calibrations and instrument set-up.
3.1. Aircraft propeller blade inspection
During fabrication of aluminum-alloy blades for propellers used in civil and military aircraft, the fillet area is typically cold-worked (i.e., cold-rolled) with a high-strength steel roller. This operation introduces high compressive stresses to a significant depth, sometimes over 1 mm, and thus, substantially increases the fatigue resistance of the blades. The quality of this operation was traditionally verified by assessing the level of residual stresses through shallow blind hole drilling on selected blades or by X-ray diffraction (Yentzer, 1999). For production quality control, there are drawbacks in each of these methods of residual stress verification.
An alternative method for discriminating between adequately rolled blades, poorly rolled blades, and non-rolled blades is to use the MWM sensor with grid methods. Figure 2(a) shows an MWM measurement in the fillet region of a propeller blade. The cold work quality is characterized by the ratio of electrical conductivity measured in the axial and circumferential directions, across and along the rolling direction. The electrical conductivity measurements are made with the longer MWM winding segments oriented first in the axial direction of the blade and then in the circumferential direction. Each of these directional conductivity measurements are performed at two axially separated locations within the fillet (Yentzer, 2000). They reveal anisotropy introduced by the plastic deformation during cold rolling. The conformability of the MWM allows for accurate measurements even in this area of complex curvature. Figure 2(b) shows a representative quality control chart for the propeller blades. The data are plotted in terms of the conductivity ratio (axial/circumferential) characterizing the cold work at two selected locations in the blade fillet area. The measurements from acceptably rolled blades fall in an acceptance box, illustrated as the lower left quadrant of Figure 2(b). Unrolled blades and "suspect" blades fall outside this acceptance box. Since the MWM, with grid methods, is able to identify improperly rolled blades either during manufacture or in service, the blades can be rerolled to restore the required level of compressive residual stresses. The MWM procedures have been successfully implemented by the U.S. Air Force for C-130 propeller blades, and close to one thousand blades have been inspected so far.
Fig 2: (a) MWM inspection of an aircraft propeller blade. (b) Quality control chart used for evaluation of cold-rolled propeller blades.
3.2. Shotpeen quality control
Another method for introducing compressive residual stresses at the surface of fatigue-critical areas of components is through the process of shotpeening. In this process, a high-velocity stream of small beads (shot) or a special flapper tool are used to plastically deform a near-surface layer. The intensity of the shotpeening process is generally measured with Almen strips placed at various positions around the part. Within the plastically deformed, i.e. cold worked layer, high compressive residual stresses are locked in; these compressive stresses are balanced by tensile residual stresses in the unaffected "substrate", that is in the base metal which has not undergone cold work during shotpeening. The electrical conductivity of the cold worked layer is distinctly lower than the conductivity of the underlying base metal. This difference in conductivity is readily revealed by MWM measurements over a range of frequencies.
|Fig 3: Multiple frequency measurement results for shotpeened Al 2024 alloy.|
Figure 3 shows the results of multiple frequency measurements using the MWM with grid measurement methods for Al 2024 samples shotpeened to Almen intensities of 0.005, 0.012, and 0.017, Scale A. In these MWM measurements, the unpeened sample conductivity was essentially constant with frequency, which validated the quality of the reference part calibration performed prior to the measurements. For the peened samples, the effective conductivity varies with frequency. At the low frequencies, the sensor MWM depth of sensitivity is larger than the cold worked layer thickness, and the effective conductivity approaches the conductivity of the unaffected material alone. At the high frequencies, the MWM depth of sensitivity is smaller than or comparable with the cold worked layer thickness so that the effective conductivity mostly reflects the reduction in conductivity associated with the cold work from shot peening. Clearly, a higher peening intensity leads to a greater reduction in the effective conductivity at the surface. It is necessary for these conductivity measurements to be insensitive to surface roughness. This capability is provided through the Grid methods for independent conductivity and lift-off measurements. Once parameters based on MWM conductivity measurements are correlated with Almen intensity of a training set, shotpeening intensity can be determined at any location on shotpeened parts using MWM with Grid methods.
3.3. TBC Characterization
Thermal barrier coatings (TBC) are often used on gas turbine blades to protect the substrate material and allow for more efficient operation at elevated temperatures. The TBC generally consists of a metallic bond coat with a ceramic topcoat. Early versions of the MWM with Grid methods have proven successful at characterizing these coatings (Goldfine, 1998b). Recently, an automated algorithm has been developed for the independent nondestructive determination of the metallic bond coat thickness and its electrical conductivity as well as the ceramic topcoat thickness. Variations in the electrical conductivity of the bond coat generally correspond with variations in the porosity of the coating. The ability to map the electrical conductivity of the bond coat then allows for nondestructive porosity characterization or, at least, detection of areas with higher porosity. Day-to-day repeatability of the ceramic coating thickness measurements with the MWM is better than 2 m m (Goldfine, 1998b).
Figure 4(a) shows a representative set of measurements for various metallic bond coat thicknesses on a nickel superalloy substrate material. No calibration standards with coatings are required for these measurements. A reference part calibration was performed on the uniform substrate material alone. MWM measurements on the uncoated sample, with and without placing a 25-m m thick shim between the sensor and the test material, yielded a conductivity that was essentially constant with frequency and was not affected by the lift-off variations. The effective conductivity of the coated samples varies with frequency. At the low frequencies, the sensor responds to both the coating and the substrate properties, as the MWM depth of sensitivity is larger than the bond coat thickness. At the high frequencies, the sensor response to the bond coat is greater as the MWM depth of sensitivity becomes comparable to the bond coat thickness.
Fig 4: (a) Representative measurements of the effective conductivity.
(b) Representative metallic bond coat thickness measurements.
Figure 4(b) shows a comparison of the results from the automated algorithm for determining the bond coat thickness. These results are obtained within seconds after the actual measurements are performed and demonstrate a real-time nondestructive inspection capability. The procedure is particularly accurate in the 100 to 350 mm range of bond coat thicknesses. Ongoing development efforts are extending this algorithm to determine a fourth unknown, for example conductivity of the substrate, in addition to the currently determined three unknowns:
3.4. Magnetic Permeability Measurements
The MWM with Grid methods provides a capability to perform magnetic permeability measurements for characterization of magnetizable materials and for assessment of stresses. Magnetic permeability measurements with the MWM have also proved to be effective for fatigue damage assessment of initially solution annealed austenitic stainless steels. Fatigue damage in austenitic stainless steels can result in progressive formation of martensite of deformation (DeBacker, 1999) and thus increasing magnetic permeability. This is illustrated in Figure 5 for two stainless steel specimens provided by Siemens. One specimen was fatigue tested while the other was not. Surface scans with the MWM windings oriented perpendicular and parallel to the axis of the specimen show a bi-directional magnetic permeability in the fatigued specimen. The magnetic susceptibility is largest in the loading direction as the fatigue alters the microstructure of the stainless steel, creating a magnetic phase in the initially nonmagnetic material.
Fig 5: (a) MWM Measurement scans of bi-directional magnetic permeability along two austenitic stainless steel specimens. One was fatigue tested while the other was not. In the fatigue-tested specimen, MWM scans were made on diametrically opposite sides. (b) Measurement grid conductivity MWM data from (a) used to determine magnetic susceptibility,|
c = m /m 0-1 = m r-1.
3.5. On-line fatigue monitoring
In numerous applications, it is desirable to monitor fatigue "on-line", as it occurs. For example, for the inspection of aircraft structures, it is desirable to detect and monitor fatigue damage, crack initiation and crack growth in the earliest stages possible in order to validate the integrity of the structure. This is particularly critical for aging aircraft, where military and commercial aircraft are being flown well beyond their original design lives. Difficult-to-access locations, e.g. in the fuel tank of an aircraft, can particularly benefit from this capability. Other potential applications include
on-line monitoring of coupon, component and full-size article fatigue tests. Additionally, when test coupons with cracks of specified size are required or when new material alloys are developed, the ability to monitor the fatigue process on-line with permanently mounted MWM sensors provides a new powerful capability.
|| Fig 6: (a) Fatigue test apparatus with on-line MWM-Array crack initiation and growth monitoring. (b) Conductivity versus fatigue cycles. (c) Conductivity versus axial position.
|© AIPnD , created by NDT.net|||Home| |Top||