With the development of non-destructive testing technology, eddy current testing is widely used in the thickness measurement of coating layers such as oil and gas pipelines, railway vehicles, and engine cladding due to the advantages of non-contact, no couplant, oil resistance, high sensitivity, fast detection and easy automation, but the measurement range of the existing eddy current Thickness Gauge of the coating layer in CHINA is mostly several hundred millimeters, and the accuracy is only ±3% H at a large range, which is difficult to meet the high-precision detection needs of thick coating, and the relevant research at home and abroad is mostly focused on thin coatings and thin metal plates.
Eddy current thickness measurement principle
When the eddy current probe detects that the coil is close to the surface of the coating, the lifting distance of the probe relative to the metal substrate x is the coating thickness d, and the detection signal strength and distribution are affected by the conductivity, permeability, thickness of the substrate material, as well as the coil excitation current, frequency and lifting distance By calibrating the relationship curve between the amplitude of the eddy current signal and the standard lifting distance, the coating thickness can be obtained according to the change of the signal amplitude during the actual detection.
Eddy current Thickness Gauge platform
In this study, the ECC-2004 intelligent multi-function electromagnetic Detector was selected with an absolute eddy current probe that can directly measure the impedance change of the coil, and a stainless steel plate of 198mm×100mm×5mm was used as the matrix specimen. The lifting operation is realized by the IS09001 standard test block, the U-groove guide rail is fixed on the bracket, the mobile bracket moves up and down with the help of rolling steel balls in the guide rail, the probe is fixed on the bracket through the fixture, and the lifting distance is accurately controlled by adding different standard test blocks between the fixing nut and the guide rail, and the eddy current signal generated by the lifting is input to the Detector through the probe converter, and then processed by pre-amplification, phase sensitivity detection, balanced filtering, phase rotation, gain amplification, etc., and then output to the impedance plane after A/D conversion.
Analysis and processing of test data
In the experiment, the excitation frequency of the eddy current probe was set to 18181Hz, the pre-setting was 25dB and other parameters, and the eddy current signals of different standard test blocks were collected with the 0mm lifting signal as the reference point, and it was found that the signal amplitude within 0~5mm had a nonlinear relationship with the lifting distance, and the signal tended to zero at 10mm. In order to solve the problem of large error of ECC-2004's built-in calibration method, sampling at 0~5mm intervals of 0.1mm, fitting with 4th polynomial, power function, composite index and Gaussian function, the results show that the sum variance and root mean square error of Gaussian function are the smallest and the determination coefficient is the highest, and the dispersion error is less than 0.089mm verified by 5 sets of data, which is a reliable calibration model.
conclusion
In order to improve the measurement range of the current domestic eddy current Thickness Gauge and enhance the measurement accuracy of thick coatings, this paper compares and analyzes the calibration effects of the four polynomials, power functions, composite exponential functions and Gaussian functions on the eddy current signal and the lifting distance, and deeply compares and analyzes the SSE, RMSE, R-square and maximum error of the fitted data and the original data, and systematically studies the calibration method to improve the eddy current thickness measurement.
The results clearly show that compared with other fitting methods, the Gaussian function fitting has the smallest lifting error and the highest measurement accuracy. In the range of 0~5mm, the error is only 0.089mm, which is significantly improved compared with the ±3%H (i.e. 0.15mm) pick-up error in the domestic large range. It can be seen that the measurement range and measurement accuracy of eddy current thickness measurement can be greatly improved by optimizing the fitting function of the calibration curve, and provide a feasible new method and new idea for the high-precision eddy current detection of thick coatings.
