ROBUST PARAMETER DESIGN OF SURFACE EDDY CURRENT PROBES. THE CASE OF MEASURING GEOMETRIC ANOMALIES IN A STATIONARY TEST OBJECT

The aim of the paper is to develop a method for increasing the signal-to-noise ratio of eddy current measurement of geometric anomalies in static planar objects without actually eliminating the inherent effects of noise factors. This is achieved by means of Taguchi's robust parameter design of rectangular frame surface probes, which allows determining the optimal configuration of their constructions. On a specific example, a robust configuration construction of the eddy current probe design is found, i.e., its technical variant that ensures a reduction of the output signal variance near its average value, i.e., resistance to noise disturbances, due only to the appropriate determination of the values of the controllable design and operating parameters of the probe without eliminating uncontrollable interference inherent in the test objects. For the robust design of a number of eddy current meters with different functionalities, a universal magnetodynamic model of the probe was used, which, together with the application of orthogonal arrays, allows the creation and implementation of Taguchi-design of experiments. The software that implements this model has been verified, including by comparing it with the results of calculations on test’s examples performed using the finite element method. The accuracy achieved in this case allows us to assert the adequacy of the created computer program. The data obtained as part of the Taguchi-design of experiment were used to evaluate design options using the “larger is better“ quality loss function and the signal-to-noise ratios calculated on its basis, which made it possible to select the optimal combination of design and operating parameters of the eddy current probe. The reliability of the found optimal configuration of the eddy current probe design was proved by confirmatory calculations. The research results were also experimentally verified on a prototype. References 21, figures 6, tables 9.