Pulsed Eddy Current Inspection of Support Structures in Steam Generators

Abstract

 Abstract-Degradation and fouling of support structures in nuclear steam generators (SGs) can lead to SG tube damage and loss of SG efficiency. Inspection and monitoring of support structures combined with preventative maintenance programs can alleviate these effects and extend SG life. Conventional eddy current inspection technologies are extensively used for detecting and sizing indications from wall loss, frets at supports, cracks and other degradation modes in the tubes, as well as assessing the condition of support structures. However, these methods have limited capabilities when more than one degradation mode is present simultaneously, or when combined with fouling. Pulsed eddy current (PEC) combined with principal components analysis (PCA) was examined for inspection of 15.9 mm (5/8") Alloy-800 tubes and surrounding stainless steel (SS410) support structures. Clear separation of PCA scores associated with tubes from those associated with ferromagnetic SS410 supports permitted measurement of tube-to-support gaps, in either the presence of tube fretting or variation of relative position of the tube within SS410 supports. For concentric tubes, frets could be sized independently of SS410 hole diameter variations, which in other materials could represent support corrosion. Capability to clearly separate scores was attributed to large differences in relaxation times for diffusion of transient fields through the tube compared with diffusion into the ferromagnetic support structure.

Evaluation of machine learning tools for inspection of steam generator tube structures using pulsed eddy current

“…The experimental apparatus used here is described in detail in [5] and [10]. Four 25 mm long 410 SS samples, with hole IDs shown in Table 1, were used to simulate different states of corrosion in a uniform ferromagnetic drilled support structure.…”

“…The tube was accurately positioned within the hole using a micrometer apparatus. The PEC probe, described in detail elsewhere [5,8,24], consisted of a single drive coil wound coaxially with the probe body and 2 arrays of four pickup coils, placed at 90° intervals around the surface of the probe both above and below the drive coil. The probe was a close fit to the alloy 800 tube.…”

“…The MPCA used for this work differs from conventional PCA in that it retains the signal mean, producing a least squares interpretation instead of variance minimization [6]. This technique allows for incremental reproduction of signals through linear combinations of principal components and their scores, significantly reducing dimensionality, while retaining most information originally contained in the data [5].…”

“…PEC signal response can be viewed as consisting of a spectrum of discrete frequencies, far greater in number than what is normally captured by multi-frequency eddy current testing, but with an associated complexity in signal response. In order to address the subtle variations and extract dominant features in the measured PEC signal, a modified principal components analysis (MPCA) has been applied [5][6][7]. MPCA is based on a minimization of sum square residuals interpretation of results instead of a variance minimization interpretation as used in conventional PCA for feature extraction [6].…”

Angular Spectral Density and Information Entropy for Eddy Current Distribution

“…The conventional EC testing method is based on the single-frequency current exciting waveform and is used in many fields, such as defect detection and stress measurement. Multi-frequency EC testing [8] was proposed to detect defects in different depths, and pulse current exciting current [9] was proposed to enrich the frequency component. Chirp excitation [10], whose frequency changes with time, was used for frequency sweep.…”