• Journal of the Korean Society for Nondestructive Testing
• /
• v.27 no.5
• /
• pp.383-392
• /
• 2007

In this paper, we study the prediction of depth and width of a defect in steam generator tube in nuclear power plant using neural network. To this end, we first generate eddy current testing (ECT) signals for 4 defect patterns of SG tube: I-In type, I-Out type, V-In type, and V-Out type. In particular, we generate 400 ECT signals for various widths and depths for each defect type by the numerical analysis program based on finite element modeling. From those generated ECT signals, we extract new feature vectors for the prediction of defect size, which include the angle between the two points where the maximum impedance and half the maximum impedance are achieved. Using the extracted feature vector, multi-layer perceptron with one hidden layer is used to predict the size of defects. Through the computer simulation study, it is shown that the proposed method achieves decent prediction performance in terms of maximum error and mean absolute percentage error (MAPE).

• Journal of the Korean Society for Nondestructive Testing
• /
• v.29 no.4
• /
• pp.374-382
• /
• 2009

Steam generator(SG) tube, as a barrier isolating primary to the secondary coolant system of nuclear power plants(NPP), must maintain the structural integrity far the public safety and its efficient power generation capacity. And SG tubes bearing defects must be timely detected and taken repair measures if needed. For the accomplishment of these objectives, SG tubes have been periodically examined by eddy current testing(ECT) on the basis of administrative notices and intensified SG management program(SGMP). Stress corrosion cracking(SCC) on the SG tubes is not easily detected and even missed since it has lower signal amplitude and other disturbing factors against its detection. However once SCC is developed, that can cause detrimental affects to the SG tubes due to its rapid propagation rate. Accordingly SCC is categorized as prime damage mechanism challenging the soundness of the SG tubes. In this study, reproduced EDM notch specimens are examined for the detectability and quantitative characterization of the axial ODSCC by +PT MRPC probe, containing pancake, +PT and shielded pancake coils apart in a single plane around the circumference. The results of this study are assumed to be applicable fur providing key information of engineering evaluation of SCC and improvement of confidence level of ECT on SG tubes.

• Proceedings of the KIEE Conference
• /
• 2005.10c
• /
• pp.53-55
• /
• 2005

The signals of the eddy current testing(ECT) for the examination of the steam generator(SG) tubes in the nuclear power plant(NPP) determine the existence, size, and kind of defects using the variation of impedance signals when a testing coil, driven by alternating current, passes through the SG tube contains defects. The aim of this paper is building a database of the RPC probe signals on the basis of the sizes variation of defects and frequency variation of probe. In this paper 3-D numerical analysis of the ECT signals using the finite element method is performed. Through this study, it is shown variation of magnitude and phase of impedance according to variation of defect size and frequency. From the result of this paper, we can obtain the information which is useful in defect discrimination of SG tube in nuclear power plant.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.30 no.1
• /
• pp.13-19
• /
• 2010

In this paper, we study a new estimation system for the prediction of steam generator tube defects. In the previous research works, defect size estimators were independently designed for each defect types in order to estimate the defect size. As a result, the structure of estimation system is rather complex and the estimation performance gets worse if the classification performance is degraded for some reason. This paper studies a new estimation system that does not require the classification of defect types. Although the previous works are expected to achieve much better estimation performance than the proposed system since it uses the estimator specialized in each defect, the performance difference is not so large. Therefore, it is expected that the proposed estimator can be effectively used for the case where the defect type classification is imperfect.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2011.04a
• /
• pp.693-699
• /
• 2011

The outermost SG tubes have more structural problems than inside tubes. Many studies have used a uniform added mass coefficient for all of the SG tubes during the FIV analysis. The purpose of this study is to find out the added mass coefficients for each tube in a cylindrical shell. When a number of tubes are increased, added mass coefficients are also increased. And added mass coefficients at outermost tubes are less than those of inside tubes. According to gap changes between outermost tube and cylindrical shell, added mass coefficients are decreased with increasing the gap. When the gap has very large value, it shows that the added mass coefficient is asymptotically converged to the value of the tube array in a free fluid field.

• Nuclear Engineering and Technology
• /
• v.46 no.6
• /
• pp.869-874
• /
• 2014

It was discovered in a Korean PWR that an extensive number of very short and shallow cracks in the SG tubes were undetectable by eddy current in-service-inspection because of the masking effect of sludge deposits. Axial stress corrosion cracks at the outside diameter of the steam generator tubes near the line contacts with the tube support plates are the major concern among the six identical Korean nuclear power plants having CE-type steam generators with Alloy 600 high temperature mill annealed tubes, HU3&4 and HB3~6. The tubes in HB3&4 have a less susceptible microstructure so that the onset of ODSCC was substantially delayed compared to HU3&4 whose tubes are most susceptible to ODSCC among the six units. The numbers of cracks detected by the eddy current inspection jumped drastically after the steam generators of HB4 were chemically cleaned. The purpose of the chemical cleaning was to mitigate stress corrosion cracking by removing the heavy sludge deposit, since a corrosive environment is formed in the occluded region under the sludge deposit. SGCC also enhances the detection capability of the eddy current inspection at the same time. Measurement of the size of each crack using the motorized rotating pancake coil probe indicated that the cracks in HB4 were shorter and substantially shallower than the cracks in HU3&4. It is believed that the cracks were shorter and shallower because the microstructure of the tubes in HB4 is less susceptible to ODSCC. It was readily understood from the size distribution of the cracks and the quantitative information available on the probability of detection that most cracks in HB4 had been undetected until the steam generators were chemically cleaned.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.21 no.5
• /
• pp.518-525
• /
• 2001

As steam generator (SG) tubes have aged, new and subtle flaws have appeared. Most of them start growing from outside the tubes. Since signals from outer diameter (OD) defects are very weak compared to those from inner diameter (ID) defects in the conventional eddy current testing due to skin effect, this paper studies the feasibility of using remote field eddy current (RFEC) technique, which has shown equal sensitivity to ID and OD defects in the ferromagnetic pipe inspection. Finite element modeling studies show that the operating frequency needs to be increased up to a few hundred kHz in order for RFEC effects to occur in the nonmagnetic SG tube. The proper distance between exciter and sensor coils is also found to be about 1.5 OD, which is half the distance used in the ferromagnetic pipe inspection. Defect signals obtained by the designed RFEC probe show equal sensitivity to ID and OD defects and the existence of linear relationship between defect depth and phase signal strength. These results tell us that RFEC inspection is feasible even in nonmagnetic steam generator tubes.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.11
• /
• pp.2095-2101
• /
• 2008

In this paper, we consider a performance improvement of neural network for predicting defect size of steam generator tube using early stopping. Usually, neural network is trained until MSE becomes less than a prescribed error goal. The smaller the error goal, the greater the prediction performance for the trained data. However, as the error goal is decreased, an over fitting is likely to start during supervised training of a neural network, which usually deteriorates the generalization performance. We propose that, for the prediction of an axisymmetric defect size, early stopping can be used to avoid the over-fitting. Through various experiments on the axisymmetric defect samples, we found that the difference bet ween the prediction error of neural network based on early stopping and that of ideal neural network is reasonably small. This indicates that the error goal used for neural network training for the prediction of defect size can be efficiently selected by early stopping.

• Nuclear Engineering and Technology
• /
• v.33 no.2
• /
• pp.145-155
• /
• 2001

This simulation first allows us to define a transition zone from a bubble to jet mode of the argon out-flow and hereinafter to define a similar area for water-steam leak in the KALIMER SG (Korea Advanced Liquid Metal Reactor Steam Generator) using a water mock-up system, taking into account the KALIMER leak classification and tube bundle design, as a simulation of a real water-steam into sodium leak. in accordance with leak conditions in the KALIMER SG, the transition from bubbling to jetting is studied by means of turbulence regime simulation for argon out-flow through a very small orifice, which has the equivalent diameter of about 0.253 mm. finally the noise generation mechanism is explained from the existing experimental data. We also confirmed the possibility of micro-leak detection from the information of the bubbling mode through simulations and the experiment in this study.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.24 no.1
• /
• pp.63-70
• /
• 2010

In this paper, we study the structure of neural network for predicting defect size of steam generator tube. After extracting the features from the eddy current testing (ECT) signals, multi-layer neural networks are used to predict the defect size. In order to maximize the prediction performance for the defect size, we should carefully choose the structure of neural networks, especially the number of neurons in the hidden layer. In this paper, it is shown that, for the prediction of defect size, the number of neurons in the hidden layer can be efficiently determined by using cross-validation.