Simulations and Measurements of Sodium Effect on the ECT Signal for SG Tubes of FBR
公開日:
カテゴリ: 第5回
1. Introduction
A small test sodium cylindrical tank, 2.2 m high and 0.6 m in diameter was filled with sodium at 500 degrees (see Fig. 1).In fast breeder reactors with sodium coolant, the integrity of steam generator (SG) tubes plays an important role in the reactor safety. These tubes are bundled in a cylindrical vessel with sodium located outside of the tube and water flowing inside of it. Due to the high chemical reactivity of sodium with water, penetration of SG tubes wall have to be avoided by all means and therefore, earlier defect detection before wall penetration during the in-service inspection (ISI) is required to increase the safety of SG tube component. Eddy current technique (ECT) is used to inspect the SG tubes [1], after sodium from the vessel is drained and the SG tubes are cooled down to the room temperature. After avoided by all means and therefore, earlier defect detection before wall penetration during the in-service inspection (ISI) is required to increase the safety of SG tube component. Eddy current technique (ECT) is used to inspect the SG tubes [1], after sodium from the vessel is drained and the SG tubes are cooled down to the room temperature. After sodium with water, penetration of SG tubes wall have to be avoided by all means and therefore, earlier defect detection before wall penetration during the in-service inspection (ISI) is required to increase the safety of SG tube component. Eddy current technique (ECT) is used to inspect the SG tubes [1], after sodium from the vessel is drained and the SG tubes are cooled down to the room temperature. After draining and cooling, there is a thin sodium layer adhering to the outer surface of SG tubes. Due to the high electrical conductivity of sodium layer, the eddy current signal will be modified. In the present paper, the authors investigate the effect of sodium on several types of outer tube defects with variable depths and widths using numerical finite element simulations. The numerical model of sodium is build after validation of experimental results using a test sodium tank experiment. The phase-amplitude diagram is constructed for defects with depths up to half of the tube wall thickness and 15 mm wide.Fig 1. View of the test sodium tankOvidiu MIHALACHE, Japan Atomic Energy Agency, Fast Breeder Reactor Research and Development Center, 1 Shiraki, Tsuruga-shi, Fukui-ken, 919-1279, Tel: 0770 39 1031 (ext 5211), email: mihalache.ovidiu@jaea.go.jpTwo SG tubes, one made of austenitic stainless steel (the Superheater tube) and the other made of Cr-Mo alloy (the Evaporator tube), were introduced in the tank and kept for more than two hours, in order to have a good wetting phenomenon (see Fig 2). After draining and cooling of the- 184 a?“iu MIHALACHE, Japan Atomic Energy Agency, F der Reactor Research and Development Center, ki, Tsuruga-shi, Fukui-ken, 919-1279, Tel: 0770 (ext 5211), email: mihalache.ovidiu@jaea.go.jp Member Non-Member Non-Member Non-Memberonsodium from tank, the eddy current effect from an artificial defect (outer groove, 10 mm wide and 20% from tube wall thickness) was measured in order to evaluate the sodium effect. In addition, after performing the experimental measurements, the sodium deposits located on the outside tube surface were measured and weighted to evaluate the sizes of sodium layer thickness. It was found out that for the Superheater tube the sodium layer thickness is between 2 and 6 um, while inside of defect the sodium deposit thickness was 40 um. The thickness of sodium deposit was greater for the evaporator tubes (up to 50 um) with a 50 um sodium layer located inside of the defect.A view of the sodium deposits on the evaporator and Superheater tubes are shown in Fig. 3 and 4.Fig 2. View of adhesion of sodium to the SG tubesFig 3. Sodium structures on the Evaporator tubesFig 4. Sodium structures on the Superheater tubes near thehelical support3 Numerical Model to Simulate Sodium ECT SignalNumerical simulations were conducted using a two-dimensional axisymmetric finite element code [2] which can model the electromagnetic interaction between electromagnetic field produced by the ECT coils system and ferromagnetic/austenitic SG steel tube and calculate the electromagnetic disturbance due to a defect in the tube.The presence of sodium layer on the outer tube surface modifies the electromagnetic signal during eddy current measurements due to the sodium high electrical conductivity (23.8x10o S/m). Based on the experimental measurements of the sodium layer thickness it was constructed a model of the ECT detection system that was validated with the outer groove signal (see Fig 5-7) in both absence or presence of sodium and under a tube SP [3].OD 20% 10 mm. 150 HzOD 2096 10 mm, 150 Hz (sodium layer 0.03 mm)ToExperiment --- SimulationExperiment --- SimulationImaginary [V]Imaginary [V]-10511.50.511.50 0.5 Real [V]-0. 5 0Real [V]Fig 5. Simulations and measurements of ECT signalfor outer groove evaporator185TIelical SP, 150 ILZHelical SP, 150 Hz (sodiam layer 0.03 mm)Experiment - SimulationExperiment ---- SinulationImagA-nary (V)Imaginary [VI-0.750.25-0.75 -.50.8 0.751.25 0.50.7510.250 0.25Reul [V-0.75 -0.5 -0.25 0Real IVFig 6. Simulations and measurements of ECT signalfor helical SP of evaporatorSP- SP + defect- SP + defectIm [V7Im [V]TEXPERIMENT 20 kHzSIMULATION 20 KHzRe[V]Re[V Fig 7. Simulations and measurements of ECT signal for outer groove under straight SP of Superheater in thepresence of sodium layerIn a previous paper was also confirmed by matching numerical simulations with experimental measurements that the sodium layer thickness adhering to the Evaporator tubes is 50 um [4]. However, it was not evaluated the thickness of the sodium layer for Superheater tubes. In the present analysis, the numerical model was relatively well confirmed by the experimental measurements in various situations for both Superheater and Evaporator, using axisymmetric FEM simulations, as: a) defect on the free tube area; b) defect under helical or straight tube SP ; c) absence or presence of a thin sodium layer.3. Estimation of phase-amplitude defect signal due to variation of sodium layer thicknessDespite the small values of the sodium layer thickness, their variations can influence the defect signal. However, experimental measurements can be performed only in a limited number of cases, raising the question of theextrapolation of the results to other types of defects with variable depths and widths. Based on the numerical electromagnetic model validated in both situations with sodium/without sodium layer, it were performed finite element simulations for outer circumferential defects with depths up to 50% from the tube wall thickness and length ranging from 0.25 to 15 mm. In the model it was consider that the thickness of sodium layer can varies between 20 and 50 um for the evaporator tubes and between 2 and 6 um for the Superheater tubes respectively. Due to the variable sodium layer thickness, the signal from an outer defect on the SG tube will change in amplitude and phase, depending also as how much sodium will fill the defect.Variations of Defect Amplitude due to sodium layer (150 Hz)Variations of Defect Phase due to sodium layer (150 Hz)1.5 deg4.106defect depth [6]defect depth [ool19 3.361.8 deg2224$ $ $ 19 22 24 defect width (mm)4 6 8 10 12 defect width (im)Variations of Defect Amplitude dne to sodium layer (450 Hz)Variations of Defect Phase due to sodium layer (450 Hz)4.5 deg9.20.defect depth [6]defect depth [%]5.2 deg6 deg0.08910.2 deg2244 6 8 10 12 defect width (mm)defect width (nim)Fig 8. Amplitude and phase of defect signal in evaporatortube covered by a thin 20-50 um sodium layerIn Fig. 8 is shown the variations in the defect signal amplitude and phase when the sodium layer thickness adhering to Evaporator change from 20 to 50 um. The error in the phase and amplitude signal increases as the frequency of excitation system is increasing from 150 to 450 Hz. It can also be observed that there is a small variation in both amplitude and phase change for different types of defect, so186it can be assumed that phase and amplitude change constantly for all defectsVariatious of Defect Amplitude due to sodium layer (10 kHz)Variatious of Defect Phase due to sodium layer (10 kHz)to the variations in frequencies.0.9 deg0.995Conclusiondefect depth [6]defect depth [6]defol2.81.9 deg1905/07/02 12:00:001.40.9 deg0.095Numerical ECT sin absence/presence of surface were confirndefect width [mm]defect width [mm]Variations of Defect Amplitude due to sodium layer (20 kHz)Variatious of Defect Phase due to sodium layer (20kHz)5 deg0.011defect depth too)defect depth [6]soli deg10 deg15 deg 20 deg1.730 deg Lagudefect width [mm]defect width [mm]Variations of Defect Amplitude due to sodium Layer (40 kHz)Variations of Defect Phase due to sodium layer (40 kHz)Referencessoh1.8A°5.6 deg3.3A°defect depth (0)defect depth [%]11.2 deg[1] ***, Nonde Volume 4, Ele 1986. [2] O. Mihalac Analysis of Electromagnet0.0182144 6 3 10 12 defect width [mm]AA 2 4 6 8 10 12 14defect width (mm)Fig 9. Amplitude and phase of defect signal in Superheatertube covered by a thin 2-6 um sodium layerFig. 9 shows the variation in phase and amplitude of signal for defects in Superheater tubes at 10, 20, 40 and 80 kHz frequencies. The phase of the defect signal is more sensitive to the variations in the sodium layer thickness at higher frequencies.Numerical ECT simulations of the defect signal in the absence/presence of sodium layer on the outer SG tube surface were confirmed by the experiment measurements in various situations, validating the numerical electromagnetic model of problem. The ECT signal from outer defects change due variations in sodium layer thickness. It was found, using FEM simulations, that phase of signal in Superheater tubes can change between 5 and 25 degrees for various sodium layer thicknesses. The amplitude and phase of defects in Evaporator tubes remain relatively constant for all class of defects and increase at higher inspection frequency[1] ***, Nondestructive Testing Handbook (second edition), Volume 4, Electromagnetic Testing, ISBN 0-931403-01-4, 1986. [2] O. Mihalache, a??Advanced Remote Field Computational Analysis of Steam Generators Tubesa??, IOS Press, Electromagnetic Nondestructive Evaluation VII, Studies in Applied Electromagnetics and Mechanics Vol. 26, pp 220-227, 2006. [3] O. Mihalache, T. Yamaguchi, S. Miyahara, T. Yamashita, a??ECT signals of SG tubes Covered by Sodium using a Sodium-Loop Mockup Test (II) Validation of Multi-Frequency ECT Algorithms to Suppress SG support Plate Signala??, Annual Meeting of the Atomic Energy Society Japan, Kokura, 27-28 September, 2007. [4] 0. Mihalache, K. Yokoyama, M. Ueda and T. Yamashita, a??Evaluation of the Effect of Sodium in Steam Generator Tubes using Remote Fielda??, Studies in Applied Electromagnetics and Mechanics, Vol. 24, pp.223-230, 2004.187 a?“
“ “Simulations and Measurements of Sodium Effecton the ECT Signal for SG Tubes of FBR“ “Ovidiu MIHALACHE,Toshihiko YAMAGUCHI,Masashi UEDA,
A small test sodium cylindrical tank, 2.2 m high and 0.6 m in diameter was filled with sodium at 500 degrees (see Fig. 1).In fast breeder reactors with sodium coolant, the integrity of steam generator (SG) tubes plays an important role in the reactor safety. These tubes are bundled in a cylindrical vessel with sodium located outside of the tube and water flowing inside of it. Due to the high chemical reactivity of sodium with water, penetration of SG tubes wall have to be avoided by all means and therefore, earlier defect detection before wall penetration during the in-service inspection (ISI) is required to increase the safety of SG tube component. Eddy current technique (ECT) is used to inspect the SG tubes [1], after sodium from the vessel is drained and the SG tubes are cooled down to the room temperature. After avoided by all means and therefore, earlier defect detection before wall penetration during the in-service inspection (ISI) is required to increase the safety of SG tube component. Eddy current technique (ECT) is used to inspect the SG tubes [1], after sodium from the vessel is drained and the SG tubes are cooled down to the room temperature. After sodium with water, penetration of SG tubes wall have to be avoided by all means and therefore, earlier defect detection before wall penetration during the in-service inspection (ISI) is required to increase the safety of SG tube component. Eddy current technique (ECT) is used to inspect the SG tubes [1], after sodium from the vessel is drained and the SG tubes are cooled down to the room temperature. After draining and cooling, there is a thin sodium layer adhering to the outer surface of SG tubes. Due to the high electrical conductivity of sodium layer, the eddy current signal will be modified. In the present paper, the authors investigate the effect of sodium on several types of outer tube defects with variable depths and widths using numerical finite element simulations. The numerical model of sodium is build after validation of experimental results using a test sodium tank experiment. The phase-amplitude diagram is constructed for defects with depths up to half of the tube wall thickness and 15 mm wide.Fig 1. View of the test sodium tankOvidiu MIHALACHE, Japan Atomic Energy Agency, Fast Breeder Reactor Research and Development Center, 1 Shiraki, Tsuruga-shi, Fukui-ken, 919-1279, Tel: 0770 39 1031 (ext 5211), email: mihalache.ovidiu@jaea.go.jpTwo SG tubes, one made of austenitic stainless steel (the Superheater tube) and the other made of Cr-Mo alloy (the Evaporator tube), were introduced in the tank and kept for more than two hours, in order to have a good wetting phenomenon (see Fig 2). After draining and cooling of the- 184 a?“iu MIHALACHE, Japan Atomic Energy Agency, F der Reactor Research and Development Center, ki, Tsuruga-shi, Fukui-ken, 919-1279, Tel: 0770 (ext 5211), email: mihalache.ovidiu@jaea.go.jp Member Non-Member Non-Member Non-Memberonsodium from tank, the eddy current effect from an artificial defect (outer groove, 10 mm wide and 20% from tube wall thickness) was measured in order to evaluate the sodium effect. In addition, after performing the experimental measurements, the sodium deposits located on the outside tube surface were measured and weighted to evaluate the sizes of sodium layer thickness. It was found out that for the Superheater tube the sodium layer thickness is between 2 and 6 um, while inside of defect the sodium deposit thickness was 40 um. The thickness of sodium deposit was greater for the evaporator tubes (up to 50 um) with a 50 um sodium layer located inside of the defect.A view of the sodium deposits on the evaporator and Superheater tubes are shown in Fig. 3 and 4.Fig 2. View of adhesion of sodium to the SG tubesFig 3. Sodium structures on the Evaporator tubesFig 4. Sodium structures on the Superheater tubes near thehelical support3 Numerical Model to Simulate Sodium ECT SignalNumerical simulations were conducted using a two-dimensional axisymmetric finite element code [2] which can model the electromagnetic interaction between electromagnetic field produced by the ECT coils system and ferromagnetic/austenitic SG steel tube and calculate the electromagnetic disturbance due to a defect in the tube.The presence of sodium layer on the outer tube surface modifies the electromagnetic signal during eddy current measurements due to the sodium high electrical conductivity (23.8x10o S/m). Based on the experimental measurements of the sodium layer thickness it was constructed a model of the ECT detection system that was validated with the outer groove signal (see Fig 5-7) in both absence or presence of sodium and under a tube SP [3].OD 20% 10 mm. 150 HzOD 2096 10 mm, 150 Hz (sodium layer 0.03 mm)ToExperiment --- SimulationExperiment --- SimulationImaginary [V]Imaginary [V]-10511.50.511.50 0.5 Real [V]-0. 5 0Real [V]Fig 5. Simulations and measurements of ECT signalfor outer groove evaporator185TIelical SP, 150 ILZHelical SP, 150 Hz (sodiam layer 0.03 mm)Experiment - SimulationExperiment ---- SinulationImagA-nary (V)Imaginary [VI-0.750.25-0.75 -.50.8 0.751.25 0.50.7510.250 0.25Reul [V-0.75 -0.5 -0.25 0Real IVFig 6. Simulations and measurements of ECT signalfor helical SP of evaporatorSP- SP + defect- SP + defectIm [V7Im [V]TEXPERIMENT 20 kHzSIMULATION 20 KHzRe[V]Re[V Fig 7. Simulations and measurements of ECT signal for outer groove under straight SP of Superheater in thepresence of sodium layerIn a previous paper was also confirmed by matching numerical simulations with experimental measurements that the sodium layer thickness adhering to the Evaporator tubes is 50 um [4]. However, it was not evaluated the thickness of the sodium layer for Superheater tubes. In the present analysis, the numerical model was relatively well confirmed by the experimental measurements in various situations for both Superheater and Evaporator, using axisymmetric FEM simulations, as: a) defect on the free tube area; b) defect under helical or straight tube SP ; c) absence or presence of a thin sodium layer.3. Estimation of phase-amplitude defect signal due to variation of sodium layer thicknessDespite the small values of the sodium layer thickness, their variations can influence the defect signal. However, experimental measurements can be performed only in a limited number of cases, raising the question of theextrapolation of the results to other types of defects with variable depths and widths. Based on the numerical electromagnetic model validated in both situations with sodium/without sodium layer, it were performed finite element simulations for outer circumferential defects with depths up to 50% from the tube wall thickness and length ranging from 0.25 to 15 mm. In the model it was consider that the thickness of sodium layer can varies between 20 and 50 um for the evaporator tubes and between 2 and 6 um for the Superheater tubes respectively. Due to the variable sodium layer thickness, the signal from an outer defect on the SG tube will change in amplitude and phase, depending also as how much sodium will fill the defect.Variations of Defect Amplitude due to sodium layer (150 Hz)Variations of Defect Phase due to sodium layer (150 Hz)1.5 deg4.106defect depth [6]defect depth [ool19 3.361.8 deg2224$ $ $ 19 22 24 defect width (mm)4 6 8 10 12 defect width (im)Variations of Defect Amplitude dne to sodium layer (450 Hz)Variations of Defect Phase due to sodium layer (450 Hz)4.5 deg9.20.defect depth [6]defect depth [%]5.2 deg6 deg0.08910.2 deg2244 6 8 10 12 defect width (mm)defect width (nim)Fig 8. Amplitude and phase of defect signal in evaporatortube covered by a thin 20-50 um sodium layerIn Fig. 8 is shown the variations in the defect signal amplitude and phase when the sodium layer thickness adhering to Evaporator change from 20 to 50 um. The error in the phase and amplitude signal increases as the frequency of excitation system is increasing from 150 to 450 Hz. It can also be observed that there is a small variation in both amplitude and phase change for different types of defect, so186it can be assumed that phase and amplitude change constantly for all defectsVariatious of Defect Amplitude due to sodium layer (10 kHz)Variatious of Defect Phase due to sodium layer (10 kHz)to the variations in frequencies.0.9 deg0.995Conclusiondefect depth [6]defect depth [6]defol2.81.9 deg1905/07/02 12:00:001.40.9 deg0.095Numerical ECT sin absence/presence of surface were confirndefect width [mm]defect width [mm]Variations of Defect Amplitude due to sodium layer (20 kHz)Variatious of Defect Phase due to sodium layer (20kHz)5 deg0.011defect depth too)defect depth [6]soli deg10 deg15 deg 20 deg1.730 deg Lagudefect width [mm]defect width [mm]Variations of Defect Amplitude due to sodium Layer (40 kHz)Variations of Defect Phase due to sodium layer (40 kHz)Referencessoh1.8A°5.6 deg3.3A°defect depth (0)defect depth [%]11.2 deg[1] ***, Nonde Volume 4, Ele 1986. [2] O. Mihalac Analysis of Electromagnet0.0182144 6 3 10 12 defect width [mm]AA 2 4 6 8 10 12 14defect width (mm)Fig 9. Amplitude and phase of defect signal in Superheatertube covered by a thin 2-6 um sodium layerFig. 9 shows the variation in phase and amplitude of signal for defects in Superheater tubes at 10, 20, 40 and 80 kHz frequencies. The phase of the defect signal is more sensitive to the variations in the sodium layer thickness at higher frequencies.Numerical ECT simulations of the defect signal in the absence/presence of sodium layer on the outer SG tube surface were confirmed by the experiment measurements in various situations, validating the numerical electromagnetic model of problem. The ECT signal from outer defects change due variations in sodium layer thickness. It was found, using FEM simulations, that phase of signal in Superheater tubes can change between 5 and 25 degrees for various sodium layer thicknesses. The amplitude and phase of defects in Evaporator tubes remain relatively constant for all class of defects and increase at higher inspection frequency[1] ***, Nondestructive Testing Handbook (second edition), Volume 4, Electromagnetic Testing, ISBN 0-931403-01-4, 1986. [2] O. Mihalache, a??Advanced Remote Field Computational Analysis of Steam Generators Tubesa??, IOS Press, Electromagnetic Nondestructive Evaluation VII, Studies in Applied Electromagnetics and Mechanics Vol. 26, pp 220-227, 2006. [3] O. Mihalache, T. Yamaguchi, S. Miyahara, T. Yamashita, a??ECT signals of SG tubes Covered by Sodium using a Sodium-Loop Mockup Test (II) Validation of Multi-Frequency ECT Algorithms to Suppress SG support Plate Signala??, Annual Meeting of the Atomic Energy Society Japan, Kokura, 27-28 September, 2007. [4] 0. Mihalache, K. Yokoyama, M. Ueda and T. Yamashita, a??Evaluation of the Effect of Sodium in Steam Generator Tubes using Remote Fielda??, Studies in Applied Electromagnetics and Mechanics, Vol. 24, pp.223-230, 2004.187 a?“
“ “Simulations and Measurements of Sodium Effecton the ECT Signal for SG Tubes of FBR“ “Ovidiu MIHALACHE,Toshihiko YAMAGUCHI,Masashi UEDA,