An analysis on eddy current flowmeter - A review

Eddy Current Flowmeter – A Review , S.Poornapushpakala, C.Gomathy I.Sylvia ,B.Krishnakumar P.Kalyanasundaram IGCAR,Kalpakkam,India [email protected] [email protected] [email protected] Sathyabama University chennai,India [email protected], [email protected], [email protected] these system coupling to the fluid is via the relatively weak external field of the coils [1]. Abstract - Measurement of flow in liquid metal systems exhibit unusual problems because of the severe operating environments like high operating temperatures and the chemical reactivity of most liquid metals. The liquid metal flowmeters should have the capacity to withstand high gamma radiation levels and rapid thermal transients. The commonly used flowmeters in liquid metal systems are permanent magnet Flowmeter, direct current electromagnet Flowmeter and alternating current electromagnet flowmeter. In Alternating current electromagnet Flowmeter there are three types namely transverse field Flowmeter, Eddy Current Flowmeter and E-Core Flowmeter. The subject of analysis is the Eddy current Flowmeter. Eddy currents are generated when a moving conductor experiences changes in the magnetic field generated by a stationary object, as well as when a stationary conductor encounters a varying magnetic field. A flowmeter which works on this principle is the Eddy Current Flowmeter. This paper gives a thorough survey on eddy current flowmeter, its construction, simulation results and its performance under various test conditions in different organization I. II. PRINCIPLE In the design suggested by Lehde and Lang, the coils are arranged inside a streamlined capsule located in the flow stream [1]. In P.F.R design, the coil system and connecting leads are to be placed inside a thimble. In these system coupling to the fluid is via the relatively weak external field of the coils [1]. Shercliffe suggested a design which will allow the fluid to flow through the coils, where the magnetic field is strongest, appears to provide a better coil-fluid coupling.The eddy-current probe type flow sensor developed by United States Atomic Energy Commission in the early 1970’s consists of three coils formed on a metallic bobbin enclosed in a protective housing. The sensor is isolated from the liquid metal by a sealed inert thimble. Figure 1.0 shows the sensor – thimble arrangement which is used for housing the flow sensor that isolates the flow sensor from the coolant [3]. The primary coil is divided at the center to serve the connection of the coil wire to the sheathed leads. The primary winding is excited by constant current source of 0.5 A at a frequency of 1000Hz [3]. INTRODUCTION The Eddy Current Flow Meters (ECFM) are much suitable for flow measurement in high-temperature liquid metal systems.Deterioration of the properties of permanent magnets and other ferromagnetic parts may seriously worsen the performance of other types of flowmeters like permanent magnet flowmeter[1,2]. In order to measure the coolant flow rate through the individual fuel subassemblies in a Nuclear Reactor, a flow measurement device fitted with eddy current flow sensor has been developed by many research and development laboratories.Lehde and Lang patented the flow measurement based on eddy current principle in 1948[1].This concept was proposed as a flow failure alarm in the British Prototype Fast Reactor in the early 1960s [1]. Later the eddy current flowmeter has been developed and used in Hanford Engineering Development Laboratory (US), Argonne National Laboratory (US) and in Indira Gandhi Centre for Atomic Research (India) for monitoring the flow of sodium through the core in a Nuclear Reactor. In O-arai Engineering center (Japan) the Eddy current flowmeter is used for void detection at the LMFBR core exit. In the design suggested by Lehde and Lang, the coils are arranged inside a streamlined capsule located in the flow stream [1]. In P.F.R design, the coil system and connecting leads are to be placed inside a thimble. In 978-1-4244-918-4/10/$26.00©2010IEEE CONSTRUCTION AND WORKING Figure 1.0 Sensor –Thimble Arrangement (Courtesy [3]) The eddy current flowmeter developed in the Department of Atomic Energy, India is employed in the Primary Sodium 214 Pump (PSP) of the PFBR to measure pump flow. The flow sensor consists of three coils which are wound on a magnetic core and assembled in a Stainless Steel thimble. Figure 2.0 shows the schematic of ECFM.The primary coil (P1) is excited by a constant current source of 0.2A at a frequency of 700Hz [4].The two secondary coils (S1 & S2) are wound symmetrically on either side of the primary on the same iron bobbin. Figure 4.0 III. Figure 2.0 A Schematic of ECFM (courtesy P.Sharma et al.[10]) P1 is primary coil,S1 is upstream side secondary Coil,S2 is downstream side secondary coil TESTING & PERFORMANCE ANALYSIS Dry test and sodium loop test: Tests on an eddy-current flowmeter were performed with a high temperature model. Wiegand of Argonne National Laboratory performed the dry test by moving an aluminum rod through the flow channel at controlled and measured velocity. From the Dry test the Theoretical and Experimental results were found to be matching except some slight deviation and it is found that the sensor gives a maximum sensitivity and an in-phase signal at a particular excitation frequency [2]. In sodium loop test the ECFM was exposed directly to the sodium. In sodium loop test the deviation of test and theoretical values were more than the dry test. The residual coupling between the primary and the secondary coils was high. This introduced a residual signal which leads to measurement errors even though external balance arrangements were made [2]. But the optimum frequency giving maximum sensitivity was found to be close to the theoretical value. Figure 5.0 and 6.0 shows the flowmeter sensitivity from Dry and Sodium loop test respectively. It was inferred from these tests that the ECFM was most sensitive to the fluid velocities in the channel wall region and the sensitivity to fluid velocities decreased as it approached the channel axis [2]. Hence it was proved that the sensitivity is a function of fluid velocities besides the excitation frequency. The eddy current flowmeter used in Power Reactor and Nuclear Fuel Development Corporation, Japan consists of three or five coils wound on a bobbin in which the magnetic core made of pure iron is placed. The probe is inserted into a 316 SS well which protects the probe from sodium. Nakamoto et al. [5] found that the optimum excitation frequency was 375Hz for MK-3 Probe which is used for void detection in sodium flow. The eddy-current flowmeter applies the distortion of the imposed magnetic field as the means of measuring flow rate. When the eddy current flowmeter sensor is placed in a pipe filled with static sodium and if primary coil is excited then equal transformer voltage is generated in the secondary coils because of equal flux linkage due to symmetry. Since the two secondary coils are connected in phase opposition, the resultant output voltage is zero. Figure 3.0 shows the series opposition connection of the secondary coils. This figure illustrates the polarities at a fixed time in a cycle [3].Figure 4.0 shows the eddy current components in the coils and the flow of liquid metal. Figure 3.0 Series opposition connection of Secondary coils (Courtesy [3]) If the eddy current flowmeter sensor is placed in a flowing sodium, the induced voltages in the both the secondary are not equal because the induced voltage is subtractive to transformer voltage in upstream secondary coil and additive to transformer voltage in downstream secondary coil hence the resultant voltage is proportional to sodium flow [7,10]. Figure 5.0 Flowmeter Sensitivity from Dry Test 215 secondary outputs were recorded at different excitation frequencies from 200Hz to 1000Hz [7].The procedure were repeated for various sodium temperatures of 400 ºC and 500 ºC. From the test results it was found that the sensor with pure iron former surrounded by SS 410 shield gives good signal output. Temperature error 250ºC to 500ºC ranges is ± 2 % of full scale. The optimum excitation frequency was observed that 400Hz from the frequency response test [7]. IV. ELECTRONICS OF ECFM The electronics of the sensor requires a drive unit and a signal processing unit with temperature compensating circuits. The drive unit gives a constant current output at constant frequency, to the primary coil. The signal processing unit consists of pre amplifiers, low pass filters, and low pass filter was 0.15s followed by micro controller circuits which gave an overall response time of 1.5s. To minimise response time analog circuits or high speed micro controller were planned [4]. Figure 6.0 Flowmeter Sensitivity from Sodium Loop Test In FFTF Probe type eddy current flowmeter the dry testing were performed by keeping the ECFM in dry design test housing where the sensor is shielded from sodium by thimble made of stainless steel [6]. In wet test design housing the ECFM was exposed directly to sodium. Testing in both the cases were made at different sodium flow rates from 0 to 600gpm and it was exposed to various temperatures ranging from 600 ºF to 1100 ºF. Figure 7.0 shows the electronics used for the performance test of ECFM. The electrical unbalance of ECFM was nullified by adjusting the voltage and gain of the amplifier [6]. It was found from the test results that the overall average dc output signal from dry design test housing was 62% of that in the wet design test housing [6]. Also the overall average sensitivity of dry test housing was 44% of that in wet test housing. Signal from secondary coil S1 PreAmplifier Precision Rectifier RC Filter Signal from secondary coil S2 PreAmplifier Precision Rectifier RC Filter Figure 8.0 Block diagram of signal processing unit V. SIMULATION OF ECFM Finite Element Method (FEM) is an important tool in the design of electromagnetic devices. The ECFM was modeled in FEM for both static and dynamic analysis [8,10]. For modeling SS 410 and Fe-Ni were selected as core and shield material respectively. The model was subjected to a transient analysis of various sodium velocities at temperatures of 400 and 500°C. The result exhibited a low eddy current generation in the down stream coil than in upstream coil. Veerasamy et al. [7] compared the calculated value from FEM analysis with the actual measured values in sodium and inferred that the differences in these values could be attributed to variation of the properties with temperature. Figure 7.0 ECFM Excitation & amplifier Block Diagram In PFBR the ECFM was subjected to sodium testing in two phases where flow was measured with reference to a permanent magnet flowmeter [7]. In first phase testing the shield material used was Fe-Ni tube and in second phase SS410 tube. The phase I testing was performed with constant primary excitation of 200mA, 433Hz.The testing was done at various sodium temperature ranging from 250ºC to 500ºC.The secondary outputs were recorded for different sodium flow rates from 0 to 30 m3/hr. In phase II testing the sodium flow rate was made constant at 20 (m3/hr) and at 300ºC.The Eddy current flowmeter (ECFM) has been modeled using FEM based software COMSOL 3.4 [8,9,10] The model was a 2D axi-symmetric model which incorporates both the Electromagnetics and Navier-Stokes modules. The actual coils of ECFM was made of Mineral Insulated Cable but in the simulation, modeling of mineral insulated cable was not possible so cable has been approximately modeled using zero conductivity (actual coils will have some conductivity) and 216 specified surface current over coil area. Due to the mismatch of the coils some difference in the actual values of the induced voltages were observed in the simulation. But the difference of the coil voltages remained the same [10]. Using this model simulation was performed for various sodium velocities at temperatures of 400 & 500°C with the constant primary excitation of 200mA and at a constant frequency of 400Hz. It was observed that the simulation result show a close agreement with the experimental results except at static sodium. When there is no sodium flow the experimental result gave some voltage due to the misalignment of the coils, but the COMSOL simulation does not give any such output. Figure 2.0 and Figure 3.0 shows the comparison chart between the experimental and simulation results of ECFM at 500°C & 400°C respectively. Prashant Sharma et al.[8] also compared the analysis of ECFM using ANSOFT (2D/3D electromagnetic field simulation software) and COMSOL (Multiphysics Modeling and Simulation software). According to Vidal et al. [11] simulation of FEM analysis of an eddy current water Flowmeter was done using MagNet which a FEM based electromagnetic field simulation software for low frequency applications. laboratories of nuclear power plants. The simulation using COMSOL was made for iron and SS410 bobbin. The misalignment of coils was compensated by suitable electronics circuitry for matching the zero output at static sodium. The electronics of ECFM comprises of a signal conditioning circuit and a control logic circuit. The overall response time of the electronics was 1.5s. An objective of minimizing the response time to less than 100ms was taken up. Design of the signal conditioning circuit using ORCAD software is being developed for reducing the response time. ACKNOWLEDGMENT This work was supported by Indira Gandhi Centre for Atomic Research, Department of Atomic Energy, Kalpakkam, India. The Authors are thankful for the support extended by the Director of IGCAR, Dr.Baldev Raj. REFERENCES [1]. David E.Wiegand - ‘Summary of an Analysis of the Eddy current flowmeter’ :IEEE Transaction on Nuclear Sciences Vol 15,Issue 1 February 1968 [2]. David E. Wiegand & Charles W. Michels - ‘Performance tests on an eddy-current flowmeter’,IEEE Transactions on Nuclear Science Vol.16, Issue 1, February 1969. [3]. Eddy-Current Probe Type Flow Sensor For Liquid Metal ServiceUnited States Atomic Energy Commission Division of Reactor Research and Development, June 1973[On line] [4]. R.Veerasamy,C.Asokane,K.Narayanan,R.Dhanasekaran,K.Swaminathan and R.Prabhakar – ‘Eddy current flowmeter for incore flow measurement in fast reactors’,8th National seminar on Physics Technology of sensors (NSTPS-8) in Feb.2001 at IGCAR,Kalpakkam. [5]. K.Nakamoto,S.Tamura,K.Ishii,H.Kuwahara,N.Ohayama and T.Muramatsu – ‘Application of an Eddy-current type flowmeter to Void detection at LMFBR Core exit’,Nuclear Engineering and Design 82 (1984) 393-404 [6]. T. J. Costello , R. L. Laubham, W. R. Miller, C. R. Smith –‘FFTF probe type eddy current flowmeter-Wet Vs Dry performance evaluation in sodium’,American Nuclear Society International Meeting,Washington D.C,December,1972 Figure 9.0 Comparison of Experimental & Simulation result at 500°C [7]. Veerasamy R., Sureshkumar S., Asokane C., Sivakumar N.S, Padmakumar G., Dash S.K., Sreedhar B.K., Chandramouli S., Gurumoorthi K. and Vaidyanathan G. - ‘Eddy Current Flow Sensor Development And Testing For LMFBR Sodium Pumps’, 15th International Conference on Nuclear Engineering Nagoya, Japan, April22-26,2007,ICONE15-10213 [8]. Prashant Sharma, S.K.Dash, B.K.Nashine, S. Suresh Kumar, R. Veerasamy, B. Krishnakumar, P. Kalyanasundaram, G. Vaidyanathan – ‘ Eddy Current Flowmeter for Sodium Flow Measurement and its Performance Prediction’, FRTG/IGCAR/NDE-2008. [9]. Prashant Sharma, S.Suresh Kumar, B.K.Nashine, R. Veerasamy, B. Krishnakumar, P. Kalyanasundaram, G. Vaidyanathan – ‘Development,computer simulation and performance testing in sodium of an eddy current flowmeter’,Annals of Nuclear Energy 37 (2010) 332338. Figure 10.0 Comparison of Experimental & Simulation result at 400°C VI. CONCLUSION [10]. Prashant Sharma, S.K.Dash, B.K.Nashine, S. Suresh Kumar, R. Veerasamy,B. Krishnakumar, P. Kalyanasundaram, G. Vaidyanathan – ‘ Performance Prediction of Eddy Current Flowmeter for Sodium’, Proceedings of the COMSOL Conference 2009 Bangalore. The Eddy Current Flow Meter was considered for the analysis since it is much suitable for flow measurement in high-temperature liquid metal systems. This paper reviewed the construction & working principle of ECFM at various [11]. N. Vidal,G. Aguirre-Zamallo,J.M.Barandiaran, A. Garcia-Arribas, J. Gutierrez –‘ FEM analysis of an eddy current water flow meter’, Journal of Magnetism and Magnetic Materials 304 (2006) e838–e840. 217 218