Avalable onlne www.jocpr.com Journal of Chemcal and Pharmaceutcal Research, 204, 6(5): 520-526 Research Artcle ISS : 0975-7384 CODE(USA) : JCPRC5 Dgtal Electrcal Resstance Tomography System and ts Expermental Research Lfeng Zhang* orth Chna Electrc Power Unversty, Chna ABSTRACT Electrcal resstance tomography (ERT) s a new technque developed n recent years, whch ams to measure mult-phase flow. The conductvty dstrbuton mages of cross secton can be obtaned, and then the gas holdup can be calculated usng ERT system. Combnng ERT technque and cross correlaton measurement method, the velocty can be calculated. In ths paper, the dual-plane dgtal ERT system was desgned, and experments were carred out on a small-scale gas/water two-phase flow expermental set-up. The on-lne reconstructed mages, gas holdup estmatons and correlaton velocty can be obtaned. Expermental results showed that the reconstructed mages wth hgh spatal resoluton and real-tme performance can be obtaned. The software has rch functons and user-frendly nterface, whch can meet the needs of vsualzed measurement for two-phase flow n practcal applcaton. Key words: Two-phase flow; Electrcal resstance tomography; Gas holdup; Cross correlaton ITRODUCTIO Industral flow processes are complex n nature, whch often nvolve a varety of components n a combnaton of gas, lqud and sold phases. Measurements of the phase dstrbuton and nterfaces n a multphase flow are very mportant for the process []. Process tomography (PT) technque s a new technque that has developed rapdly n recent years and whch has great potental and wde ndustral applcaton prospect for drect characterstcs analyss for multphase flows [2]. Electrcal resstance tomography (ERT) s one nd of PT technque and has been proved to be a powerful tool for mappng the concentraton and velocty dstrbutons n two-phase flow [3, 4]. ERT s used when the contnuous phase s conductve n two-phase flows, such as gas-water two-phase flow n vertcal ppe, through whch the cross-sectonal conductvty dstrbuton mages can be obtaned at two adjacent planes along the ppelne. Then, reconstructed mages can be cross-correlated to obtan the velocty profle of the flow. The non-ntrusve measurement of velocty profles wthn process equpment s a challengng tas because few nstruments can measure the velocty profle of a mult-phase flow n a ppelne. ERT provdes a unque opportunty to do so by cross correlaton usng a dual-plane ERT sensor [5]. In ths paper, the structure of the dgtal dual-plane ERT system was descrbed. The method for measurng the gas holdup and the velocty based on the gas holdup cross correlaton was presented. Experments were carred on usng the desgned dgtal dual-plane ERT system. Expermental results showed that the reconstructed mages wth hgh spatal resoluton and real-tme performance can be obtaned. 520
DIGITALERT SYSTEM PRICIPLE OF ERT A typcal ERT system s shown n Fg., whch ncludes the array sensor, the data acquston system (DAS) and the mage reconstructon and dsplay unt. In ths paper, there were 6 rectangle stanless steel electrodes wth the wdth of 0 mm and the heght of 20 mm for each plane. The voltage sgnals between adjacent measurement electrodes were measured by DAS. Then, these data were sent to mage reconstructon computer. Usng mage reconstructon algorthm, the reconstructed mages can be obtaned. Fg. : Structural dagram of ERT ARCHITECTURE OF DIGITAL ERT SYSTEM The modular functonal desgn method was adopted for the dgtal ERT system, mplemented as a set of hardware crcut card unts and software modules. Data from the dual-plane ERT sensor were ntegrated by a host PC. An archtecture vew for the twn-plane ERT system s shown n Fg. 2. Fg. 2: The ERT system platform It ncludes common functons as follows. drect dgtal syntheszer (DDS) whch s realzed by Feld-Programmable Gate Array (FPGA) and measurement (MEAS), whch usually nclude exctaton sgnal generaton and sgnal condton; multplexng (MUX), for routng exctng and measurement sgnals; power supply (PWS), to provde electrcal power for the hardware system. The varous cards are all organzed usng a custom-defned bacplane. Control sgnals are provded from a FPGA chp that embraces DAS and communcaton functons. Data were transferred to the host PC through the Unversal Seral Bus (USB) nterface. The dgtal ERT system for dual-plane measurements can be shown n Fg. 3. Fg. 3: The dgtal dual-plane ERT system 52
For the desgned dgtal ERT system, the snusodal current (Hz~MHz, adjustable) exctaton sgnal s generated by FPGA, whch s sent to MEAS card to excte a par of adjacent electrodes and the resultng potental dfference between another par of adjacent electrodes s measured, whch are also called projected data. The current source s fed va the bus to the multplex to select exctng electrodes. Each measurement card supports up to four electrodes. Eght cards may be used to realze dual-plane 32 electrodes (6 electrodes for each plane). Measurements are routed to a 4 bt, 0 µs, samplng hgh precson analogue to dgtal converter. Resultng data are then routed va USB 2.0 nterface to the host PC for further processng. The data acquston rate of each plane s 500 frames/s for dgtal dual-plane ERT sensor, whle the onlne mage reconstructon speed s 20 frames/s for each plane [6]. SOFTWARE OF ERT SYSTEM Software of ERT system controls the operaton and data process, and was developed based on the vsual C++ 6.0 software. One typcal user nterface confguraton n Fg. 4 shows an onlne operaton for ERT system. Fg. 4: The onlne operaton software vew There are sx areas n the man GUI (Graphc user nterface). Area shows the real-tme cross-sectonal mages of each plane. The onlne real-tme mage reconstructon speed s 20 frames/s. Area 2 shows the calculated gas holdup for both planes. Area 3 provdes the calculated correlaton velocty curve. Area 4 s a man parameter confguraton panel for system operaton, ncludng exctng frequency, exctng and measurng mode, electrode number and data transfer mode. Area 5 s an algorthm parameter confguraton panel, ncludng teraton number, teraton factor and regularzaton factor. Area 6 s an mage dsplay confguraton panel, ncludng selecton of the mage color, grd and smooth effect, flter threshold, 3D rotaton, mage zoom and translaton. CALCULATIO OF GAS HOLDUP AD CORRELATIO VELOCITY The grey values of the reconstructed mages are normalzed from 0 to, whch delegate gas and lqud, respectvely. The grey value of each pxel corresponds to ts phase fracton value [7]. The gas holdup α can be estmated accordng to (): M α = g A A () = Where g and ppe. M s the total pxel number n cross secton of the ppe. A are the grey value and area of the thpxel, respectvely. A s the cross-sectonal area of the The prncple of cross-sectonal flow velocty measurement along the flow drecton s taggng sgnals generated by flow turbulence or suspended movng partcles. The prncple dagram of flow velocty measurement usng cross correlaton can be seen n Fg. 5. 522
Fg. 5: Flow velocty measurement usng cross correlaton method If the flow regme can be consdered frozen durng transton from plane to plane 2, and the dstance between plane and plane 2 s L, then the transt tme τ of the th pxel of the mage from plane to plane 2 can be calculated from the followng cross-correlaton functon [8] R ( j ) = x( ) y( + j ) j = 0,, 2,, m and m <, =, 2,, M (2) = Where stands for the samplng nterval of each mage, j s the dscrete delay tme and s the length of samplng mages for cross correlaton, M s the total number of the pxels n ppe, x (t) and y (t) are the grey values of the th pxels of plane and plane 2, respectvely. When the cross-correlaton functon reaches the pea value, the transt tme τ = j (3) τ can be obtaned. Where j corresponds to the pea value of the cross-correlaton functon. Then, the velocty v can be calculated accordng to (4). v L τ = (4) It s mportant to choose a sutable dstance L for cross correlaton [9]. Both the smlarty of sgnals and the dynamc behavours of the system should be taen nto consderaton. The velocty measurement based on pxel-pxel cross-correlaton has been adopted by many researchers to obtan velocty profle [0]. But the calculated velocty curve fluctuates very qucly based on ths method. In our experment, the cross-correlaton velocty measurement method based on gas holdup of the two planes was presented as follows: R ( j ) = = α (5) ( ) α2( + j ) Where α and α 2 are gas holdups of the plane and plane 2, respectvely. RESULTS EXPERIMETAL SETUP The expermental set-up was showed n Fg. 6. The experments on the ar water two-phase flow n the vertcal ppe were desgned to verfy the performance of dgtal ERT system. The experment devce s movable and contans the followng man parts: a Perspex ppe, two planes of 6-electrode sensng arrays, ERT DAS system, an ar compressor, four solenods and valves, the mage reconstructon computer, one flow rate meter whch s used to measure the water flow and one water pump wth an nverter by whch the speed of the motor can be regulated n order to control the water flow rate. 523
Fg. 6: Expermental set-up Water s crculated va the motor and the water pump. Water flow rate can be controlled n the range of 0 to m/s. Gas flow rate s controlled through four solenods and gas valves. The dstance between the two sensng planes s 50 mm. Bubble flow, slug flow and churn flow can be obtaned usng ths set-up. RECOSTRUCTED IMAGES AD GAS HOLDUP In ths experment, bubble flow and slug flow were tested. Reconstructed mages were shown at 20 ms/frame for each plane usng pre-teraton algorthm []. Fg. 7 showed the mages of bubble flow and slug flow over the magng plane at dfferent tme durng ther pass through the magng cross-secton. It s clearly that the temporal seres of the reconstructed mages can acceptably reflect the change of the cross-secton dstrbuton of the nvestgated two-phase flow. (a) (b) Fg. 7: Reconstructed mages: (a) Bubble flow, (b) Slug flow For bubble flow and slug flow, the moton of bubbles can be clearly seen from the bottom plane to top plane. Frst, the bubbles can be seen n bottom plane reconstructed mages, and then n the top plane reconstructed mages. The place of bubbles n cross secton changes durng ther motons. ERT provdes a new approach to vsualzng the complex moton of the bubbles n gas/water two-phase flow. 524
The gas holdup curve of each plane s shown n Fg. 8. It gves the real tme estmaton of gas holdup for each plane. The gas holdup curves of bottom plane and top plane are n good correlaton, and they reflect the gas change. Fg. 8: Curves of gas holdup for dual plane CORRELATIO VELOCITY The length of samplng mages s 00, and the real tme correlaton velocty curve s shown n Fg. 9. From Fg. 9, t can be seen that the correlaton velocty s n consstent wth the reconstructed mages, whch can reflects the velocty of the bubbles. Fg. 9: Curve of correlaton velocty The velocty s very low at pont A when there are few bubbles. When the flow regme s slug flow, the velocty reaches a pea value at pont B. Wth the decrease of bubbles, the velocty becomes lower at pont C and D. The velocty reaches another pea value at pont E, when the flow regme becomes slug flow agan. Because ths set-up s very small, to obtan more accurate velocty estmate, more calbraton wor should be done. But from Fg. 9, t s clear that the cross correlaton method based on the gas holdup s feasble for correlaton velocty estmaton. COCLUSIO The desgn of dgtal dual-plane ERT system and the small experment set-up were descrbed n ths ppe. The ERT system has standard modules and s easy to be expanded. Ths system has been employed successfully n ths small expermental set-up. 525
Experment results showed that the reconstructed mages can reflect the change of the cross-secton dstrbuton of the gas/lqud two-phase flow, the estmaton of the gas holdup can be obtaned and used n ndustry control. The presented cross correlaton velocty measurement method base on gas holdup was feasble. Further wor wll focus on mprovements of the precson of the estmatons of the gas holdup and correlaton velocty. Calbraton experments on large-scale experment set-up wll be done. Acnowledgments The author wshes to than the atonal atural Scence Foundaton of Chna for contract5306058 and the Fundamental Research Funds for the Central Unverstes for contract204ms42, under whch the present wor was possble. REFERECES [] W Warsto; L-S Fan.Chem. Eng. Sc.,200, 56(), 6455-6462. [2] MS Bec; RA Wllams.Meas. Sc. Technol., 996,7(3), 25-224. [3] H Jang; M Wang; RA Wllams.Chem. Eng. J., 2007, 30(6), 79-85. [4] M Wang;ADorward; R Mann.Chem. Eng. J., 2000, 77(4),93-98. [5] HX Wang; LF Zhang.Meas. Sc. Technol., 2009, 20(), 4007 (8pp). [6] X Deng; F Dong; LJ Xu; XP Lu; LA Xu.Meas. Sc. Technol., 2009, 2(8), 024-03. [7] ZY Huang; BL Wang; HQ L.IEEE. Trans. Instrum. Meas.,2003, 52(2), 7-2. [8] S Lu; Q Chen; HG Wang; F Jang; I Ismal; WQ Yang.Flow Meas. Instrum., 2005, 6(4-6),35-44. [9] U Datta; T Dyaows; S Mylvaganam.Chem. Eng. J., 2003, 94(8),87-99. [0] V Mosorov; D Sanlows; Ł Mazurewcz; T Dyaows.Meas. Sc. Technol., 2002, 3(2), 80-84. [] HX Wang; C Wang; WL Yn.IEEE Trans. Instrum. Meas., 2004, 53(4), 093-096. 526