.... A. .... .1 . Z vulva-.5.“- . V . ......o’ .21..“ . . . . .. . . .... . ...... .... .- . . - n I 1“. u u . . .259... . . V. . . V . , .. V. . .o .. ~ . I . .. . . V . ‘33.; V ’ . . . . ... H.....N.L.‘u.hnw Ton“. .. o...... .swmoo_u”.taamvm .....~3 “what-fig . . . . . . o . V . . . . a... o ‘f . . . .. . ._ ...-8.3.. .. ... . V. .. . . n .. . .33.... . a 3. ..............~m.e. . 0.0.3... . V . o. V) ’0‘ . . . I n . . . . O. I . . . v .,V . . . .. . V . . .V o. . ... . I... o .. _ H . . . . . . . . .c a. at L... . . . ..Mu!....~. . . . V .V ...... . . t . us. . . . . ’Lb~0.v. ... V . . ... n3. ....3 . 3 .~ ..m ‘m. .... «...-.s‘i V V . . v . .V .... .. V . . '0‘. ... V.. ..- ...p . . . - ......" V V . .. . . .. 5 V. o... . . ...~o.. 5... 0‘ . . . . . .. .mooaudLOovo a . V F‘s-.0... . 003 2 1.... .V . . V. . ...: s. . c . 00“... ..va...uM..-‘u‘lnvo.uv . . . . V... . u‘. 6:..3 . . g on... . ..O. t. .0. I“ I. ...! .. V . . I. 0.1. .... 2.5.3.. ... .LI. . . . . .. V . ... yrio. v: fact-s.- . . . 3......” V . . . . V V . . V ...“.c :2 V “$9.59... «.435 . . . . . . air-...... 5. .. . . . . . . . . . . . V . too... ...; . . . ... V .. _ V V .0. . .0.‘- :0 . . . . V . V Cu...’ . 3;.- . . . . . . .l .‘ . .... .... . . V . . .... .. . .....V . V . 4.9!... ...) ...... .’..~: . . . J .8 ‘ 213:1.“ l . .5 3! \otfl... 34-. .... . O. ‘ «s..... .0»! a. .‘....4 ‘13. . .. .2... .1. t ... . V I. . o: . .90... ...... . 5‘68. . r c .. o. i...” 2.29.. . ...wfliu...3 . uV. I . . 30 g: .. .33: av U. . . . . . .3. .... .....r... .. t V . . . ...(HH. . . no: . 1-1.. 0 z ’ . . V ...? ‘ Cllun.vo-o. . . . ICA’. . . . o..‘..¢. . ‘u. 0' .. .. . ...! 0.... . v. .. . . . . V . V. ._ :3 . luff... 2... V . . . . O . . o ’D- 3.. . .A .0 a < . . . ..3. 33”.... .t ”.....ultor. . to .X 9. 1|- . V .. ... .J J.........t..a:b..£. 3 1.1.”? . .. .l“ V . . .. 5933.. . "29.2.... t. 1.. . .2373; .11.. ...-29.. .3.3-IOV. . .13. V .3043 V. . .....E 9.... . . . o. ... . ... OI“. . 0.... Vac o..fi£1. . . . .1 3: 33., Q: . .. . .. . . 2.0.1.2.. .. .o 0 V .v .I 1...... . .... .....a .90- . . o . u .3! O9:- . . . . . . .. V. ”to... {is ..oflhf nugvV- ...... ... ...-... V . . . . . .u . . .n .. . .... .s. V . . . ...: . . ... .2... 8-... . .t.. . ... “I. 9H0. _ . .V ’1 :0 ...-0.3.... . . ...-'0. .3 . . . . Hu‘uooqov‘ . .. to . ..4.. V V ., 3.2.0.8339. V . . : p5..o.V.-..IJ. . .... . V . a . u. . . . . V . ... 0.3.2.. a . .V V V V A. o. - . . . o0"...... . . . I...o ’3" .n‘. o _ “......o... .. o ‘v. on. . V. V nfi’h \h v. ......05 .50... on. . . ..v 004".)‘n 4.9.... ...-.1 . . 3., . o . . . . . . .'$,8.f . . .V .. . o... 0'... . .‘z .. V. V :l‘v \‘u . V .339; .. V . V . . ouV . . . l . V . . . . . . V on... . .an... .0.’V . . . . ......c- 5:...li o .o '0'. .. ..v "Jo—flux...” - . . . . . . . .l a. .o ......u....{ 5A... 9.. ..| . ~ ~ n I. O n «o. locQo- 3. . .o o v.17. .... 00.9.... .1; ..b . . .cV. ‘0. _ .t3 1 . o: . . . . V . 03“ .3- 3' . n1 0 220:1: .... a o. 2!. u... ... . V . . . ... 39* I... :73: . .slot..£.. . ... . . V . V V . c . to, .uflloo. . .-:';..o‘ ...r .95.. . . V . V . . . . “We... . . . . . . V . . V . S r... a: 1:: ..Vvo.-o .3 I . \. . . V . . . .. .V .. V . . . .. V . . . V...» 3’31}... .... ... 33...; .06... ‘.?o.u!.20 .- . . . . . . . V ... . . . . . ...o. o. ...-3 4.. ....oc. . .. .. .. . . . . .‘i. .7; .. .. a. . on!‘ us... . . . . V . . . .33110. . ciao: .. V... . . .. . . .o ....a ......r ...... . . . . .. . . . .... _. . . . in .n . ... .0. 3“ .- ...) 320..."- 33: ‘1). .. V . . . . . . 31.....- 9. J. . . . . . . . : _. .V. V . . .V V .. V n. s. . . y . ... . . . . V .V .32.)“: . .. 2.5.... V . bf .12... . . . V . _ .. .. V .o O. 9.7A.‘V1.V.. ..O‘ .. . . . . . . . . ....X. 3“ . . ‘.L. ...... . 3.. . .~.V..~ko...q . .... . .21. . .3 . .14.... . . .. V . .0 . I. -' .... .I‘. . :. ...?! . .. ..Jd‘z o.‘ as. ...-.4, ... .... . . ..... . Q... .4 . . . . 3r.1..ol~m.-V‘1.ncl. l§ .... up . . . 0. .V c 3.13”...- .. . r to 2. .... J ...». .n . 0 v V . . . V . . . 3 . V . .V obi-...c! . . ....O‘...Q-o:...u . . . . ... 3...}! V ...»... . . . .. . V .. . . . .."¢J‘o 0.... . .. ‘ .n o I V .. . . .. . . ..I.-.. .o ..gC..-../. of or. ‘ .... v . no. 5. . . .. .... .. .. . V . . it... . .hn‘..’ld‘ ..L . .v .0} ...V. 0.79....“ . V . . .... .. . . . .. . . ... . . V . .o‘ £28 ... ...}. ... . . .. . .a. 20.1.3 . .. Vo ... : . { . . . . . . . . :3ch ..o . 1.3.7. .. . . V .1‘0. no .. . . .. . . o. . . a . . . . nail... .... o I. ‘o ...-...?- Q 2... . h . . .V V. . ...... 3:93.. . . . .V... . V . . :0.‘.... .30. . .1 .... . V V . 2 V .93 .‘ . atoz ., . . .. . . ... .V .. V... 9.". . .. . 39.0.... .... . . . . . . . . . . . . . .. . . 6..., :0 .....n'o.:.oo.-.ou . . . . . . . . . . . ...-Va, .: \.7A . ... o . . . . . . 3V. 1 ..l . 3...: ' ova-V ? ...... «...? Caxtou‘) . V ...a 63-... to... n; V V .. . . . . . . . . . . .095... s....<.u. . . A we . . . V . . . ...].y: .03.. ... u: . .... ...:7... ,V a. .. .. T . ..w. .92. . I. . ... oi ...V...‘ 4 - . . . . 4.‘.o'. . ... . . V.. . . . . ...l 0:.- . . o. 11......0 3... Q. . . . .V V V ..oio‘ 4 a . . . . .C I Q . . . . n .- ... V ... . . . . . . . .V . if...» ...-iv. 0. 7 . .V . . . . . . u." ...-fa..- . ..3V. . V . .. .. .. Va... . . V .V . . . .. 327-. O“..3o...9.o .V... Div .....1.. . . . . . . V .. . V .c .. u .. . .. . . . . . .v.'a6.v 4. V .V . . .. ...-o s.” 1.0. . . , Q. l. '0... . . . . . . V. . . . .. ‘1’)? . . . . . V . V. .0...I.. . ... ..l V . . . . . V . A ...: a . .psuV-o....?~ .. . ... . .— .1. . V v . . . 04.. o. ll. . . . .V .. . . O .....V . . .. . ... . . V . .. . .. . V .. . V . . . V . . . . .... . A . V . .V . ... 2),. . . .. V. . .. ....\ .... .. 0.. V . . . . . . ... . .. . ....s‘. nvh‘uV:o. 4.-.’I . . . - . . . .h..-:. ...-0 . . . . . u.’ .v .‘ o... 1" ..v.'. ¢ . . . . . . . .. V- ...d .\~ . .0..- . . . V . . . . «a... .V . . ....I .5... . . . . . .V . . . . V '3. . . . . . . . . . . . . . . . V .. . . . . . y .. D: a .. . . . . V . . . . . V . .- . . V . . . V . . . . .V . . . . . . . . . . . .. . . .. . . . . . . . . . . . . v. 0» ‘._ . v . . . V V . $1.2.qu '0 ...-at. . . . . . . . . . . . . . .0. O. _ 3.5... D. . . . . .V V (y. 3.. . . a . a 9!. o. 4...... .0 .:.'J”.. .. a: . .n u . 34?; 303‘ V. .U-no v n. O . V V... ... o. . .... . . . 9 H. V 1 .. v.V.ooocuI VV~.o... .....l: . ... . . V. . .. . .V 9, .V .... .. c... . r.v .00. .V I... .0 V. ..... ... . .A.'I.t .-.)I4 .0 .- . ...:...........V 2 vl' .O.’a .3 327.1. . .. t . 0 -- . .. ‘ - u. ... .‘ 0....- b‘ Vn4 “Conch coin... ... V .v. .. oo. . .3... r: ... .....V. . .4. ......faS. J... V. ... .... .12. V. ... . .... .... ...: u.§...§§n%m$§.:. V... 3 ....m.... 3%... _ .... V .V; .................._. ......u. ..3... :fifi.§.. 33..-... ‘ . .s.‘. t .‘ v. . . .l 316 ....I. ..I “a. ..3..‘..V.‘.o.r..:..~. — foil: .‘L. .0 ... u c o .0350 ’=- ... .. . .4 . V. —. O. 4. §. .0 ..n...-n§ .0. o. n... vi.‘ ....R....3~b~.. ._ 51......9 ... y f . .. o .. .v_§.:.'wm.’n.-H”.U.‘nco.urq a L l c. O. . c‘.:' a... I .Vuo: - ... . ...: ...... I ‘ ... 03:39.9, scion-.4!- . .- .. . .o \ ...: . . ... y. ‘. ‘3... H": "-94.9 v! ...v....' . O ..... . _. . ...wanfififiv‘nquuuu.m~mnmnMrnvdmmrgfirignmfit?. . ... ......u: . J '1 1! ...:- l 2004 UBRARY Michigt. " Tate University This is to certify that the thesis entitled MOLECULAR TAGGING MANOMETRY: A FEASIBILITY STUDY presented by Mr. Rajat Basu has been accepted towards fulfillment of the requirements for the Master of degree in Mechanical Engineering Science H M (twig/l rm; ( Major Professorw 4/2 2/2 a 027 Date MSU is an Affirmative Action/Equal Opportunity Employer ....-.----—-—.—.--.—q—--I-I-I-c-l-o-I-o-c-I-u-u-o—u-u-u-o-n-n-o-o- 0—-—I—c-l-u-D-I-0-l-I-U-I-I-D-I-l-.-I-I-I-‘_'—- PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5108 K:lProi/Acc&Pres/ClRC/Dale0ue indd MOLECULAR TAGGING MANOMETRY FOR GAS PHASE FLOWS: A FEASIBILITY STUDY By Rajat Basu A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Mechanical Engineering 2008 ABSTRACT MOLECULAR TAGGING MANOMETRY FOR GAS PHASE FLOWS: A FEASIBILITY STUDY By Rajat Basu The primary objective of the research conducted was to determine the feasibility of a proposed planar whole-field, non-intrusive, molecular based Optical diagnostic technique to measure pressure in the body of gas flows; a capability that has so far remained elusive. The technique relied on the quenching of a molecular tracer’s phosphorescence lifetime. The tracer used was acetone and the specific quenching mechanism relied on was oxygen quenching and acetone’s self-quenching. A combined photochemistry and detection parameter based mathematical model was developed and was used to obtain Optimal experimental conditions in order to detect a change in acetone’s phosphorescence lifetime with change in pressure. An experimental set-up was built and measurements were made to obtain the required ‘proof—of-concept’. The results obtained showed that the proposed technique is feasible and worth pursuing. The agreement of the experimental data with the model’s predictions was within 10-12%. ACKNOWLEDGEMENTS To date, completing my master’s thesis has been the single largest intellectual accomplishment of my life; something that I could have never achieved without the involvement of certain people. I would like to take this opportunity to thank them. First and foremost I would like to thank my principal advisor, Dr. Manoochehr Koochesfahani for the faith he had in me in agreeing to let me work at TMUAL and be my advisor. My research interest when I first came to Michigan State University was IC engines. However, afier speaking to Dr. Koochesfahani I decided to make a switch to Laser Diagnostics. The research I was engaged in proved to be one of the best learning experiences of my life and every bit as exhilarating and challenging as I hoped research would be and more. Being part of a group which maintains the highest standards of research made me not just a better researcher but also a better person. When I first started, the going was rough and my progress slow. Dr. Koochesfahani was patient and encouraging, always helping me along to gain a clearer vision of the problems that I worked on and appreciate its subtleties. I will be eternally grateful for the time that he spent teaching me all the tools I needed to get going and start working independently. From a personal standpoint, Dr. Koochesfahani is somebody who I have the greatest respect for. His unwavering adherence to honesty, integrity and his personal drive to maintain a high standard in everything he does inspite of what the world around him does or says has inspired me to do the same. I hope this acquaintance lasts beyond my years at TMUAL. iii I would like to thank my co-advisor, Dr. Ahmed Naguib for his continual involvement in my work and his crucial inputs whenever I hit a roadblock. He never once refused to give me time inspite of his hectic schedule, particularly at the end of my master’s when schedules were tight: something for which I can’t thank him enough. I would also like to thank Dr. Tonghun Lee for agreeing to be on committee on short notice. A lot of his inputs helped me solve crucial design problems even before he was formally on my committee. The advice that he gave me during our long conversations related to topics research and my career in general kept me motivated and focused. No lab environment is complete without great lab mates. The help and support I got from Shahram (who in my opinion is the most calm, composed and patient person I have ever known) right from the first day till the last was incredible. Nothing in words would ever be able to adequately express how grateful I am. I would like to thank Vibhav, Alan and Sagar for all their help and keeping me company during experiments which ran late into the night. I will always cherish the times we spent together. I would like to thank my fi‘iends Pranav, Arjun, Siddharth, J ayesh, Zahid and Sagata who encouraged me when the going was tough and shared in my joys I would like to thank my family, my mother Nita, my father Shyamal and my brother Rahul for all their love, support and understanding while I completed my master’s. Inspite of being far away from home, I always felt their presence and that gave me the strength to keep going when the going was the hardest. iv Finally I would like to thank my better half and my greatest support system, Nupur. Being halfway across the world and yet understanding perfectly every trial and tribulation I went through. Thank you for your love and patience. I would like to dedicate this thesis to my mother; who never understood what I did during my master’s but understood its importance and what it meant to me. For believing in me when no one else did. Thanks Morn! TABLE OF CONTENTS LIST OF TABLES ................................................................................ viii LIST OF FIGURES .............................................................................. ix LIST OF SYMBOLS ............................................................................ xiii CHAPTER 1 : BACKGROUND AND MOTIVATION .................................. l 1.1 . Concepts related to lifetime ........................................... 5 1.2 . Applications of molecular tagging .................................... 9 1.2.1 . Molecular tagging velocimetry ................................ 9 1.2.2 . Molecular tagging thermometry .............................. 11 1.3 . Molecular tagging manometry: motivation. . . . . . 17 1.3.1 . Related pressure measurement techniques .................. 17 1.3.2 . Principle of Molecular tagging manometry and potential advantages ........................................................... 21 1.4 . Molecular tagging therrnometry for gas flows ..................... 22 CHAPTER 2 : ACETONE AND ITS PHOTOCHEMICAL PROPERTIES........ 23 2.1 . Oxygen quenching ...................................................... 23 2.2 . Self-Quenching and its wavelength dependence ................... 25 2.3 . Emission and absorption spectra ..................................... 28 2.4 . Photolysis ............................................................... 29 CHAPTER 3 : MOLECULAR TAGGING MAN OMETRY ........................... 32 3.1 . Analysis and aspects to consider ..................................... 37 3.1.1 . Acetone and oxygen concentration related calculations. 37 3.1.2 . Signal response for different compositions ................. 40 3.1.3 . Gating schemes ................................................ 47 3.1.4 . Detector related issues: ghost image and integrating fluorescence light ................................................ 51 3.2 . Experiment .............................................................. 56 3.2.1 . Setup ............................................................ 56 3.2.2 . Procedure ....................................................... 64 3.3 . Results and post experimental analysis .............................. 68 3.3.1 . Experimental settings .......................................... 68 3.3.2 . Results ........................................................... 72 3.3.3 . Possible sources of error ...................................... 74 CHAPTER 4 : MOLECULAR TAGGING THERMOMETRY ....................... 82 4.1 . Temperature dependence of acetone triplet lifetime on temperature ........................................................... 84 4.1.1 . Set-up ............................................................ 84 vi 4.1.2 . Details of the experiment .................................... 86 4.1.3 . Experimental procedure .................................... 88 4.2 . Results .................................................................... 90 4.3 . Analysis .................................................................. 93 CHAPTER 5 : FUTURE DEVELOPMENTS ........................................... 97 5.1 . Provision for making mixtures with different compositions. . 97 5.2 . Adjustable sensitivity of signal response by varying delay ....... 98 5.3 . Dual tracer based system for combined thermometry and manometry experiments .............................................. 101 APPENDIX A. ANTOINE FITTED EQUATION IN TABULATED FORM IMAGES OF STATIC TEST CELL ........................................ 106 APPENDIX B. ATTENUATION AND VARIATION OF INITIAL PHOSPHORESCENCE INTENSITY (10) WITH PRESSURE FOR MTM EXPERIMENTAL RESULTS ........................................ 114 APPENDIX C. PRE-EXPERIMENTAL CHECKS: SEEDING CHAMBER TEMPERATURE ............................................................ 122 APPENDIX D. CHANGES IN PHOSPHORESCENCE LIFETIME MEASUREMENTS DUE TO VARIATIONS 1N LASER INTENSITY .................................................................. 132 APPENDIX E. FLEXIBLE MIXTURE COMPOSITION SYSTEM FOR MTM EXPERIMENTS .............................................................. ‘40 LIST OF REFERENCES ...................................................................... 1““ vii Table A.1 Table B.1 Table B.2 Table C] Table C.2 Table D.l Table D.2 LIST OF TABLES Antoine equation in tabulated form ....................................... Drop in laser power in static cell due to attenuation ................... Expected and calculated variation in initial intensity with attenuation taken into account ............................................. Example of variations in lifetime corresponding to changes in seeding chamber temperatures. Concentration of acetone based on its vapor pressure at 22°C at room pressure ............................. Example of temperature variations inside the seeding chamber corresponding to changes in temperature of the bath .................. Sources of error in an average image due to variations in the laser and their respective magnitudes .......................................... Example of variation in lifetime with change in oxygen concentration for MTT ..................................................... viii 107 117 119 129 130 133 138 Figure 1.1 Figure 1.2 Figure 1.3 Figure 1.4 Figure 1.5 Figure 1.6 Figure 1.7 Figure 1.8 Figure 1.9 Figure 1.10 Figure 2.1 Figure 2.2 Figure 2.3 LIST OF FIGURES Exponential decay model ............................................... Typical MTV image pairs and the resultant velocity field [Gendrich, C. P., Koochesfahani, M. M. and Nocera, D. G. [1997]]). (a) The grid imaged 1 ms after the laser pulse. (b)The same grid imaged 8 ms later. (c) The velocity field derived from (a) and (b) ................................................................. Use of MTV in IC engine flow studies [[Koochesfahani, M. M., Goh, A. C., and Schock, H. J. [2004]] .............................. Study of supersonic jet flows using MTV [W. R. Lempert, N. Jiang, S. Sethurarn, M. Samirny, [2002]] ........................... Phosphorescence exponential decay curves: intensity decay with time for different temperature [[Hu, H., Koochesfahani, M. M., and Lum, C. [2006]] ................................................... Change in a tracers temperature response curves with time. [Hu, H. and Koochesfahani, M. M. [2006]] ................................ Delayed phosphorescence images for a typical MTT experiment [H. Hu, M. M. Koochesfahani, and C. Lum. [2006]] ............... Image pair used for MTV&T measurements.[Hu, H. and Koochesfahani, M. M. [2006]] .......................................... (a) Mean velocity profile and (b) mean temperature distribution obtained from the image pair shown above (Fig. 1.8). [Hu, H. and Koochesfahani, M. M. [2006]] ..................................... Surface pressure measured on a high speed civil aircrafi. [Bel], J., Schairer, E., Mehta, R. [2001]].................. Quenching of acetone triplet lifetime due to oxygen ................ Quenching of acetone’s triplet lifetime due to self-quenching. . Variation of quantum efficiency of acetone triplet emission with acetone pressure and excitation wavelength. [O’Neal, E. and Larson, C. [1968]] ........................................................ ix 10 11 11 12 14 15 l6 17 20 24 26 27 Figure 2.4 Figure 2.5 Figure 2.6 Figure 3.1 Figure 3.2 Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 3.7 Figure 3.8 Figure 3.9 Figure 3.10 Comparison of acetone and oxygen quenching of acetone triplet lifetime .................................................................... Acetone absorption spectra .............................................. Photolysis of acetone to form biacetyl ................................. Variation of acetone triplet lifetime with acetone and oxygen concentration (3D) ....................................................... Variation of acetone triplet lifetime with acetone and oxygen concentration (2D) ....................................................... Variation of vapor pressure of acetone with temperature [A. Lozano, B. Yip, R. K. Hanson [1992]] ................................ Variation of phosphorescence signal ratio with change in pressure: [02]= 9.375 x10—3 moI/I (concentration of oxygen in air at 1 bar) ......................................................................... Variation of phosphorescence signal ratio with change in pressure: [02] = 9.375 x10.4 mol / I (1/10th of the concentration of oxygen in air at 1 bar) ................................................. Variation of phosphorescence signal ratio with change in pressure: [02]= 9.375 x10—6 mol / I (1/1000th of the concentration of oxygen in air at 1 bar) .................................................... Variation of phosphorescence signal ratio with change in pressure: [02] = 2.067 x10—7 mol / 1 (concentration of trace oxygen in the nitrogen used at 1 bar) .................................................... Signal ratio versus lifetime: comparing the use of equal and unequal gate periods ...................................................... Comparison of relative intensities of fluorescence and phosphorescence .......................................................... MTM flow circuit ......................................................... 28 29 30 35 36 38 43 45 46 49 53 58 Figure 3.11-A Figure 3.11-B Figure 3.11-C Figure 3.12 Figure 3.13 Figure 3.14 Figure 3.15-A Figure 3.15-B Figure 3.15-C Figure 3.16 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 5.1 Figure 5.2 Figure 5.3-A Schematic of connections leading upto the static test cell used for MTM ................................................................... Dimensions of the static test cell used for MTM ..................... Photograph of the static test cell used for MTM ..................... An exploded isometric view of the test cell window ................ 15x15 pixel region fi'om where counts are obtained in the static cell experiment of MTM ................................................. MTM experimental results .............................................. MTM experimental results: sources of error- variation of oxygen concentration .............................................................. MTM experimental results: sources of error- variation of acetone concentration .............................................................. MTM experimental results: sources of error- variation of quenching constants K q and K a ....................................... Photolysis of acetone to form biacetyl .................................. MTT flow circuit (heating added) ...................................... Details of the heating apparatus added to flow circuit. . . . . . . . . .. Variation of acetone triplet lifetime with temperature for different seeding chamber temperatures .............................. Variation of acetone triplet quantum efficiency with temperature for different acetone concentrations. [O’Neal, E. and Larson, C. [1968]] .................................................................... Error propagation in MTT ............................................... Variation of signal response with pressure: delay being varied. . .. Variation of signal response with pressure: delay being varied: Log .plot ................................................................. MTM: variation of lifetime with pressure and temperature: Tracer 1 ................................................................... xi 59 60 61 63 71 72 76 77 78 80 85 86 91 92 93 99 100 103 Figure 5.3-B Figure 5.4 Figure B.1 Figure B.2 Figure B.3 Figure C.1 Figure C.2 Figure C.3 Figure C.4 Figure D.l Figure D.2 Figure D.3 Figure E.1 Figure E.2 Figure E.3 MTM: variation of lifetime with pressure and temperature: Tracer 2 ................................................................... 104 Superimposing of isolines from lifetime contour plots of two tracers for simultaneous temperature and pressure measurements 105 Absolute location of imaging region location with respect to the test cell entry window ................................................... 1 l7 Extinction co-efficient of acetone as a function of excitation wavelength ............................................................... 118 Variation of initial intensity with pressure: MTM experimental results ..................................................................... 120 MTT flow circuit .................................. 124 Variation in intensity of image before and after the constant temperature bath was put in ............................................. 127 Variation in temperature with time for nozzle and seeding chamber bath before and after the constant temperature bath was put in ........................................................................ 128 Temperature change inside seeding chamber over time compared to changes in bath temperature .......................................... 131 Change in lifetime with change in local laser intensity. [Kaskan, et al., 1949] ................................................................ 134 Signal ratio versus delay: exponential decay curves ................. 136 Variations in lifetime with variations in local laser intensity ....... 137 Auxiliary/ reference tank ................................................ 141 MTM schematic of flexible mixture composition system 142 Reservoir or blowdown tank ........................................... 143 Images in this dissertation are presented in colour. xii LIST OF SYMBOLS dt MTV MTT MTM PIV PSP Initial intensity of emission after a laser pulse. Intensity of emission at time‘t’. Signal obtained by integrating the emission till time‘t’. Exposure time or gate period. Time for which an emission is recorded Single exponential lifetime. Refers to fluorescence or phosphorescence depending on subscript used. Signal ratio. Typically ratio of light integrated in the second image to that of the first. Delay as defined in this thesis; the time between the end of the first image and the beginning of the second. Generic definition of delay; time between the beginning of the first image and the beginning of the second image. Molecular tagging velocimetry. Molecular tagging therrnometry Molecular tagging manometry. Particle image velocimetry. Pressure sensitive paints. The highest lifetime of a tracer when completely unquenched. Triplet oxygen quenching constant for acetone. (8000 l/mol.us) Triplet self quenching constant for acetone. (12 l/mol.us) Concentration of acetone in mol/l. Concentration of oxygen in moi/1. Parts per million (by volume in this thesis). Refers to concentration of oxygen in the nitrogen cylinder used. xiii NTP STP Normal pressure and temperature. (20°C and latm) Standard pressure and temperature. (273.16 K and latrn) Factor of compression. The numerical factor by which a mixture is compressed or expanded in a MTM experiment. C f =1 for room pressure. Peak pressure to which the test cell is filled for a particular experiment. Factor of compression. The numerical factor by which a mixture is compressed or expanded in a MTM experiment. Fc_e =1 for room pressure. Subscript used for fluorescence. Subscript used for phosphorescence. Quantum efficiency. Absorption cross-section. Figure of merit for acetone fluorescence. Figure of merit for acetone phosphorescence. Flow control valve. Solenoid valve. National pipe thread. Universal gas constant (8.314 J/mol.K) Extinction co-efficient. Number of photons. xiv CHAPTER 1: BACKGROUND AND MOTIVATION The primary objective of this thesis was to determine the feasibility in practically implementing a proposed non-intrusive, molecular based optical diagnostic technique to measure pressure in gas flows. The technique proposed is a molecular based technique; meaning that the ‘barometer’ or ‘pressure transducer’ in this case is a molecule. Given the practical limitations in reducing the physical size of a measuring device, the availability of a non-intrusive technique to measure various flow parameters is always beneficial. The practical implementation of this technique if proven to be feasible would find its relevance in many compressible flow studies like IC engines, high performance turbines, non-linear acoustic research, shock wave studies, among others. The generic term for this kind of molecular based optical diagnostic technique is called ‘molecular tagging’. The growing interest in research topics related to micro-fluids, small scale turbulence, molecular diffusion, boundary layers, molecular mixing, reacting flows, etc., molecular tagging has greater scope today than ever before. The technique of molecular tagging relies primarily on exciting tracers in a fluid flow by a light source of appropriate wavelength and intensity. The light source could be a laser, a flashlarnp or a monochromatic light source depending on the nature of the study. The tracers in turn emit light, the duration, intensity and wavelength of which depends on the intrinsic photochemical properties of the tracer, its dependence on the nature of the illuminating light and the presence of other species. The light emitted can be broadly classified into a short-lived emission caused by singlet-singlet transitions or a longer emission caused by a triplet-singlet transition. They are typically referred to as fluorescence and phosphorescence, respectively. Fluorescence occurs when an electron in the atom goes from an excited singlet state to a singlet ground state (a ground state of the same multiplicity) and emits a photon. This transition is quantum mechanically allowed and has a high probability of occurring. Triplet-singlet transitions however are quantum mechanically forbidden and have a low probability of occurring. As a result phosphorescence emission is more efficient than fluorescence emission and has a longer lifetime of emission. For the tracers used, the order of magnitude difference in the emission ‘lifetimes’ between fluorescence and phosphorescence is typically 100-1000 times. Typical tracers produce both kinds of emission simultaneously. However in many cases, the flow conditions and/or the presence of other species makes only fluorescence emission detectable. Elevated temperatures and certain species called ‘quenchers’ ‘turn off” or quench phosphorescence emission. However when detectable, phosphorescence emission is extremely useful. This is primarily due to its longer lifetime coupled with its dependence on the presence of other species and local flow conditions. If this dependence is calibrated for appropriately, by measuring the phosphorescence emission of a tracer in a flow one can obtain a lot of additional information about the flow field in an indirect manner. A good example of the use of the longer emission period of phosphorescent tracers is for velocity measurements [Koochesfahani, [1999]]. This technique is referred to as MTV or ‘molecular tagging velocimetry’. In this case the lifetime of emission was long enough so one could detect the physical displacement of an emitting molecule in a flow field. This displacement when quantitatively acquired for an entire flow field would provide the corresponding velocity field. These velocity fields after post-processing could also provide the kinetic energy and the vorticity of the flow field. The dependence of phosphorescence on flow conditions like the local flow temperature has been used to study thermal-mixing [Hu, et al. [2006]]. However when used in this manner, a calibration needs to be performed in order to ascertain the quantitative phosphorescence emission dependence on temperature of the specific tracer being used. Certain classes of flow problems require the use of the fluorescence emission as well. The use of fluorescence is seen to a large extent in concentration measurements [Lozano, et al [1992]] and studies involving mixing and reacting flows [Koochesfahani, et al. [1986]]. Fluorescence tracers are also useful when studying flow systems with very high temperatures since very high temperatures are known to quench the phosphorescence emission of a tracer completely [Groh, et al. [1953]]. Fluorescence however is detectable even if the tracer is subjected to several hundred degrees. Compared to other non-intrusive flow measurement techniques, the use of molecular tagging has some obvious advantages. For example when using traditional particle tracking techniques, some of the primary concerns are seed particles lagging the flow, oil deposits on wind tunnel viewing windows (in case PIV where seeding is done using oil) or damage to the flow apparatus due to clogging or eroding as seen in case of PIV measurements where the flow is seeded with sawdust or buoyant micro spheres. In molecular tagging the tracers in can be added externally, pre-mixed with a carrier fluid or produced in the flow (typically observed in reacting and combusting flows). Despite its success as a highly effective and reliable non-intrusive flow diagnostic technique, there are certain situations where the use of these molecular tagging techniques is not appropriate. As mentioned before, the emission characteristics of the tracers get affected by several flow parameters. The presence of quenchers, the temperature of the flow, concentration of the tracers itself, local laser intensity, how long a tracer has been in a solution etc., all affect the emission characteristics of a tracer. If the measured flow parameters quantitatively depend on the measured emission, it is imperative that the emission’s dependence on the flow conditions and composition is quantitatively understood and taken into account. Phosphorescence is particularly sensitive to these issues compared to fluorescence, given the difference in the emission time scales of the two processes. These aspects will be covered in more detail in subsequent chapters with emphasis on how these issues affected the experiments conducted during this project. Solutions to eliminate or at least mitigate these problems have also been suggested and implemented. Before getting into any details, it is advisable to go over a few basic concepts related to the terminology typically associated with the kind of research discussed in this thesis. 1.1 CONCEPTS RELATED TO LIFETIME The kind of emission we will be dealing with during this thesis is almost exclusively phosphorescence. The basic premise of the techniques discussed is the variation of phosphorescence lifetime at different points in a flow stemming from variations in the local flow condition and composition. Some allusion to fluorescence could be made primarily as a means of testing experimental set-ups and corroborating some results indirectly. The phosphorescence emission observed in most tracers used for studies of this nature is almost always exponentially decaying. That is, the decay of emission intensity with time is exponential. Figure1.l is a typical decay curve showing the variation of intensity with time. InFigure 1.1, t1,t2 = The time when the first and second exposures start, respectively. At = The time for which the emission is recorded. Referred to as ‘exposure time’ or ‘gate period’. td = Time between end of the first gate and beginning of the second gate 4) = Intensity at time '0' N. o At Intensity (I) —-- At t1 . td Delay (t) —-- Figure 1. 1: Exponential decay model In general we have for an exponentially decaying model: It : [0 e—t / T Where, I t = Intensity of emission at a time 't' after a laser pulse. 1 = Lifetime of the tracer. This number refers to the time when the intensity drops to 36.7% of its initial value (b). Now imagine a flow containing a tracer with lifetimer , being illuminated by a single laser pulse due to which its initial intensity is b after which 4, starts to decay. If we now start imaging this emission at time t and image it for an exposure time of At , then we can say that: t+At -t/ St = I [0 e rdt t (Note that the flow is being excited by only a laser pulse and not a continuous beam.) -. S, =—Iorle’(’+A’)/’— e’mJ (Eqn.1.1) St = Signal integrated over ‘ At’ what we refer to as gate period or commonly known as ‘exposure time’. Note: At] ¢At2 i At (If the gate periods are unequal.) Referring to Figure 1.1 if we integrate the emission of the tracer twice (for the same time At) following a single laser pulse, once at t1 and once at t2 , then we get the signal integrated over the two exposure times as S ,1 and SQ . Sn being the signal integrated over the first exposure time and 5,2 is the signal integrated over the second exposure time we have: St :_10Tle—(r1+At)/r_e—tl/rJ 5:2 = 40$ [e—(rz +At)/r _ e"2 n] 5:2 —=S R which we will denote as ‘Signal ratio’. t1 Based on the above we can write: SR =e—(t2 -—t1)/r = e—(td +At)/ r (Eqn. 1.2) At this time it is important to clearly explain how the term ‘delay’ has been defined in the work presented. In many references, the term delay is denoted as 'dt' which is the time from the beginning of one exposure time image to the beginning of the second. However in this thesis what is referred to as ‘delay’ is denoted as 'td' which is the end of the first image and the beginning of the second. Specifically, dt =td +At1 This is so is because the camera used allows this parameter to be changed independently. This also makes the analysis and processing of data a lot simpler. It should be noted that the phosphorescence emission need not be exponential always. The emission can also comprise of multiple exponentials, polynomials or a polynomial in combination with an exponential. For the systems discussed in this thesis, the first order behaviour is typically dominated by a single exponential function and other parameters have a very small effect if at all. Therefore using a single exponential definition of lifetime substantially simplifies the analysis and introduces no discernible error. For all the analysis shown, ‘lifetime’ refers to the single parameter exponential phosphorescence lifetime of acetone unless mentioned otherwise. Throughout this report the above nomenclature will be followed and any new terms introduced will be defined. 1.2 APPLICATIONS OF MOLECULAR TAGGING As mentioned before, molecular tagging can be used to measure a number of different quantities in both gas and liquid flows. We now look at a few specific molecular tagging applications in detail. This will help the reader understand better the motivations for the work conducted here. 1.2.1 MOLECULAR TAGGING VELOCIMETRY The use of molecular tagging to measure velocity fields is arguably its most popular application. One of the approaches is to rely on the long lifetime of the phosphorescent tracer. In a typical experiment, in order to make velocity measurements, one would excite the tracer/S in a flow in the region that needs to be investigated. The tracers are excited or ‘tagged’ by a pulsed laser which causes them to emit light which can be captured by a photo-detector/camera. Depending on the intensity of the emitted light, either an intensified or an ordinary CCD camera may be used. Following the laser pulse, two images of the region are taken in quick succession. Comparison of the two images provides us with information regarding the pixel-wise displacement of the excited molecules in the time elapsed between the two frames. This in turn, coupled with the imaging scale, will give us the velocity of the flow at those points. For whole field measurements, a pulsed laser grid is used as shown in Figure 1.2. By comparing the pixel-wise displacements at the nodes of the grid, one can obtain the entire flow field. Once the complete velocity field is obtained (after post processing the velocity fields obtained in an appropriate manner) it can be further used to obtain other flow quantities such as the vorticity distribution, velocity gradient, strain-rate and kinetic energy of the flow as well. MTV has been used in both liquid flows [Koochesfahani, et a1. [2002]] as well as gas flow studies [Stier, et al. [1997]]. (a) (b) (c) Figure1.2: Typical MTV image pairs and the resultant velocity field [Gendrich, et al. [1997]]). (a) The grid imaged 1 ms after the laser pulse. (b)The same grid imaged 8 ms later. (c) The velocity field derived from (a) and (b). A variety of different flow fields that could not be acquired using traditional velocity measuring techniques have been successfully acquired using MTV, proving the usefulness of this technique in a multitude of engineering flow problems. Some unique applications include studies of flow inside an IC engine during the intake stroke 10 [Figure1.3 Stier, et al. [1999]], study of supersonic microjets [Figure1.4 [Lempert, et a1. [2002]] and alloy solidification studies [Lum, et al. [2001]] (a) Figure1.3: Use of MTV in IC engine flow studies [Stier, et al. [1999]]. Image (a) is the reference image and image (b) is a delayed image, SOUS after the first image. 012 3 4 5 6 7 s 9(mm) Figure1.4: Study of supersonic jet flows using MTV [Lempert, et al. [2002]] ll 1.2.2 MOLECULAR TAGGING THERMOMETRY Molecular tagging thermometry or MTT is a relatively new application of molecular tagging. This technique relies primarily on the dependence of the phosphorescence lifetime of a tracer on the local flow conditions; specifically temperature. By measuring the phosphorescence lifetime at a point in a flow, and knowing the lifetime’s dependence on temperature through appropriate calibration done before, one can obtain the local temperature in the flow (see Hu, et al. [2006) and Hu, H. and Koochesfahani, M. M. [2006]). 1.2 L — - — exponential fit : A T = 50 °C 3 V T = 40 °c . ’= l \ T = 25 C i 2 > \ \ \T‘ -- O . 92 0 8 .---..\..:_.....- 2. \ . C ’ i - i ' (9 l .8 + \ ‘ \‘Ej‘ A \ {L ‘ g 0.6 ‘ \ \ E i \ \ E3 \ .- q 5 o 4 ' \ 7‘ ~rL ' Z ' I X \ V \ ‘ £1 0 2 i \ A \ 7 l ' ’ V\ ~ . \AL \ A j 0 4 L A - - A - - - - A - ~— 0.5 1.0 1.5 2.0 2.5 3.0 Time delay to after laser pulse (ms) Figure1.5: Phosphorescence exponential decay curves: intensity decay with time for different temperature [[Hu, et al. [2006]] 12 Figure1.5 shows the typical change in the phosphorescence decay curves with temperature. The decay exponent or 2’ (what we refer to as lifetime) changes with temperature. Using a laser sheet and applying this procedure at every point on the sheet, one can derive a complete temperature field. This technique has tremendous potential and has been used successfully in the study of different temperature fields. MTT requires one to calibrate a system in order to know not just qualitatively but also quantitatively the phosphorescence lifetime dependence on temperature. Also the number of factors affecting the phosphorescence lifetime may be more than one in some cases. Hence the effect of each of these needs to isolated and studied on its own. Their effect on the experiment and each other needs to be understood; particularly when more than one of these factors is at play together and the experimental set-up does not allow one to eliminate all but one effect. This will be shown in subsequent chapters. In some cases the tracer itself undergoes changes. For example in the case shown in Figure1.6 (a), the dependence of the tracers’ lifetime on temperature changes with time from when the tracer solution was prepared to the time it was used; an ageing effect of sorts. Figure1.6 (b) shows how this effect had been compensated for. The universal curve shown in Figure1.6 (b) is obtained by normalizing the lifetime at any temperature to its lifetime at 25°C. This in essence says that by knowing the temperature at any one point in an unknown flow, one can obtain the specific lifetime versus temperature curve for the tracer used (whatever its age may be) from this normalized universal curve and from there the rest of the flow field can be now obtained quantitatively. l3 2882 .2 .2 8338:6682 as. .m BE 38% mama 65 2836 a 6:8 688862 3 938388 :33 258.:— mo 32?» 838% we counta> A3 BE: ES, 823 Dmaommfi 059.8388 208.5 a E omega 6.3.3»:— SV 3 GOV SEEOQEE. Gov EBSOQES. om mv ow mm om mm om om mv ow mm om mm om "l A A n. O N Ml 4" A M . o m U ...... - N o W. -- P _ . . mu. . _ m . 9 m o ...... u. u n w :1 m w _ . II. a m o w i E _w_Eo:>_oa . II a w . 55963” o o.— M i .223me 9.10 i- r. it v 16. 86. ECoE — u 1... BE EcoE F a . wagchoE m a N._. no .- CofiszoEm c .i- m .23 92:9: 0 x m. 5:“... 22:06 m I W A F fi.—- H h n o 14 In a typical MTT experiment, akin to a MTV experiment, following a laser pulse which tags the tracer molecules, an image pair is taken. In case of an MTV experiment the second of the image pair shows a ‘distorted’ grid compared to a uniform grid in the first image. Based on this one can determine the change in pixel locations of the excited molecules. In the case of MTT, the second image shows lighter and darker regions indicating colder and hotter areas respectively; the darker areas indicating a lower lifetime or higher local temperatures and the lighter areas indicating a higher lifetime and therefore a lower local temperature. Figurel.7: Delayed phosphorescence images for a typical MTT experiment (a) t = 1 ms after laser pulse. (b) t = 3 ms after laser pulse. (c) t= 5 ms afier laser pulse. ((1) t =7 ms after laser pulse. [Hu, et al. [2006]]. The tracer used is a phosphorescent triplex called 1- Ber_Mb-CD_ROH. 15 Figure1.7 shows typical image pairs one might expect to see. The experiment tries to capture the temperature distribution within the wake of a heated cylinder as a colder flow moves over it. The darker the region, warmer is the temperature of the flow in those regions. One of the most ingenious applications of MTT is its application in combination with MTV [Hu, H. and Koochesfahani, M. M. [2006]]. Figure1.8 shows the image pair obtained in that case. Note the qualitative variations in intensity indicating temperature gradients and the presence of grid lines seen in most planar MTV experiments. Figure1.8: Image pair used for MTV&T measurements. (a) First image acquired 1 ms after laser pulse. (b) Second image acquired 5 ms after laser pulse [Hu, H. and Koochesfahani, M. M. [2006]]. The examples shown in Figures 1.8 and 1.9 provide an excellent example showing the possibility of different molecular tagging techniques being used together to obtain information on various flow field parameters simultaneously. l6 .2683 .2 .2 .5EBEBM Ba .2 BE 2.62:6 65 ~96 mE>oE Bo: 200 088003: 65 no 8380088 Hue—EH 98 52:20 05 00 2320088 Nor—r .0500 m “N 8380088 ”,5 600.0 00 $62? .3208 N 5:3 No.0 3 8.83 NEE 58:8 2: ”Co—EH I PC 52:; I .3 mm :oswmzmsco: 23300800. Am; Deswiv 30% 850% :3 Owen: of Set 3:850 20:2:me 8:880an :38 fit use 2020 5623 :32 A3 "a; 2:3“— e> - - - v N N _ o _ N N a N N _ a; o _- N- m- -+——_4—~q4fidqq-q-fi—qd-udqdd—q—q—q_dqr 4. a o . - . a I o fiqu%-di—1uJ-___-_——_——__-_fidi—q—-_—r 0N0.0 000.0 unfit“... 03.0 ... 0m0.0 000.0 0h0.0 000.0 000.0 23860800. - M-..’.‘” -V. -~—-~ -\«I— M". O-“ r u- .. -A-'_.-...-A-- . a - _ AM 3:23 upopp—-———----- . 't'»: -.. ...... 1llllJlLJJlllllllllllllllllLllllll N LJlllllLlllllJl+llllJlllllllllllll N 17 1.3 MOLECULAR TAGGING MANOMETRY: MOTIVATION To date there exists no reliable molecular based technique for acquiring the whole-field pressure distribution within the body of a flow and the use of molecular tagging thermometry OVITT) for planar whole field measurement of temperature fields has been used almost exclusively in liquid phase flow studies. This thesis aims at studying the feasibility of using acetone as a phosphorescent tracer in gas flows to obtain whole field pressure and temperature data independent of each other. Before getting into the details of how this could be done, it is advisable to discuss briefly a few commonly used pressure measurement techniques and how MTM holds up when compared to them. 1.3.1 RELATED PRESSURE MEASUREMENT TECHNIQUES Traditional pressure measurement techniques include using pressure taps when measuring surface pressure. However there is a physical limitation where the use of pressure taps is concerned. The Size of a physical pressure tap cannot be made smaller than a certain limit. Even if one could indeed reduce the size of a pressure tap to the required degree (a pressure tap that is too small has issues of its own), there is a practical limitation on the number of physical taps one can install. This leads to a trade-off between resolution and size of the area being investigated. Further having too many taps close together may actually modify the flow being studied introducing its own error. This would be even more pronounced when the surface features on which the pressure variations are being measured are small. 18 Using pressure taps also imposes the limitation of being able to measure only surface pressures. Whether the flow is compressible or incompressible, pressure information in the flow is always useful. In certain compressible flow scenarios like shockwaves, strong pressure gradients exist within the body of the flow and the ability to measure such pressure gradients is particularly useful. The most ground breaking development where surface pressure measurement is concerned was the advent of Pressure Sensitive Paints or PSPS. PSPs are essentially sensor molecules or ‘lumophores’ embedded in a transparent oxygen-permeable polymer binder. These molecules glow when excited with a light source of the appropriate wavelength. When illuminated, these ‘sensor’ molecules in the binder become electronically excited and move to a higher energy level and on returning to ground state, these molecules emit energy (Radiative or non-Radiative). In some common PSPs, the amount of radiant energy given off by the molecules depends on how much oxygen the molecules have interacted with in the excited state. Typically, the emission is inversely proportional to the amount of oxygen present. Hence in a homogenous flow, lower emission implies higher oxygen concentration which in turn implies a higher local pressure provided the mole fraction of oxygen in the flow is uniform. The measured PSP emission measured could be fluorescence or phosphorescence depending on the application and the lumophores used. Most modern lumophores rely on measurement of the phosphorescence lifetime since it is a very direct way of measuring the surface pressure since the lifetime obtained is an absolute number. Figure 1.10 is an example of an application of PSPs in obtaining surface pressure data. 19 Figure1.10: Surface pressure measured on a high speed civil aircraft. The colors represent the phosphorescence intensity ratio with blue representing low pressure and red representing high pressure. [Bell, et a1. [2001]] PSPs bypass ahnost all the issues that are encountered with traditional surface pressure measurement techniques. It provides a very high spatial resolution and the limitation in measuring the pressure distribution on a surface depends purely on the ability to effectively illuminate the surface and image it. There are however some major limitations where the use of PSPs is concerned. The biggest one being the inability to measure pressure fields within the body of a flow. Another issue is the transparent coating which holds the sensor molecules. The refi'active index of this coating changes as temperature changes and this in turn changes the manner in which the test surface is illuminated and the way it is imaged. Another more important local temperature dependent effect that needs to be corrected for is the variation of the 20 emission spectrum of fluorescence and phosphorescence. The Radiative emission spectrum of PSPS is highly temperature dependent and needs to be corrected for. Changes in local illumination also occur in case a model that is coated with PSPS deforms between wind-off and wind-on conditions which introduces errors. From an operations perspective, the PSP Should also be easily applicable on the test surface, robust enough to handle the flow moving over it and be chemically inert. The PSP coating should not modify the surface characteristics (adhesion of the flow, heat transfer co-efficient, surface roughness, etc) of the test surface since that would alter the nature of the flow. Furthermore the initial cost of instrumentation and the costs involved with coating a test surface are quite high. The proposed technique for MTM can potentially address a lot of these issues and will be covered in detail in subsequent chapters. 1.3.2 PRINCIPLE OF MOLECULAR TAGGING MANOMETRY Akin to the working principle of PSPS, MTM relies on oxygen quenching of phosphorescence lifetime of a molecule. However unlike PSPS which are applied on a test surface as a coating, the lumophore (which we will hereafter refer to as ‘the tracer’) is premixed in the flow. This is the primary advantage that MTM has over the traditional PSP approach. The tracer in our case will be acetone. In essence if one can determine the dependence of phosphorescence lifetime on oxygen concentration, a known quencher of acetone phosphorescence lifetime, then by knowing 21 the mole fraction of oxygen in a flow and its local concentration based on phosphorescence lifetime observed, the local pressure in the flow can be determined. The primary goal of this project will be to design a system that can provide a proof of concept for MTM: a system that can provide experimental data showing the change of lifetime with pressure. Potentially, MTM can provide at least the same resolution as PSPS if not better while not being limited to surface pressure measurements. 1.4 MOLECULAR TAGGING THERMOMETRY FOR GAS FLOWS The use of MTT so far has been restricted to gas phase flows. Part of this thesis also aims at extending the use of MTT to gas flows. The first step towards achieving this goal will be to determine the dependence of the phosphorescence (triplet) lifetime of acetone on temperature in the absence of oxygen in order to create a thermometer based on acetone’s temperature dependent phosphorescence lifetime. This will be done by building a test rig that provides a simple jet flow, the temperature of which could be controlled. This will not only provide us with a non-intrusive, molecular based whole field temperature mapping technique for gas flows but also data that will be useful when making quantitative pressure measurements in compressible flows where temperature changes as well. Keeping these two broad objectives in mind, the work for this thesis was carried out. 22 CHAPTER 2: ACETONE AND ITS PHOTOCHEMICAL PROPERTIES The primary tracer that will be used during the work described in this thesis is acetone. Keeping this in mind it is beneficial to go over some of the tracer’s important photochemical properties, Specifically those related to its phosphorescence emission. 2.1 OXYGEN QUENCHING The quantum efficiency (typically denoted by g5) is a measure of how efficient a molecule is in converting the photons incident on it into photons that make up its emission. The quantum efficiency of a tracer’s emission is different for fluorescence and phosphorescence. The quantum efficiency of acetone’s phosphorescence emission (1.8%) is approximately 9 times that of its fluorescence emission efficiency (0.2%) [Lozano, [1992]]. As a result, the phosphorescence lifetime of acetone (200us) is a lot higher than its fluorescence lifetime (z4ns). This however makes its phosphorescence extremely susceptible to collisional quenching. There are several species which have the tendency to reduce acetone’s phosphorescence lifetime. One of the most severe quenchers of acetone phosphorescence lifetime is oxygen. Acetone phosphorescence lifetime drops from 200us to 4ns when measured in air. The phosphorescence lifetime of acetone can be expressed in terms of oxygen concentration present using the Stem-Volmer relation: l=—1-—+Kq[02] Where, (Eqn 2.1) T To 23 r = Phosphorescence lifetime of acetone after being quenched by oxygen ro= Phosphorescence lifetime of acetone in the absence of oxygen K q = Quenching rate constant due to oxygen = 8000 moi/Iris. [Groh, et al. [1953]] [02] = Oxygen concentration. Figure 2.1 shows this effect. 0 '___ ___'___ ____________ '____ - 10 C:::C ______ I: _______ I ________ |__-_W I H p—a O I N p—a O Inverse of lifetime in us'1:(l/r) -3 10 O 0.2x10“ 0.4x10'4 0.6x10'4 0.8x104 1.0x10tl Concentration of oxygen in mol/l : ([02]) Figure 2. 1: Quenching of acetone phosphorescence lifetime due to oxygen. Oxygen concentration in air at STP conditions is 9.375 x10'3 mol/l. As a result most gas phase experiments that rely on acetone’s phosphorescence lifetime are conducted in oxygen free (typically nitrogen) environments (except in our case where the pressure measurement technique will rely on the presence of oxygen). 24 2.2 ACETONE SELF -QUENCHING Another mechanism by which acetone phosphorescence is quenched, which is relevant to MTM, is the ability of acetone to quench its own phosphorescence lifetime. When present in large concentrations, acetone is known to quench its own phosphorescence emission. This property of acetone is called self-quenching. Akin to Eqn 2.1, the phosphorescence lifetime of acetone can be expressed in terms of its own concentration using the Stem-Volmer relation: 1 =——1- + K0 [act] Where, (Eqn 2.2) T TO I = Phosphorescence lifetime of acetone after being quenched. r0= Phosphorescence lifetime of acetone for a single molecule. K a = Self— quenching rate constant = 12 moi/ins. [Lozano [1992]] [act] = Acetone concentration. Figure 2.2 shows this effect. 25 O ...—A O ::::I::::::::i::::::::l::::::::I::::::::I:::: ’___~I: _______ I ________ I ________ l______ ____ |___1 .__.-_I________I ________ I ________ I ________ I____, .__—-—I-———-—————l _________ I ........ I ________ I____. ,____l___._____| ________ I ________ I ________ I_____, I I I I I ~-——I-————-——I -------- I ———————— I ———————— l—-——‘ 7 I I I I I -1 I I I I 10 t____l:—_—__-—i ........ I_______ ______I.___. IL) Tl II I l I l 5" F_| | I l \ l——-— l_-_.____._ ...—a O Tl “fix III III Tflj III III III WIT I I I I I I —_—l I I I I I I I "‘“Il I I I I I I I I I | I II II II __" — II II II llLl Inverse of lifetime in us 30/!) I b.) p— — _ p— ._ _ P— — _— >— — ... p—r O 0 0.002 0.004 0.006 0.008 0.010 Concentration of acetone in mol/l : ([act]) Figure 2. 2: Quenching of acetone’s phosphorescence lifetime due to self-quenching. Concentration of acetone at 20°C and room pressure is 10.15 x10—3 mol/l. It should be noted that acetone’s self-quenching rate constant is not really constant but tends to be non-linear. This causes its Stern-Volmer plot (variation of yr vs. acetone concentration) or Eqn.2.2 to be non-linear and not a straight line. This is generally observed at very low acetone pressures (z30mm of Hg) and is highly wavelength dependent (see Figure 2.3) [O’Neal, E. and Larson, C. [1968]]. There are some references which don’t talk about this non-linearity explicitly and quote a linear form for acetone’s self-quenching constant (see [Lozano, [1992] HGTL report no.T—284]). Given the ambiguity and based on the concentration regimes dealt with and the excitation wavelength (308nm) dealt with in this thesis, this non-linearity has been ignored. 26 l Wavelength in mm 2'0 1' A i N 1.8 ' $ 3130 1.6 ’ r; 1.4 , . 3020 ES 2970 N L2 r . 1 o , ° 28904 0.8 0.6 0.4 0.2 l) 2600 Cr 50 100 150 200 Pressure of acetone in mm of Hg. A ..1 - l ”A ...... Figure 2. 3: Variation of inverse of quantum efficiency of acetone phosphorescence emission with acetone pressure and excitation wavelength. [O’Neal, et al. [1968]] The comparison of these two quenching mechanisms is shown in Figure 2.4. As can be seen from Figure 2.4, oxygen quenching is a lot more severe when compared to acetone’s self—quenching. However there are situations where acetone self-quenching dominates oxygen quenching. This will be discussed in detail in the next chapter. 27 O ...—I O I I—n 10 I N Inverse of lifetime in us”:(1/c) 0 0.2x10‘4 0.4x104 0.6x10“4 0.8x10‘4 1.0x10'4 Concentration of oxygen or acetone in mon : ([02] or [act]) Figure 2. 4: Comparison of acetone and oxygen quenching of acetone phosphorescence lifetime. 28 2.3 ABSORPTION SPECTRA [TTTIIVTTIIll—TIIITIITTWTIITTI- NE T - A 06i- -15 ’5‘ A .. - E a. . - T s: « " 33 - i ii 34... t °° E ..10. 3 g 0 . g. " U P - ‘ E a .. " 8 o .. o I"-2.... : U: 5 0 5‘ ‘3 2:: b .5 E ’5 3 .. ‘ kl £0ILIJIJIIIJ.11LLIIIJIIIILIL O 200 225 250 275 300 (E) 325 350 ' Wavelength in nm Figure 2. 5: Acetone absorption spectra. [Lozano, [1992]] The absorption spectrum of acetone shows that the peak lies between 270 and. 280 nm —20 C corresponding to an absorption cross-section (0') of 4.7x10 m2. The primary excitation source for the work done during this thesis is 308nm (XeCl Eximer) the a for which is 1.6x 10’20 cm2. The extinction coefficient (.9) shown on the right axis will be used subsequently to calculate the attenuation of the 308nm excitation source as it passes through acetone vapor. 29 2.4 PHOTOLYSIS Another extremely important aspect that needs to be taken into account is acetone’s tendency to photolyze when illuminated. The primary dissociation process is given by [Lozano, 1992. HGTL report no. T-284]: CH3COCH3 +hU—)CH3 +C0 The free acetyl radicals tend to combine and form biacetyl. This is the reason why many static cell experiments involving acetone tend to have a greenish glow when irradiated for a long time [Noyes, et al. [1975]]. Figure 2.6 shows the transition of the blue glow typical of acetone phosphorescence to the brighter greenish glow as biacetyl produced increases. Biacetyl is essentially a wall process and is inversely proportional to the pressure of the static cell [see Noyes, et al. [1975] and Howland, et al. [1944]]. The triplet dissociation barrier is 10kcal/mol. [O’Neal, et al. [1968]] ACETONE CONCENTRATION: z 5.63 x10’5 mol /1 Static cell pressure 0.067 Static cell pressure 0.067 Static cell pressure 0.067 bar bar (t= 0) bar (t= 2 mins) (t= 5 mins) Figure 2.6: Photolysis of acetone to form biacetyl. (These images were taken with a handheld digital camera only to serve as a qualitative representation of the photolysis phenomena.) 30 For the work conducted in this thesis, care was taken that the acetone concentration was high enough, the irradiating light low and the time of radiation minimal to minimize any spurious effects due to this phenomenon. This is discussed in more detail in Section 3.3.3. 31 CHAPTER 3: MOLECULAR TAGGING MANOMETRY The central premise of Molecular Tagging Manometry is the strong dependence of acetone phosphorescence lifetime on oxygen concentration. Trace amounts of oxygen is known to quench the phosphorescence lifetime of acetone. For a flow with fixed mole fraction of oxygen, variation in the local pressure leads to variations in the local oxygen concentration which in turn affects the lifetime of acetone. Thus, in principle, by knowing the local lifetime one can obtain the local pressure. However, during the course of the thesis it was discovered that for the range of acetone and oxygen concentrations used, and their relative proportion to each other, acetone’s ability to quench its own phosphorescence lifetime was much more( a: 10 times in some cases) than the amount of quenching that was caused by the trace oxygen in the flow. The following should help to explain this idea better. Based on the Stern-Volmer relation, taking into account acetone’s self-quenching and oxygen quenching, we have: —-1 r=[;1—+[Kq[02]+Ka[act]]] ps (Eqn 3.1) 0 32 Where, To 2200/5 Kq =8000 l/mol.,us Ka =121/mol.,us [Lozano [1992]] [act]: Concentration of acetone in mol/ l. [02 ] = Concentration of oxygen in mol / l. Eqn 3.1 clearly demonstrates that while the oxygen quenching constant for acetone phosphorescence lifetime, (K q) is almost 800 times that of its self-quenching constant (K a) [see Lozano, [1992], HGTL report no. T-284] (as shown in Figure 3.1 and 3.2), if the product of the self-quenching constant and the acetone concentration is greater than the product of oxygen quenching constant and oxygen concentration, then self-quenching would dominate. In determining what form of quenching will be most effective in changing the lifetime of the tracer in situations when multiple forms of quenching operate simultaneously, the product of the quenching constant and the related concentration becomes more important. In our case, the mixture used comprised of nitrogen seeded with acetone. The seeding was done by bubbling the nitrogen through a sealed seeding chamber (constructional details will be discussed in Sections 3.2.1) containing acetone. The concentration of acetone in the flow depends on the vapor pressure of acetone in the seeding chamber which in turn depends on the temperature of the seeding chamber as will be discussed next. The nitrogen used came from an industrial high purity nitrogen cylinder with 33 <5ppm of oxygen by volume. This was the source of the trace amount of oxygen present in the flow. Later provisions were made for increasing the amount of oxygen in the flow by adding oxygen separately or modifying the manner in which the nitrogen, acetone and oxygen mixture was made. Details connected to this will be discussed later. 34 .AQMV 2008:8280 nowhxo use 6:908 233 08:8: vacuumeonqgna 0:808 00 cougar» 3 .m 0.53.”— 308 E 2830 2868 we 5:838:00 Q ~ N.O~ O a 0H 0 o OH Y A: see E :No: - 5&on :ocwbcoocoo 0H Y N 0H 9 I I 00 O I 4 0w ....a.. ..... .... ..... x02 ...... J.. - mgwwmm Wmmflwmmzzw ..... meMm W wwwww W mm;_:/00N .... . ovfi ...... . . . .m1 5 CV 68505;... cocoaeonamocm oofi / 0 l0 $11 “I swam 35 0N ow 00 0m 02. ONE... 9; 02 .AQNV 508800000 :Dmxxo 0:0 0:380 0:3 0:50.:— 00:00m80:%0:0 0:808 .80 :0:ES> "N .0 «Earn 50:: E 3.00: 0:880 00 5:888:00 800: 0:0 Damm E 2:00: 0:880 00 5:888:00 25228 him a a N0 a 5330 00 :0:8::00:0U .2 E 30 8:085 00588050850 morn ut ](Z()) uofifixo jo uonenuoouog 36 3.1 ANALYSIS AND ASPECTS TO CONSIDER In order to get a better feel for how the aforementioned quenching processes work, it is instructive to go over the actual concentration of oxygen and acetone present in the mixtures used and how they were calculated. 3.1.1 ACETONE AND OXYGEN CONCENTRATION RELATED CALCULATIONS To calculate the concentration of acetone in the flow it was relied on the fact that the maximum concentration of acetone vapor in a flow, at a given temperature, is governed by its vapor pressure at that temperature. In the present work, the temperature used to calculate the concentration of acetone used in the experiments was the temperature inside the seeding chamber. At this temperature, the vapor pressure of acetone can be calculated using the ‘Fitted Antoine equation ’, [Lozano, et al. [1992]], (Eqn. 3.2, Figure 3.3.) loglo P[Torr]= 7.125267 + ”14°08 (Eqn. 3.2) 230.002 + T °C 37 Vapor pressure vs. temperature 3 I I I I I I I I I r I I I I I I I I I I I I I I ”-_l-——I-—T———_—l—_T—_I_____7-—T-_I_---—T-—I———l—_‘ r.__'__._'__'_ _____ '__l__'______._'__..'__.'___.___'___'___'__.. I I I I I I I I I I I I A .__-'___'__'_ _____ '___'__'______'__L__'_-_____'__'___'___., ca. I I I I I I I I I I I I > __I___I__I_ _____ I-.._I__I______I__I__-I______I__I___I-_ V I I I I I I I I I I I I 5.: I l I I L I l I I l I I <32 I I I I I I I I I I I I '9 ..-_I___I_..I.____._I__.L--I.__-__I__r.__I.._-__.t-_I__ --.. a I I I I I I I I I I I I 0) »——I———I——I——————I-—+——i;———--I——I-——I—-——-+—— ——l-——« S ___:___:__'+__ Antoine'sEquation: [Lozano,etal. [1992]] __' _:___:__4 8 I I I I I I I I I I I I ‘3 *‘"I"‘1‘-1'---—‘1‘—T‘-I“'——-‘I'-T‘—-l“ "T—"I‘—"I"—" g, I I I I I I I I I I I I 81 I I I I I I I I I f r I I I I I I I I I I I g" r'_"l_——I—_I_ _____ I_-T_—I_——-__I_ ~--'l—_A_—_I__l__—I"- > I__-'___'__'_ _____ '___'__'__ ___'___'______'__'___'__., I I I I I I I I I I I I ..__.'___'_..__'_ _____ I-.. _______'.___'___'______'__'___'__., I I I I I I I I I I ,___I__. ___I__._I__L_____I___L__I__ ___'__L.__.I_._ I I I I I I I _I I I I 8 0 l I l 441 l J‘l I 1 I l 0 20 4O 6O 80 Temperature in °C Figure 3. 3: Variation of vapor pressure of acetone with temperature [Lozano, et al. [1992]]. As a sample calculation, based on Eqn. 3.2 and assuming a temperature of 20°C, the vapor pressure of acetone is 0.2474 bar. When mixed with another gas, this becomes acetone’s partial pressure in the mixture. Depending on the pressure of the other gas, its mole fraction might change but not its absolute concentration unless of course the entire mixture is compressed or decompressed. The following is a sample calculation: Using the ideal gas law we have: P‘v’ =nRT: P = 0.2474 bar= 2.474 x104 Pa 38 = J R 8.314 mol.K T = 293 K 3% =10.15 x10—3 mol/l at room pressure and 20°C. The Antoine fitted equation in a tabulated form, along with the corresponding concentration and phosphorescence lifetime expected based on self-quenching in an oxygen free environment, is included in APPENDD( A. It should be noted that this holds only if the flow doesn’t expand significantly after it exits the seeding chamber. In the case of very high flow speeds, the pressure drop fiom the seeding chamber to the point where the flow is being measured needs to be accounted for. The ratio of the concentrations will be inversely proportional to the ratio of the pressure inside and outside the seeding chamber for isothermal flow conditions. This aspect will not impact the experiments for MTM discussed in this chapter but will be discussed in Chapter 4 (Section 4.5) when temperature based lifetime study is conducted. The buffer gas used was high purity nitrogen, the source for which was an industrial nitrogen cylinder (AIRGAS) with a peak pressure of 160 bar and volume of 304 ft3 . The oxygen content was rated at <5 ppm by volume. This rating was used to calculate the amount of oxygen that would be present in the flow. The concentration was calculated for room pressure in the following manner. 39 From basic thermodynamic principles we know that one mole of a gas occupies 24.4 liters at NTP conditions. Therefore one can conclude that the amount of the gas that occupies 1 liter is l/24.4 moles. Further given that the amount of oxygen in the nitrogen cylinder occupies 5/106 parts by volume, one can say that the amount of oxygen in 1 liter of the nitrogen used, at room pressure is, moles. This gives us 24.4x106 2.067 x10—7 mon of oxygen in the flow at NTP conditions. For all our analysis we will be assuming isothermal conditions since all our experiments will be conducted at room temperature in a static cell. Hence when the pressure changes by a factor, C f the oxygen concentration also varies by that same factor. Using this as the basic framework, we now proceed to a more detailed analysis of the various aspects that we need to consider, important experimental parameters and how one would go about choosing these parameters so as to obtain optimal experimental conditions. 3.1.2 SIGNAL RESPONSE FOR DIFFERENT COMPOSITIONS For the experiments that will be conducted, the starting gas mixture used will be prepared ahead of time and then isotherrnally compressed and decompressed. This implies that the proportion of acetone and oxygen in a flow will not change with respect to each other. However their local concentrations will change depending on the local pressure. 40 Our main concern is in obtaining the maximum change in the signal as integrated by the photo-detector as a measure of the sensitivity of the technique in detecting a pressure change, it would be beneficial to study the changes in lifetime or rather signal response as a given mixture is compressed or expanded. We will use room pressure or 1 bar as the reference and observe the change in signal ratio as the pressure is changed. From Eqn. 1.2 we have: SR =e—(t2 —t1)/r = e—(td +At)/z' And fi'om Eqn. 3.1 we have: —1 r=[;1—+[Kq[02]+ Ka[act]]:l ,us 0 If we adapt Eqn. 3.1 in a manner such that we assume that the mixture was made at room pressure and we then proceed to isotherrnally compress or expand it, we can re-write Eqn.3.l as: —1 l r=[-;;+Cf [Kq[02]+ Ka[act]]:| ,us (Eqn. 3.3) Where, [02] is the concentration of oxygen at room pressure. [act] is the concentration of acetone at room pressure. C f is the factor by which the pressure of the mixture changes. C f = l for room pressure, 1 for pressures above room pressure. 41 Using Eqn. 1.2 and Eqn.3.3 gives us: _(td +At)/[—1—+Cf [Kq IOZI+Ka [act]]:l SR = e TO (Eqn. 3.4) Figures 3.4, 3.5, 3.6 and 3.7 show four plots of Eqn. 3.4 wherein the starting concentration of oxygen, compared to acetone, in the mixture is steadily decreased. Each plot represents a fixed starting oxygen concentration and shows how the signal ratio changes as the mixture pressure changes. For each oxygen concentration, four different acetone concentrations are shown. The acetone concentrations are calculated using Eqn.3.2. The idea is to illustrate how one quenching mechanism gains precedence over the other as their relative proportions are changed. The plots also serve to provide a quantitative feel for how the signal ratio changes with pressure: a measure of MTM’S sensitivity. For the plots shown, td is 0.51.15 and the exposure time, At is 5 us. 42 Signal ratio vs. pressure: both acetone and oxygen present in the mixture. 10'1 I--ELI-EIIIL--IL---:u--L-IILII-:I--:III- HIIJIIIJIIIJIIIJIIIJIIIJIIIJIIIJIIIJIIIL IIIJIIIJIIIJIIIJIIIJIIIJIIIJIIIJIlIJlIl 2 I .0 rl IIILIII 3 ) HI..-- .Wmu IIIJIII no N m IHHHI u... 7 _ _ _ _ _ _ _ m :0 III I IJIIIJIIIJIIIJIII IIIJI .II I” _ . . _ J. . V m m .m IIIIIIILIIILIIILIIILIIILIIILI mm m e m ... m _ 2.. .... .0 w _ 03 O I 4 C II I. I 1.... 1 X II I 0 0r. IIM - ..m M rm M 1P HIHm H2. 1 5. l. I II. I 00 .I. 1 2 _ .. .. .- .. I l. m m ...- n%@ms%w I I” M. m w .m- w M M M I It. cm __ m.- m m- l 2 3 4 s ...-.... P @@@@ H mg m .m m r. r. r. <... L 5m I. e e e ..W. ”H- m. .m .m mu .m m .m m w IHHH. n minim m mam meme-m IIII_ cc ”I . I c C c C _mcc Luke m I--. ms mm Msm .m-.gm.m..m. 1:8.“ Im mma “Ema: ILMWM n8 wmm mum-karma.“ _I __ __ s _ _ _ _ .3. 0.. ..me O “9WWWVMK TWWMMC 4.. 1 9 E 7. 6. 5 4. 3 2 1. 0m 0 0 0 0 0 0 0 0 0 Cm: 3.3. 1.5.0 Variation of phosphorescence signal ratio with change in pressure: Figure 3. 4 9.375 ><10‘3 mol/ I (concentration of oxygen in air at NTP conditions.) [02 I In Figure 3.4 it can be seen that all four curves (representing a mixture of air along with the four different acetone concentrations) lie on top of each other. This basically implies that the change in acetone concentration made no difference to the response curve obtained. Recalling Eqn 3.1, this makes sense since the concentration of oxygen is high, the product of K q and oxygen concentration [02] is much higher than the product of 43 K a and acetone concentration [act]. — I : fin??? : : :E:E?:? : :Oxygenconcentranon= E 09.-- ‘-—1-1-1-1—1111--—-1-1-111-1111--—-1--1“‘0‘h°fai’ .2 ' : .::::::: : :::::::: : :ntlbn0r9-375n'0" : 08—_-_|_t- I_I_IIII____I_I_I IIIIII____I__I 0101/]. I - T 7 1 . ':‘.T: F T 11111.”. : {deunwbenveenmestartsof T 1 1 :1. 111 1 11111111 1 1thefirstandsecondimgesO-5ns 1 0.7—-‘°r-_I-T-l'l’l TI"""I""l'F'l‘l'le'I""’I""T ‘T‘ r; 1 111111.‘ 1 11111111 1 :Ka=121/mol.us 1 @0.6—--—1———1—-1-1—1-1—:1—1———1—--1—1--1—:-1I1—1-—--1--1—Kq=8000VrmI4ts :~ 3 111111111 111111111 11 1 '5 05____I___I__I_I_I_I_II_I__ _I___I_I__I_I_I1I_I____I__I_ Theconcentrationof[act] at L 5 : : ::::::: : : ::::::: : :roompressurehasbeenexpressed : E 1 1 1111111 11111111 1 1intenmofthepressureofthe 1 204”"?":”:':":':‘:T:"": ":':“:‘:‘:T:‘:"":":‘ min Concennnnons T w I I Illllll l llllllll I Ishownareatlbar. I 0'3—___{-_—:—-%-:-:-:_:-:-:_---:-— -:_-:_:-:%:_:--_-:_—%' I I Illll I I I Illlll l I IIIII: 02___-1-_1_1_1_1_1111____L_J 'JJ11L1____1__ —Seedingchamber@15degC: 8.21x10'3moI/1 E E 5 E5555: E E 1 :EEEEE E —Seedingchamber@22degC:11x10'3mol/l O-ln"-1--:-1-1-1-1J.11---+‘I-f. IIH----:-- —Seedingchamber@30degC:15.1x10'3mol/l 0 man. 11- L —Seedingchamber@40degC:21.76x10'3mol/l 10'3 10'2 10" 10° 101 102 Pressure in bar abs: (P) Figure 3. 5: Variation of phosphorescence signal ratio with change in pressure: [02 ] = 9.375 x10”4 mol/ l at 1 bar (1/10’h of the concentration of oxygen in air at 1 bar) Figure 3.5 shows that the four curves (though barely discernible) have just started to separate out. This implies that the change in acetone concentration has a substantial effect on the phosphorescence lifetime of acetone and in turn makes a difference to the response curve obtained. However the product of K q and oxygen concentration [02 ] is still higher than the product of K a and acetone concentration [act] and oxygen quenching still dominates. Signal ratio vs. pressure: both acetone and oxygen present in the mixture. Io2 = fl mcm HHHL mm .amm . ”mam m ,: ma m to MIM- .Omsm. S .w mmmm LM 0. m III 1 Mr: aw uh mu m h .0... III w wo w fim VW m .1 mm III mm mm0m as ..m.. m =_ fl. Wmmume mm mwmm Wm urn-mmmmmmwuuuwunuhnnu-I" owm. mwmm m .................. I _ .. u M b IIIrIII III m m II ................... mm mm... .m .... .... .0 m n m....o m m I. m em. -mm-mm-mmunmm m. m m m mm mm H”- - ”H H” HHHHH on 1. -.-. a H H ”H H“ H” HHHHH finmammou- -\ ................ w M. M M -- 1 2 3 4 1..- ----------------- @@@@- I ........................ m m m m - a H uuurnnuruuu unnunuunun m ..m m mum m H“THU”_mHHWHHHHHHHHHHH W New Wu“ 4-+IT+:IIImmmm- 1.4-1.4--_1---.w--”---"-- & & a“ nu -- LIIrIIIrIIIrIIITIIITIIITII II n n H u u H _ _ _ _ a a n h u n -I m ._ ...o 1.387.654.3210] 0 0 0 0 0 0 0 0 0 Te 3:5— _aflu.m Variation of phosphorescence signal ratio with change in pressure: Figure 3. 6 9.375 x10"6 mot/1 at 1 bar ( l/lOOOth of the concentration of oxygen in air at 1 [02] bar) Figure 3.6 shows that the four curves have now clearly separated out. This basically implies that the change in acetone concentration is now making a substantial difference to the response curve obtained. The product of K q and oxygen concentration [02 ] is now and acetone concentration [act]. In this case both (1 almost half of the product of K acetone and oxygen quenching play equal roles. 45 Signal ratio vs. pressure: both acetone and oxygen present in the nuxture. 2 r ILIIILIIILIIII m n HUHHHUHHHUHHHH . _ _ I HUHHHUHHHUHHHH I _ _ _ I IJIIIJIIIJIII- .I |I_IIII.IIIII._IIIIL = . _ _ _ _ _ I Mr. m M. W IlIIIIIIIJIIII .UO MS 8 ..m _ _ . M Di. . MP S _ _ _ b wwwOMI. m _ _ . V l 8L ILIIILIIILIIII 1.0 .f..¥am.. -IIEI:I r Ot7 em my h“ . IUIIIUIIIUIII F a. b m 0 m I III IIIII - cum I. v0 .0 -..-111.1 ,mmu-momnm .n I:-+:I - u U o _. __ m a II_IIII_III .I I n-_ma «mm om II:----- I 0N2 .08 S C _ . . _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ \I _ _ _ _ . 0 nunnnnnunnn”nunUIInJIIuJIIIUHIIunuu-Ifiu”Ir 1010 IIIIrIIIrIIIrI II I_ IIIIIIIII LIIILIIII .. IIIITIIITIIIT II IIIII .I iIIIIIIILIIII .m IIIITIIITIII I _II _III IIIJIIIJIIIJIIIn a IIIITIIITI IT IIIII . ILIIILIIILIIILIIII r Flllrlll I I |_| |.l.ll|_llll.l “I a _ _ _ . _ _ IIIITII I I _IIII_IIII_IIIL- m m 01 b . _ _ _ _ . mm m m .m IIII_I I_ I IIII.IIII_IIII_IIII_- O 1... I e _ _ _ _ _ .....J m1... 0 r. _ _ _ _ _ . 01... 0 H u 1 1 _ u n u u u ..I. m -... m I. m IIHH I Hn-IHHHHHHUHHHUHHHUHHHUM 2 X . .n 0 r III IITIIITIII_IIII_IIII_IIII.- . .l. 5 111 19.. II IIWIIIWIII_IIII_IIII_IIII_. 0m 1.1 IN «MI II .IIIWIIJIIII_IIII_IIII_IIIJ- I :.:437£:I33Iw§w$ II. _IIIIrIIIFIII_IIII.IIII_IIILI C C 6 81 ..x. _ _ _ _ _ _ mm mm“ m M II I_IIII.IIII.IIII_IIII_IIII_IIIJI I .0 _ _ _ _ _ _ _ 1 2 3 4 I IrIIIFIIIFIII_IIII_IIII_IIILI I .5 . . _ . _ _ _ @@@@ _ _ _ _ . _ _ _ a mi» fl .nlv _ . _ _ _ _ _ b ..D ..D b I.II_IIII_IIII_IIII_IIII_IIII_IIII_I m m m mIQO I IITIIITIIITIIITIII_IIII_IIII.. a a a 31 .I. I IIfiIIIw-IIJIIIJIIII_IIII_IIIJI h h h InI I.II_IIII_IIII.IIII.IIII.IIII_IIII_- C C C CH H.UHUHUHHHHHHHHH m. m. m. m.- _ . _ . _ _ _ . . . . .IITIIITIIITIII_IIII_IIII_IIIJ. d .m d dfi fl _ _ _ _ _ _ . C C C e _ _ _ _ _ . _ _ _ _ _- _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ 3 b _ h — _ _ _ _ _ .0 l 9 ob 7 6. 5 A 3. 2 J 01 O 0 O 0 0 0 0 O 0 Ce 3:5— 1&5 Variation of phosphorescence signal ratio with change in pressure: Figure 3. 7 [02] 2.067 x10’7 mol/ 1 (concentration of trace oxygen in the nitrogen used at 1 bar) Figure 3.7 shows that the four curves have now separated out even more implying that the change in acetone concentration is now making even more of a difference to the response curve obtained than before. The product of K q and oxygen concentration [02 ] is now ahnost / th of the product of K a and acetone concentration [act]. In this 80 case acetone quenching dominates. For even lower oxygen concentrations, the curves would hardly change and would depend ahnost only on the concentration of acetone. 46 3.1.3 GATING SCHEMES Referring to the curve for acetone at 15°C in Figure 3.7 (red curve), for a factor of 13 drop in pressure, the maximum increase in signal ratio is #925 (from 0.1 to 0.925). Though mathematically it may appear obvious that to detect a 1 percent change in pressure one needs to detect a 0.71 percent change in signal count, with practical imaging systems, being able to detect difference in counts with such a degree of accuracy is difficult. For the imaging system used, an intensified CCD camera (12bit, P46, DiCAM Pro, Cooke Corp), that change would translate to an increase in signal count of only 29 if the entire dynamic range of the camera is being used (all 4095 available counts). In most cases it will be observed that the phosphorescence signal for the first fi'ame (S ,1) is well below the full 4095 count limit. Factors that limit the amount of light captured by the camera are laser power, amount of acetone in the flow and maximum size of the lens aperture. Also in the case of relatively highly quenched flows, the lower lifetime further reduces the amount of signal counts available; particularly so for the second frame. The maximum possible value of signal ratio (ratio of signal count as integrated during the second image to that integrated during the first) is one, as is obvious fiom Eqn. 1.2 and the fact that we are imaging an exponentially “decaying” function. All this greatly limits the use of this technique in measuring small pressure changes. In order to broaden the applicability of MTM, we need to improve the sensitivity of the signal change. Apart from using a camera with the ability to take images larger than 12bits, is to increase the maximum value of the signal ratio that we get beyond unity. This can be achieved by using a larger exposure time for the second image compared to the first and keeping the value of td as small as possible. 47 Eqn. 1.2 was derived in Chapter 1(Section 1.1) under the assumption that the exposure times for the first and second images were equal (Atl =At2 = At). If the exposure times are not equal, i.e.: At] ¢At1, then we have: Sq = _Iorle—(t1+At1)/z' _ e—tl ”J Stz = —Ior[e—(t2 +At2)/z' _ e—tz /z'] T§£=S R which we denote as ‘Signal ratio’. From these three equations we get, tl _ _ -(At2)/r SR: 8 (td +At1)/r|:l e ] (Eqn.”) 1_e—(At1)/Z' Eqn.3.5 shows that if the first exposure time is smaller than the second exposure time, the — At )/2' l—e ( 2 denominator of the term [I —(At1 ) / z- ] is smaller than its numerator thus making the —e term larger than one. In fact it can be much larger than one. Using unequal gates (exposure times) can give us a higher sensitivity using the same instrumentation. The effect is even more pronounced for shorter gate periods and longer lifetimes. There is of course a practical limitation on how long the second frame can be made before the phosphorescence signal gets buried under the camera noise. The plot in Figure 3.8 demonstrates how the signal ratio changes with pressure for the case when we use different gate periods and when we use equal gate periods. 48 Signal ratio versus pressure: use of equal and unequal gates m. m. unnmnunnunnw IIIJ IIIIIII 1 5 5 I... ....... 1 0 .2 IIII ....... I m. m = = IIIL ....... I 5 5 ta ta _ _ o 2 m m ....... .. = = IIII. IIIIIII _I td tam w _ _ m m .m .m _ _ 5 5 ----.111-.. ....... <5 5 II 1 nunrnnnnunnr m\m y .. .s. I. “I _ m m m m11_1-11.1-11- R. mu m. E. III.“ IIIIIII _I .m .m .m .w _ _ u u u IIIIIIIIIII .m m w m u _. Illl. lllllll r. . _ _ _ _ _ _ . _ .I llllll _ IIIIIII L IIIIII rI IIIIIIII III-“Ina- nnnnunuumunquIu IIIII I IIIIIIII _II'|-'- III Il'lr IIIIIIII _IIIII IL IIIIIIP _ _ IIIIIIII wII II JIIIIIIIflIIIIIII IIIIIIII _I I IIILIIIIIIIP _ _ _ _ fi IIIIII _ IIIIII J IIIIIII “I _ _ . _ _ _ _ _ III II. lllllll J IIIIIII 1 Hum IIII. IIIIIII L IIIIIII r rII III. IIIIIII J IIIIIII I III III. lllllll l. IIIIIII 1 lllllll .1 4| IIII.I _ " I'll—l _ T ..... u- _ r = = N am m m H nunuu.n.mlW.n HHHHH mm mm .m IIIIIIII “I “H em ........ _- e e7. mm m IIIIIIII .I mmoo.) .hu . _ W.I.H.AC.0 am IIIIIIII _I N7 1 U10 nf m x n] _ gomm4 em IIIIIIII _I math/.601. m . m Ico am _ 0 0A1 _ . . _ _ 7 00 6 4 I'llIIII"'I-Ilnl'l. IIII’l 'Il'l I‘liL Illll Illl IIII. 10° Pressure in bar abs: (P) lua'l I II. I '1 IIIIIIL -|'l .1 r'Il 0 1 I'll. [l'l Ilil 1"1 [I'lll I'll I'll. IIIII. IIII. .Illvl we Hoes. :36 Figure 3. 8: Signal ratio versus lifetime: comparing the use of equal and unequal gate periods. The oxygen concentration for the plot shown is 2.067 x10’7 moI/l (similar to the concentration of trace oxygen in the nitrogen used at 1 bar) and the concentration of acetone is 10.94x10—3 mol/I (based on vapor pressure of acetone at 22°C). The concentrations are measured at room pressure and temperature. For each gating scheme two I d ’s have been plotted, 2.5us and 0.5 us which is the shortest t d the DiCAM Pro can handle. In order to make the comparison fair, the amount of light integrated is made the 0.5 us integrates the same for the two cases, i.e. the unequal gate period curve with td 11.5us same amount of light (lus for gate period1+0.5us for t d +10us for gate period2 worth of light) as its equal gate period counterpart (5.5 [13 for gate periodl+0.5 us for 49 td +5.5 us for gate period2 = 11.5Irs worth of light). The same holds for the pair of curves with Id =2.5us. Figure3.8 demonstrates two things: one that using unequal gates can increase the sensitivity of MTM. Second, the larger the gap between the first and second image (I d ), the steeper is the response curve. The option of using unequal exposure times, though promising, should be used carefirlly when acquiring images since a ten-fold drop in signal ratio means a ten-fold drop in signal count too. One needs to be careful that the counts measured after the drop isn’t too small or the percentage of noise present in such a signal will be large and this will introduce errors of its own. Further, there is a limit to how small one can make td due to constraints of by the imaging system. The DiCAM PRO allows the minimum value of td to be 0.5 us. Even then, there is a possibility of a small fraction of the first image that was integrated on the CCD getting superimposed on the second image. This is because the P46 fast phosphor used in the CCD has its own finite decay time. There is a finite time period before which the phosphor of the CCD glows brightly enough for it to be seen on the second frame integrated. This is referred to as a ‘ghost image’. Discussions’ pertaining to this is covered next. 50 3.1.4 DETECTOR RELATED ISSUES: GHOST IMAGE AND FLUORESCENCE LIGHT. The imaging system used was an intensified CCD camera (12bit, P46, DiCAM Pro, Cooke Corp.). The DiCAM PRO has the option of operating it in a mode called the ‘dual- frame’ mode wherein two images can be taken in rapid succession following a single trigger pulse. The camera however does impose a lower limit of 0.5us on the delay between the end of the first image and the start of the second image. We refer to this delay astd . This is the minimum time that the camera CCD needs to off-load its charge from the first image and start taking the second image. However, even when this 0.5us time delay is maintained and the entire charge on the CCD has been removed, the glow of the phosphor on the CCD remains and this gets integrated on the second frame. The specifications of the camera provided by Cooke Corporation has shown that for a td of 2 us, around 1% of the first image is present in the second image and for a rd of 0.2-0.4ps, around 10% of the first image is present in the second image. The P46 phosphor does not have a single exponential lifetime, hence trying to calculate what the decay rate would be is hard. However it is reasonable to assume that any value of td greater than 1.5us should be good enough to neglect the ghost-image effect. Another possible way to remedy this is to use unequal gate periods. In that case since the amount of light integrated during the first frame is generally smaller given its smaller gate period, not only will the actual amount of ghost be smaller, but also the percentage of the total light on the second image due to the ghost will also be lower since the second image will have integrated a lot more light given the longer duration (provided the value of td and At] isn’t too high). 51 Another issue that one needs to be paid attention to is the light that is being integrated by the CCD is purely phosphorescence light and not fluorescence light since the entire premise of MTM is based on phosphorescence lifetime measurement. As discussed in Chapterl, the fluorescence lifetime of a tracer is typically a few orders of magnitude lower than its phosphorescence lifetime. When a flow containing a tracer is excited by a laser pulse, both phosphorescence (unless quenched) and fluorescence emissions are observed simultaneously. Fluorescence is much brighter and lasts for a short duration while phosphorescence is less bright but lasts much longer. Figure3.9 illustrates the basic idea. It should be noted that Figure 3.9 is only a broad representation and not a plot specifically for acetone. In case of acetone, the slope of the fluorescence curve is even steeper. Broadly speaking, the flow conditions that we are dealing with don’t really affect fluorescence lifetime when compared to its effect on phosphorescence lifetime of the tracer. Fluorescence though short lived is extremely bright for the period that its emission is detectable and imaging that emission will introduce a substantial amount of error and it therefore becomes important to understand when a flow is imaged, depending on the gate period and the delay that we use, how much error is introduced in phosphorescence detection. The following calculations illustrate how this is can be done. The primary reference used is [Lozano, et al. [1992]]. 52 COMPARISON OF FLUORESCENCE AND PHOSPHORESCENCE H O —-——1 ——-—< ———< —1 —1 OO | I I ' ' ' ' In Phosphorescence l l l I I I i In Fluorescence I l | I I ’ ‘——l—————"‘I*"_——"'1———‘"--—I—-——--—‘I——_‘ l l | l | Instantaneous intensity (arbitary units): (It) 0 . 20 40 . . 60 so 100 Trme elapsed after excrtatron laser pulse 1n us: (t) Figure 3. 9: Comparison of relative intensities of fluorescence and phosphorescence for a general tracer. If the subscripts ' f ' and ' p' represent fluorescence and for phosphorescence respectively, then from basic photochemistry theory it follows:. NxC x¢f,px0'=(10f,p)x(rf,p) (Eqn.3.6) Where, ¢ = Quantum efficiency a = Absorption Coefficient I 0 = Initial intensity 53 r = Exponential lifetime This is in essence a measure of the number of photons emitted by a unit volume of the fluid containing tracer concentration ‘C’ with ‘N’ number of photons striking the unit volume of fluid. For acetone we have: If = 4ns 1,, = lOus (this is the figure used because this is a more realistic number for acetone phosphorescence lifetime we expect to measure compared to the 200118 reported in chemistry papers where experiments are conducted with acetone present in very low concentrations and the self-quenching issue doesn’t come into the picture.) For the calculations shown below, N and C are the same for both florescence and phosphorescence cases since the fluid being excited by the laser is the same. For the sample calculations shown, the gate period is 5 us and the signal is integrated from t= 0.1115 to 5.51.13. Here we recall Eqn. 1.1: St =_IOTle—(t+At)/T _ e—t/TJ Based on the assumed delays and exposure times, using Eqn.l.l and Eqn. 3.6 we have: Sp =— N x Cx (15], x ale—(O'1+5)/10 — e—S/IOJ and (Eqn. 3.7) —3 —3 Sf =-NxCx¢f xa[e'(°'1+5)/(4"10 )_ e—5/(4x10 )] (Eqn. 38) 54 S Taking the ratio of S f and S p and plugging in the necessary numbers we get 19; is P equal to zero. Therefore we can be sure that maintaining an zOJus delay after the laser pulse before taking the first image will ensure none of the light integrated is due to fluorescence. Since the jitter of the Excirner is z i0.lus, a delay of at least 0.25us after the laser pulse is maintained while taking images. With the above understanding of the fundamental aspects upon which MTM is based, the details of the actual experiments and associated results are discussed next. 55 3.2 EXPERIMENT The experimental set-up basically comprised two main parts: a simple T-section shaped static cell, optically accessible from three sides and a flow system capable of seeding nitrogen with acetone and filling the test cell. The purpose of the test cell was to hold a pressurized mixture of acetone-seeded nitrogen and oxygen and maintain its pressure long enough in order for the mixture to be illuminated by a laser and imaged. The test cell would be filled to the peak pressure desired and decompressed in steps by opening an outlet valve and allowing the pressure to drop to the desired value. The block diagram shown in Figure 3.10 shows the layout of the flow circuit used to produce the acetone seeded nitrogen mixture. Figures 3.11-A shows the main parts of the static test cell and the connections leading into it. Figure 3.11—B shows the important dimensions of the test cell and Figure 3.11-C shows an actual photograph of the test cell. 3.2.1 SET-UP As shown in figure 3.10, the nitrogen flows out from a standard high purity nitrogen cylinder (AIRGAS, high purity nitrogen, <5ppm of oxygen) through a dual stage pressure regulator (Reg) which is the first flow rate control device. This provides the upper limit and bulk control of the flow rate. The flow then proceeds to a flow control valve (F C V) which provides finer control of the flow rate. From there the nitrogen flows to an on-off type solenoid valve (SC). The outlet of the solenoid valve leads to a damper, which is essentially a tank with a volume large enough to smooth out small scale pressure fluctuations, prevents splashing of acetone in the seeding chamber and help maintain a stable flow rate. The outlet of the damper leads into the seeding chamber containing the 56 acetone. Inside, the nitrogen is directed to the bottom of the seeding chamber and bubbles through the acetone, coming out from another outlet which then flows out from a tube (PARKER, Nylonll tubing, 1 ” OD, 0.062” wall thickness and z 2m in length). Nylon 2 11 was used given its resistance to acetone as a solvent. The acetone used in the flow (J.T. BAKER, 9006-33, CAS # 67-64-1) was excited using a 20ns pulse of a 308nm, XeCl excimer laser (Lambda Physik LPX 200) and imaged using an intensified CCD camera (12bit, P46, DiCAM Pro, Cooke Corp.). Typical energy level per pulse was 125- 145m]. The test cell can be filled to a maximum pressure of 25bar. The test cell was filled to 14bar and left for a week to check for leaks. No drop in pressure was recorded. This assured that there would be no change in pressure during an experiment. For the experiments conducted the maximum pressure used was 13.2bar (gage). There was a pressure gage mounted on the test cell rated for a maximum of 400psig. (ASHCROFT: Model: 10088 02L 400.) There are four l/2”NPT threaded ports on the test cell. Details pertaining to these are described afier Figure 3.11-C. There are also two MI” NPT threaded ports on the test cell, one of which is used to mount the pressure gage and the other was provided to hold a T type Cu Constantan thermocouple (not installed for the experiments conducted) with a maximum rated temperature of 260 °C . 57 3:86 32.: $212 "A: .m «...—mum 93mg «Soho 30E bag-8:2: mfiwwfi 330202 IIIIIIIIH .n 53 .533 0:598:88 0:32:00 .8:ng a :owOHEA 00m 33m :owobE _ adj 3:03 0:803. 302:3 mama: 30:53 .35 .693 / 5on 7:833:31.- «8280 III 0:895: II 5%:on wE—eoom ”.830 6m ....Ol._ :83 .834\ to: 8.593 \\: 320:3 EB Home-H to: em; owew 0.5%on “Swamp: 9:535 awed “39:0 35:8 32m “DU-m 03$» gone—om "Ix/m 58 3:085:03 32.2 :0.“ com: :00 38 0:06 2: 000: mafia“: 20:00:08 .00 emumfionom 3?“ w .m 0.53% .200 $0: 05 mo 00-305 ”SE-E 033 3:82 30m :0w0E: 00000.0. 0:00.00».N 309:3 wfiwaafi : \ 300:3 X :08 Home-H BOMVHHR \ 30:0 .595 Z \: *1/00: 03:5 to: 8283/ 0w0w 8381/5 59 90mm L'-‘-——1'50mm——-D----‘—135mm—D-w 35mm J J J Figure 3. ll-B: Dimensions of the static test cell used for MTM. 60 4Smm—-~ ‘— .252 go.“ wow: :8 52 235 05 mo :QBwBQE "0-: .m PEEK 61 The connections into the test cell (referring to Figure3.11-A) and their respective functions are as follows: 1. A% ” NPT connection with a manual valve was used to fill acetone and nitrogen in to the test cell. The inlet pipe attached to this connection came from the outlet of the seeding chamber. This will be referred to as the inlet port. 2. There was a similar manual placed at the exit of the seeding chamber. 3. A% ” NPT connection with a manual valve, referred to as the vacuum port, was connected to the vacuum pump. 4. A% ”NPT connection with a 14 ” NPT manual valve that was used as the outlet for the test cell. This was used to reduce the pressure of the test cell by opening it partially and letting out the pressurized acetone and nitrogen mixture. This will be referred to as the purge or exit port. The purge port was directed into a fume hood. 5. The pressure gage mounted on the test cell had a manual valve connected to it, which was kept closed when the test cell was being purged so that the gage doesn’t get damaged when the system was being evacuated. The structure of the test cell window is shown in Figure 3.12. The two leftmost flanges shown, house the quartz window. The whole assembly along with the rubber washers (black discs) forms the optical window. Three such windows are mounted on the flanges that are integral with the T-section. In the Figure 3.12 only one window is shown. The T- section and the flanges are made of stainless steel. 62 The sealing for the test cell is achieved using four washers per optical entry point: two that provide sealing between the window and the flanges and two that seal the window and the flange that is integral with the T- section. The washers used for sealing are made of Aramid and are 1/16th “ thick and 50mm OD and 35mm II). Aramid was used given its resistance to acetone as a solvent. There is a Buna-N o-ring around the quartz discs (inside diameter 1.75” and the diameter of the annulus is 3/ 16”). An inert vacuum grease (Dow Coming) is used to obtain additional sealing. Each window was held together using 8 3/8” bolts. O-RING ARAMID WASHER Figure 3. 12: An exploded isometric view of the test cell window. 63 3.2.2 PROCEDURE The next step was to come up with a method by which a mixture of known composition and low oxygen concentration could be made. The idea behind trying to reduce the concentration of oxygen is that oxygen is far too strong a quencher and if its concentration is comparable to its concentration in air, almost no phosphorescence is detectable. In order to get the needed conditions, alternate cycles of vacuuming the test cell and then filling it with a high pressure mixture of acetone seeded nitrogen. The test cell was filled to a peak pressure of 13.2bar (gage) or 190psig and vacuumed to a pressure of 2” of Hg abs. The following steps are the detailed steps that were followed: 1. First the nitrogen from the cylinder is allowed to flow freely through the seeding chamber into the test cell. The inlet port and the outlet ports of the test cell are kept open. This is done to bubble the oxygen dissolved in the acetone out through the purge port of the test cell. The system is allowed to flow freely for around 3-5 minutes. This is generally enough to purge the acetone of the oxygen dissolved in it. . This process was limited to a maximum of five minutes or else the acetone level in the seeding chamber dropped too much. After this, the inlet port was closed and the exit port kept open. Following this the vacuum pump is turned on for a while. This refills the test cell with air. Note that at this time the line leading right upto the test cell inlet port from the seeding chamber was filled with nitrogen and acetone due to the line purging process just described. 64 6. After a couple of minutes, the purge port and the vacuum port were closed. 7. The vacuum pump is then turned on again. If we assume that the test cell is now filled with pure air at one atmosphere (208000 ppm by volume (mole fraction of oxygen in air) or9.375><10—3 ”7011 ), then purging it to 2” of Hg will drop the ' 30' 28 '3 mol — ‘4 mol concentration of [02]to l: 30 :lx9.375x10 K —6.25x10 /1. The pressure inside would now be 2” Hg absolute or 0.96psia. Atmospheric pressure is taken as 30” of Hg. Note that the ‘ppm’ value remains unchanged = 208000ppm. That means the molar fraction doesn’t change but the absolute concentration does. 8. We shut off the vacuum pump and the valve it is connected to. We then fill the test cell again with nitrogen and acetone upto a pressure of 190 psig. This should drop the concentration of [02] from 208000 ppm to: [fl]x208000ppm = 976.1 ppm. However its molar concentration in the 204 4 mixture will be unchanged since the volume of the test cell is fixed. Pressure is now 204.4psia. 9. Here it is important to note that we are assuming that the nitrogen that we are filling has no oxygen at all, which is not accurate. This will be corrected for in the end. 10. We now evacuate the test cell again to 0.96 psia. The ‘ppm’ value once again remains unchanged at 976.1 ppm. The concentration of the oxygen in the test cell 65 however drops to: 0'96 ’4 mol — ‘6 mol [——]x6.25x10 {—2.935 x10 1 204.4 11. The chamber is filled upto 204.4 psia again. This keeps the [02] in the mixture at the value obtained above and only drops the ‘ppm’. 12. The nitrogen that we use has 2.067 x10—7 "10% of oxygen at 1 atm. Therefore if the test cell is filled to 204.4 psia or a gage pressure of z13.2 atm, then the concentration of oxygen in the nitrogen will be 13.2x2.067x10_7 ”70% or = 2.73x10'6 mol/l at 13.2 atm. Therefore the total oxygen concentration will now be 2 5.66x10—6 mol/I (adding the oxygen in the test cell at 13.2bar as calculated in step 10) at 204.4psia. This is approximately the concentration of the starting mixture that will be used for the experiment. This is will be expanded to room pressure in steps, with data being collected at each step. Precaution was taken when the seeding chamber was being filled to make sure that there were no acetone droplets coming out of pipe from the seeding chamber that was connected to the test cell. The filling of the test cell was done extremely slowly in order to make sure that the concentration of acetone in the flow remained constant. Since the seeding chamber was at 13.2bar and the test cell vacuumed, the tendency of the nitrogen from the seeding chamber would be to rush into the test cell This avoids the issue of the acetone concentration changing due to expansion as it moves from inside the seeding chamber where the mixture was made to outside where it is measured because of the 66 pressure drop across the flow system (Section 3.1.1). The manual valves at the exit of the seeding chamber and the inlet of the test cell are never kept open together to prevent the flow from the seeding chamber to rush into the test cell. The only time that the valves are kept open together is when the test cell is very close to the pressure upstream or inside the seeding chamber (z 13.2bar). For the calculation shown in the procedure above, the 2 purge-fill cycle has been employed. However if an arbitrary number of such cycles were to be performed, the calculation for ‘n’ such cycles is shown next. Let, [02],, be the concentration of oxygen in mol/l after ‘n’ cycles of vacuuming and purging just described and [00]be concentration of oxygen in mol/l initially at room pressure. The generalized method then based on the procedure described becomes: ' ‘ "[3041— [00] 30 [02 ]n = Where, Vc = Inches of mercury below atmospheric pressure after evacuating the test cell. (29.9” of Hg corresponds to latm) Pf = Pressure ratio or compression ratio of test cell fill compared to atmospheric pressure (typically 12-13.5). 67 And [30 V ] is the dilution factor when the test cell is evacuated to V6 inches of Hg — c absolute. The above equation therefore becomes: _- X71 [304/CT- 3O [0 1,.= [0 x _— 2 210 Pf h d This when corrected for the true oxygen content of the nitrogen gives us: 1— [30-VC]2_ 30 [0 ]n= [0 l x _— 2 2 0 Pf L .. xn + [Pf x2.067x10’7] "10% (Eqn. 3.9) This is the oxygen concentration of the mixture at Pf and holds for ‘n’>1. 68 3.3 RESULTS AND POST EXPERIMENTAL ANALYSIS Having taken into consideration these theoretical and practical aspects, an experiment was conducted to ascertain if MTM has the potential to be used as a tool to measure pressure changes in a gas phase flow. 3.3.1 EXPERIMENTAL DETAILS For the experiment conducted, the fill procedure used was the one described in Section 3.2.2. This results in a starting concentration of 4.2><10—7 mol/l of oxygen and 8.33 x10—4 mol/1 of acetone at 1 bar or 5.56x10—6 mol/l of oxygen and 10.94x10_3 mol/l of acetone at 13.2 bar (this corresponds to the average concentration of acetone due to its vapor pressure at 21- 22°C when mixing with the nitrogen inside the seeding chamber.) Before starting the experiment, the test cell was filled with the desired mixture at 13.2 bar (abs). The mixture was then systematically decompressed by opening the purge valve till the pressure dropped to the desired level. The valve was opened slowly and partially in order that the mixture in the test cell doesn’t cool too much due to expansion. If done to rapidly this could cause the acetone to condense and/or change the temperature of the mixture thus changing the isothermal conditions assumed for the experiment. Readings were taken for five different pressures by dropping the pressure from 13.2bar to lbar in equal steps. Each time the pressure was allowed to drop, at least ten minutes elapsed before the next set of readings were acquired. This ensured that the mixture was back up to room temperature. The pressure was lowered only till 1 bar even though the provision of lowering the pressure still fiirther using the vacuum pump existed. This was because the concentration of acetone dropped too much after vacuuming and the biacetyl build-up 69 due to acetone photolysis was substantial. Readings acquired then were likely to be spurious. The imaging system consisted of the CCD camera and laser described in Section 3.2 and were synchronized using a Stanford Research Systems digital delay generator (model DDGS35). The beam size used was z 8mm in diameter. Around 20% of the total beam’s cross-section was used. The energy-per-pulse of a complete beam was z135mJ. The given mixture was imaged using the following four combinations of the first and second exposure times respectively: (Even though they would essentially provide the same information, the idea of using four different readings for the same mixture is to test the integrity of the theoretical analysis done before and the robustness of the experimental data across the theoretical framework.) 1. 0.5us and 5ps 2. 1.5us and 5ps 3. 0.5us and 3us 4. Ins and lus The DiCAM was operated in the dual-frame mode; i.e. following a single laser pulse, two frames were acquired. The signal ratio was calculated by taking the ratio of the signal counts of the second image to that of the first image. Spatially averaged counts were obtained from a 15x15pixel region (4mm x 4mm in physical space) (see Figure 3.13) in a temporally averaged image obtained by averaging 40 instantaneous images of the mixture taken at 10Hz. A similar average background image was obtained from 40 images taken 70 of the test cell vacuumed and the laser pulsing. This average background was subtracted from the average image before the counts were obtained. The delay period, td which is the time between the end of the first image and the beginning of the second, was 1.5us which takes care of the ‘ghost image’ issue discussed before. The delay of the start of the first image from the laser pulse was 0.24us which is good enough to avoid all fluorescence light. Figure 3. 13: 15x15 pixel region from where counts are obtained in the static cell experiment of MTM. (The region where the spatial standard deviation was the lowest was used to acquire the signal count since the laser intensity variations there were the smallest and effects due to changes in phosphorescence lifetime due to variations in local laser intensity is minimized.) The time for which the mixture was irradiated with the laser for a given pressure, was under 45 seconds which eliminates any errors introduced by photolysis of acetone into biacetyl. 71 See To? mg u an _ E T03 0:208 mo :oumbcooaoo 28 See 73x Nd u :3 _ “m 10H 2235:0280 5&sz .8393 3:085qu 3212 u: .m PEME fit ”was :3 E asthma: mo 83$on :2 v4: ”moumw _mscm _ _ _ “33.580 ——i—————__q ”Sac Ecofitomxm I aim was mam.“ ”mofiw $2625 ”woSQEoU mnm 23 gm; ”mofiw 3525 ”Saw _ScoEtoaxm 1 mam ES mimd ”mofiw 333:3 682580 mim was mimd 63mm 33625 ”8% _ScoEtomxm I mim use mimd ”mofiw Bacon: 6039800 vim 28 9.6.0 ”wowam 12525 “5% _ScoEtoaxm I (as) :opel [eufirg q _ _ _ _ _ _ 3 II. _ m1_nmoumw_mscm H. T [JIIITTII r_+_I_I_I"ILII.. _ _ a _ _ _ _ _ _ _ _ _ _ _ _ J S h F» p 3:58 3:08:on mp...- 72 3.3.2 RESULTS The experimental results obtained, along with the results computed using Eqn.s 3.1, 3.4 and Eqn 3.5 are shown in Figure 3.14. From the results shown in Figure 3.14 it can be seen that there is a clear dependence of the signal ratio on the pressure ratio. In the best case (unequal gates: 0.5 us and Sus) the signal ratio goes up as high as 8.73. Over the 13.2 factor of change in pressure, the maximum change in the signal ratio was z2.6 (1.06 at 13.2bar to 2.73 at lbar for the unequal gate 0.5 us and 3us case) and the lowest was z 1.72 (for the equal gate 1 us case). This clearly shows that, if used properly, unequal gate periods give us the option of not only increasing the sensitivity of MTM but also the option of adjusting its sensitivity. Though the magnitude of the change isn’t as predicted, the trends agree well with the computational model. The difference between the obtained and the computed values was 241%. The error bars shown in Figure 3.14 represent the error in the signal ratio calculated based on the variation of RMS of the variation of the signal counts obtained from the average image compared to the 40 instantaneous images it was obtained from. For a 15x15 pixel window, the variation was i 3.5% and i 7.6% when calculated for a 2x2 pixel window. The errors bars shown on the experimental data in Figure 3.14 and Figure’s 3.15-A through C represents this error. Now that we know that the technique proposed can indeed be used to detect reasonable changes in pressure, it is advisable to go over some of the sources of error and if they serve to explain some of the discrepancy in the magnitude of the computed and experimentally obtained values. 73 3.3.3 POSSIBLE SOURCES OF ERROR The likely dominant sources that could cause errors in the MTM measurement made are: . Error due to variation in initial concentration of oxygen. . Error due to variation in initial concentration of acetone. . Uncertainty in the quenching constants K a and K q . . Other photochemical processes. All the errors discussed in this section will be used to recalculate the possible change in the computed signal response curves compared to the curves obtained based on estimated acetone- oxygen concentrations and values of the quenching constants. Figure3.15-A through Figure 3.15-C. The new curves will be superimposed on the previous ones to show the magnitude of changes we can expect. ERROR DUE VARIATION OF INITIAL OXYGEN CONCENTRATION It is believed that the single largest error would be in ascertaining the starting concentration of oxygen in the test cell. The parameter the system is sensitive to most is variation in the oxygen concentration given the extreme dependence of acetone phosphorescence lifetime on oxygen concentration, particularly so when its concentration relative to acetone in a mixture increases. The static cell filling process though meticulously done, has the potential to introduce errors since it is hard to know very precisely the concentration of oxygen in the test cell before we start the purging process. We assume it is purely air. The efficiency of the purging process of the flow lines is another possible source of error. Since the test cell is filled to a high pressure, the roles of 74 acetone and oxygen as quenchers are equal as compared to a system that is a low speed freely flowing system discharging into the atmosphere where acetone self-quenching dominates. The change in the signal response curve that a i 50% change in oxygen concentration produces is shown in Figure 3.15-A about the original computed curve. ERROR DUE VARIATION OF INITIAL ACE T ONE CONCENTRATION The concentration of acetone in the test cell depends on the temperature of the seeding chamber. Even though the flow is kept as slow as possible, there are likely to be slight variations in the temperature of the seeding chamber. Typically the constant temperature bath maintains bulk changes in temperature. However a slight amount of change in temperature is inevitable. The variation was within i 1°C for the static cell experiment. Corresponding to this change in temperature, the change in acetone’s vapor pressure is z5% (APPENDIX A). The change in the signal response curve for a i 30% change in acetone concentration is shown in Figure 3.15-B. The change is plotted about the original computed curve. 75 .0023 3383800 33830 08 8 3.8.3800 303808008 8038800800 80338 8 000038 .8 0000.800 Raw 0 3 3330800800 003:0 08032 :8me 388800 08 8000.88 008— 3300 332 30 “0&3 8038800800 80308 .30 count?» .880 no 00088 ”0:308 3808398 2.32 firm—d 0.8m:— Ev ”was .89 8 0.8328 .3 08305 MB NB WI 9 4fiquaJ « l _:____ _ _/.,. ___. ll...lr.l|.ll.l..l. ..../ 8035:0083 comma: " _/ 8 03:50 8 03. 00:3 3 _ 3 088—00.. .055 8 853...; __ _ _ / -_._.___ _ __»~___ a: ”38% 35 6088800 _.__.L_._L_____... ”800 38083me I win 30 min; ”008% 33383 6038800 III: 0% 30 01mg ”38% 3383 9.: ”0.02% 35 I <- ”300 38083me I 01m 080 93.0 330w 33383 ”388800 01m 080 mimd ”008% 3383 ”8.30 1.8083me I | ___ ___J__J__J___J_____, | I \O _ _ 01m 30 mind ”003w 3:383 #17 I I +1 II: T I I 7 ”332.5?me _ ______ _ _ 01m 30 956 38% 3383 ”338800 __l.___.__._ l 33 D D ’3D"h 0:308 EJ8083mxm .E 2 (Us) 3011911911318 76 ..xmn 8 880800, 0:: 0000 88 5 .0080 38888 330830 08 8 38888 308808008 830080080 89000 8 000088 8 0000800 $3. 0 8 88890088 0.088 08802 88me 38888 05 800088 08: 3:20 8032 80 0083 80388088 088000 .8 830800, -880 .8 00088 ”38000 8808398 2 H2 um- mfim 0.8m:— An—V "80 83 8 0.588— 8 08000...— 2 «8344 a 1 a . _________ : -..__. 2 m _- 888080280 08.000 0 IIIIIIII 8 098:0 8 0.8 00?...0 082.00.. .085 8 830...!» :51” _ / 017003830033 . 11 t0 _I_I_I I, 808880 _ /_ _ _ _ 0.: 38m 88m ”086 380880me I _ 01m 80 01m; 38m 38083 _ 808980 a. I 01m 80 01mg 38w 380:3 _ ”088 _08083mxm I _ _ _ ———_— V \O (“3) zone: [2113! 01m 80 8nd 38m .0883 8088.80 I I 01m 80 mimd 33m _0883 _ _ ”08w 380880933 OIIIIC _ H 01m 80 01030080308083 _ _ I _ 3 _ _ _ _ 608880 _ _ _ _ _ _ _ _ _ _ _ 03880036520304.9083 TI_III+_+T_I_IT I+IiIIiIlI H0006 _080838m I _ _ _ _ _ _ _ _ _ _ _ _ __ ________ :2 -----I- . A: E II-II I $8008 8808:0me 2 2 77 .823 859:8 3383mm 2: 8 339:8 3038032 ck 98 wk Eon 5 88.85 Ho 338% $3 a 8 wfiwaommotoo maize 3:38“ 3:me c8388 2: 38058 3:: Eton .832 can 5&5 wk 28 wk 3:838 mafia—Baa mo counter $8.8 mo 38:8 ”3252 _Scofitomxo 32.2 "D. mfim 953% P5 ”Ba :3 E 23me mo 2:305 A: 2 S m N T $322.8 ”25:25 _:_I_.Il // E own-Eu 8 26 8?:8 0, , «2.2.8.. 355 E 533:5 m3 ”83w Baum 6239800 9: 68% Edam ”Sat _Scofitomxm Tm. vim was m1 m. # “moumw 133:3 6839800 3 m 28 m1 m. g macaw 3:325 ”Saw _Scofitomxm 1 m1 m 98 m1 238% 3625 ”BBQ—boo mam can m1 238% 13:25 ”8% 3585me I... SE 98 m1 238mm @625 ”BBQEOU man can 9.. Wonmofiw 153:3 ”Sac _ficuegmxm I (US) town Iwfim » bbbl[ D I? o~ 78 UNCERTAINT Y IN VALUES 0F QUENCHING CONSTANT S The value of the quenching constants K a and K q were obtained from A. Lozano [1992] and Groh, [1953]. Though these values are accurate from an order of magnitude of perspective, there are variations in the values of quenching constants reported in different publications. Cebul, et al. [1979] quotes two different values for K q : 6900i 500 l/mol.us and 8200 1/mol.us. This uncertainty comes partly from the inability to make very precise measurements at the time when most of these papers were first published. This in turn results in another source for errors. The change in the signal response curve that a i 50% change in both K a and K q produces is shown in Figure 3.15-C. The change is plotted about the original computed curve. As can be seen, an increase of 50% in the values of K a and K q causes the experimental data to almost coincide with the computed values. ERRORS DUE T 0 UNKNOWNPHOTOCHEMICAL PROCESSES Behaviour of tracers particularly when in the gas phase is subject to many different photochemical processes which tend to alter the emission characteristics of these tracers. One of the most commonly observed phenomena in static cell experiments is photolysis. A tracer in the gas phase is particularly susceptible to this given its low concentration compared to a liquid and the proportionately large photon flux incident on it when illuminated by a laser. In case of acetone it is extremely susceptible to photolysis and the formation of biacetyl. Figure 3.16 shows the case when acetone photolyzes to form biacetyl over time. Figure 3.16 clearly shows that the slow increase in the green color content of emission: typical of biacetyl phosphorescence. Also the greener the image gets the brighter it gets. 79 ACETONE CONCENTRATION: z2.1x10—3mol/l Static cell pressure 2.5bar (t= O) Static cell pressure 2.5bar (t= 5mins) ACETONE CONCENTRATION: z5.63x10’5 mol/l Static cell pressure 0.067 bar (t= 0) Static cell pressure 0.067 bar (t= 2 mins) Static cell pressure 0.067 bar (t= 5 mins) Figure 3.166: Photolysis of acetone to form biacetyl. All parameters were held fixed while the concentration of acetone was varied. (These images were taken with a handheld digital camera only to serve as a qualitative representation of the photolysis phenomena. The camera shutter was kept open for [sec with the laser pulsing at 10Hz) This can be attributed to the higher quantum efficiency of biacetyl phosphorescence. The quantum efficiency of biacetyl phosphorescence is 15% as opposed to 1.8% of acetone [Lozano, et al. [1992]]. The lower the concentration of acetone and longer the irradiation 80 period, more is the biacetyl formed. For our case the concentration was higher and the irradiation times were also much shorter than the period shown in Figure 3.16. Another possibility is the form of Stem-Volmer equation used. Some publications suggest that the self-quenching constant may not be linear for low acetone concentrations [O’Neal, et a1. [1968]], whereas [A. Lozano [1992]] doesn’t talk of this non-linearity explicitly and neither does [Copeland, et al. [1985]]. There is a certain amount of ambiguity in the interpretation of self-quenching and even though this doesn’t broadly affect the results. It might be worth looking into when trying to fine tune MTM as a technique. Some of the quantitative mismatch between the computed and experimental results could be due to the experiments being carried out in a static cell. There maybe certain aspects of a static cell experiment that needs to be taken into account when measuring phosphorescence emission. Most diagnostic related studies using static cells focus on fluorescence rather than phosphorescence hence data is somewhat limited. Certain aspects of the MTM experimental results (attenuation and variation of initial phosphorescence intensity I 0 with pressure) presented that do not impact the use of MTM as a diagnostic technique directly have been included in APPENDD( B. The aim of this thesis to ascertain the feasibility of MTM as a technique has been achieved. The fine tuning of this technique is the next course of action that will be undertaken. 81 CHAPTER 4: MOLECULAR TAGGING THERMOMETRY As discussed in Chapter 1, the phosphorescence lifetime of molecular tracers depends on the local temperature in the flow. Typically, as the temperature goes up, the phosphorescence lifetime drops. The reason for this is a photochemistry issue and has not been included in the scope of this thesis. Details related these photochemical processes can be found in [Sidebottom, et a1. (1979)] and [O’Neal, et al. [1968]]. There were two reasons behind why this study was undertaken: one was to a gain quantitative insight into changes that occur in acetone phosphorescence lifetime with changes in temperature. Understanding how this impacts the use of MTM in non- isotherrnal situations is important. And second is to explore the possibility of extending the use of MTT to gas phase flows with acetone as the tracer. The analysis and results presented for molecular tagging manometry assumed isothermal conditions. However, for most practical compressible flows with large pressure gradients, isothermal conditions never exist. Systems where the rate of change of pressure is high, the temperature changes are proportionately high as well. Typical examples being the temperature changes across a shockwave or the flow ahead of an IC engine piston during the compression or intake stroke. Hence restricting our analysis to purely isothermal flows limits the potential use of MTM in real life engineering problems. It therefore becomes important to determine quantitatively the dependence of acetone’s 82 phosphorescence lifetime on temperature for the range of temperatures and concentrations likely to be encountered in our experiments. It is expected that experiments would show that acetone’s phosphorescence lifetime will depend on temperature. However based on the understanding of acetone’s self- quenching, gained from the previous chapter, it would also be a good opportunity to validate if for different seeding chamber temperatures (which lead to different acetone concentrations) we indeed get different temperature dependence curves, the trends of which are in accordance with the self-quenching hypothesis. Before performing the actual experiments to determine the dependence of acetone’s phosphorescence lifetime on temperature, there are a few aspects from an experimental set-up point of view that need to be addressed. Issues like the variation of acetone seeding concentration with variations in temperature of the seeding chamber, variation in acetone lifetime with local laser intensity and other sources of error that are likely to affect the measurements that will be described next. The amount of error that these variations are likely to introduce and how and to what extent they were mitigated has been described in detail in APPENDICES C and D. 83 4.1 TEMPERATURE DEPENDENCE OF LIFETIME The flow system used was similar to the one shown in Figure CI in APPENDIX C; the difference being a heating apparatus that was added apart from the copper pipe leading out of the seeding chamber previously. Detailed description and working of the individual components of the flow circuit have been discussed in Section CI in APPENDIX C. Afier taking into consideration all of the sources of uncertainty stemming from the variation of acetone seeding concentration with variations in temperature of the seeding chamber, variation in acetone lifetime with local laser intensity and other sources of error that are likely to affect the experiments described next, measurements were made. The temperature range over which the measurements were made was typically 25-160 °C. This was repeated for different fixed seeding chamber temperatures. This would be another test for the self-quenching hypothesis discussed in the previous chapter. 4.1.1 SET-UP The layout of the set-up is shown in Fig 4.1. As mentioned earlier, it is ahnost identical to the set-up shown in Figure C], the exception being the added heating apparatus added to the flow circuit. The details of the heating system used are described next. In order to heat the flow coming out of the seeding chamber, the copper tube attached to the outlet of the seeding chamber is first wrapped with an OMEGA heating tape rated for 600W. It is powered using a KEPCO, BOP-200-2M power supply. This gives a copper tube exit temperature of 60-65°C. 84 To heat the flow further, the end of the copper tube after the heating tape is heated using an industrial hot air blower (heating gun) rated for 14.5 amps and an exit temperature of 520°C. Figure4.2 show constructional details of the heating part of the flow circuit. Acetone seeded I >1 .3 l nitrogen flow \ Heating gun\‘: / Sealed heating chamber ;. sv: Solenoid valve E - / FCV! Flow control valve Heating ”pk-J / Reg.: Pressure regulator E Sd. Cham.: ‘ : Seeding chamber __ Acetone E ==== Coolant 5 m UnseededNitrogen S“““‘“ Re . il l S 3 J l l! FCV .. Nitrogen @ ’ :18; g: 3:33.: '2' 1'11?qu Damper t , . / fl" Constant temperature water bath 7 __‘ I?“ Molecular tagging thermometry flow circuit NESLAB Figure 4. 1: MTT flow circuit (heating added) The heating chamber was made of aluminum and was insulated using glass wool. The maximum exit temperature obtained was 160 to 170 °C . The temperature was measured using a T type Cu Constantan Omega thermocouple with a maximum rated temperature of 260°C . It was placed at the exit of the nozzle before and after the flow was imaged, the temperatures being noted both times for consistency. 85 [SEALED HEATING CHAMBER] / ‘1. COPPER TUBE WRAPPED IN HEATING TAPE AND THEN COVERED WITH INSULATION Figure 4. 1: Details of the heating apparatus added to flow circuit. Care was taken that the thermocouple was in the core of the jet and the laser was blocked when temperature readings were being taken in order to eliminate any heating aspects associated with a focused laser sheet incident on the thermocouple. The thermocouple placed in the seeding chamber bath to monitor the temperature continuously. 4.1.2 DETAILS OF THE EXPERIMENT The kind of measurements made for the following experiment was very similar to the measurements made for the MTM experiment. For a given temperature, a number of instantaneous images of the flow were acquired. From these instantaneous images an average image was obtained. Within an average image a region was selected wherein the local temperature is expected to be uniform and is unaffected by any form of oxygen quenching. The spatially averaged signal counts were obtained from region of the 86 temporally averaged image. For a given temperature, sets of instantaneous images were acquired for 6-7 different delays. The delay this time corresponds to the delay of the frame from the laser pulse (the exposure time wasn’t varied). Based on this, we would get 6-7 average images of the same flow at different delays. From this we could plot a variation of the integrated signal with delay of the laser. A decaying exponential function was fitted to this data, from which the lifetime of phosphorescence could be obtained. 1. The DiCAM PRO was always used in the ‘single image’ mode. It has been shown before that the seeding chamber when used with a constant temperature bath maintains its temperature and therefore the seeding concentration. Therefore the need to use the camera in the ‘dual-frame’ mode is not necessary. 60 instantaneous images were taken for a given delay and the spatially averaged counts were obtained from a small region (typically a 5x5 pixel region where the length of the entire diameter of the jet was around 25—30 pixel) in the core of the jet and within the first diameter of the exit from the temporally averaged image of the 60 images. An average background image was taken before the experiment was started and was subtracted from the average image before the counts were acquired. To obtain the lifetime, all images were normalized to the first image which is an average image of 60 instantaneous images as well (typically taken 0.24 to 0.5 us afier the laser pulse). As shown before, the intensity does not change for the duration of the experiment. Hence normalizing the subsequent images to the first image doesn’t introduce any discernible error. 87 4. The flow pressure was kept constant for an experiment. The nitrogen cylinder pressure regulator setting was z 28-36psig (2-2.5 bar). The gate period was typically 10 to 15 ,us and the delay from the laser ranged from 0.24 to 0.5 as depending on the temperature for which the flow was being imaged. Keeping the first image at least 250ns after the laser pulse ensures that the camera captures only phosphorescence light when the acetone is excited by the laser. Maintaining the 250ns limit also mitigates the effect of shot to shot jitter of theolaser pulse (i 0.1 us for the laser used). 6. The experiment was performed for 3 different seeding chamber temperatures. 4.1.3 EDERIMENTAL PROCEDURE The typical procedure for one of these experiments is as follows: 1. The NESLAB was turned on and set for a temperature between 42-48 °C depending on the desired seeding chamber temperature. A stirrer, placed in the water bath to ensure maximum heat transfer to the seeding chamber and uniform heating, was switched on at the same time. The stable temperature for the seeding chamber, with nitrogen flowing through it, when the NESLAB is set for 42 °C and the flow pressure on the nitrogen cylinder pressure regulator is 36psig (2.5 bar) is 2: 29.5 °C . It should be noted that the temperature of the NESLAB was kept fixed between 42 to 48°C depending on the temperature required in the seeding chamber bath. Further, the flow pressure was typically held between 28-36psig using the 88 10. regulator mounted on the nitrogen cylinder. These settings were held fixed for a complete set of experiments. To heat the flow, the tape is turned on. For temperatures below 65 °C , only the heating tape is used. Beyond that, the hot-air blower (heat gun) is used along with the heating tape. Two temperatures are checked to see if the flow has reached a thermal steady state or not: the seeding chamber temperature and the gas temperature at the outlet of the copper tube. Once steady state was reached for the seeding chamber temperature, i.e. the variation is under 1 °C /min, image capturing commenced. The two aforementioned temperatures and the time were noted before and after images were recorded. Typically, 60 images were taken for 6-8 different delays depending on the amount of light (counts) available. The time taken to obtain readings for one temperature setting is typically 1-3 minutes. The procedure is repeated for subsequent gas temperatures by increasing the voltage setting for the KEPCO power supply supplying power to the heating tape for temperatures upto 65 °C and to obtain temperatures beyond that, by changing the settings knob on the heat gun. 89 4.2 RESULTS The temperature dependence experiment was carried out for three different seeding chamber temperatures (21, 27 and 315°C) and the flow temperature ranged from z25- 160 °C. The idea was to obtain not only data related to the variation of acetone phosphorescence lifetime with temperature, but also test the hypothesis that acetone phosphorescence lifetime changes discemibly with its concentration in the flow. The summary of the results is presented in Figure 4.3. Based on the results obtained, it was observed that there is a clear dependence of acetone phosphorescence lifetime on temperature. Further it was also observed that the hypothesis that the phosphorescence lifetime would change at a given temperature depending on the concentration of acetone in the flow, holds up qualitatively based on the self-quenching hypothesis and published results (see [O’Neal, et al. [1968]]). The temperature of the seeding chamber governs the vapor pressure of acetone which in turn decides the concentration of acetone in the flow. This proves that for the oxygen-acetone composition regimes that were dealt with, acetone self quenching dominates compared to oxygen quenching 90 .253 a 3m u ooh: a A 3 2535 .38 a 3.3 u 002 a A 3 case: .32 a as: u 0.8 a A 3 case: 8888388 $98.20 magnum Bogota .8.“ 2380988 55 0:50.:— ooaoomoconamonm 0:808 no cosmtm> ”mé 0.5m?— Uo E oHBwHomEoH > o: o8 of 2: o2 o2 o: 2: co ow on on ow ow em om o n wcfioaozcfiom no women 256:— Ho 83? fl wouofioa E3052 gonm 3an m 2F .d S d m w. - a / . ,- v m e u 3 flVl m m... m 35-0%.: I Ha» m. N E; don: I -. H o w. Lagoon: all / m. m E; 0.2 Tl... ,/ N. u N 3.5. .003 Elm. $3820 wfiuoom 02mg monafiomfioh m m. fl RE; .002” .T w mix m RE. -ULN mills 0:: 93 H at _ RE. -003 pl.» . rlw m N 35. -ULN ol.o — _ L _ H Ofi 91 VARIATION OF QUANTUM EFFICIENCY OF ACETONE PHOSPHORESCENCE WITH CHANGE IN PRESSURE AND TEMPERATURE -3 5X10 IIIFTIII IIII fiIII T—|_—|_—|-j_‘—'T_T—I--r—-—-l_'_l_—l_—I_T_T_r-r-|-—‘ I——l——l———|-—l—fi—4——-l——l——l————l—-|-—l—-—l-+—4——l——l——|——« I __|__l__|__|_d__l__|__|__l___‘ I I l | | l I I —‘I“I"I”‘I“"‘T‘I“I“I“—‘ L i f? i i f i fix _L__I__ r... EEZOOmmong , I ___—__‘ F—OlSOmmong I | V———-7100mmong —'_-'—" G—ElSOmmong l l ‘ G——625mmong I I PHOSPHORESCENCE QUANTUM EFFICIENCY: ((1)?) O 50 100 150 200 TEMPERATURE IN °C: (T) Figure 4.4: Variation of acetone triplet quantum efficiency with temperature for different acetone concentrations. [O’Nea1, et al. [1968]] Based on Figure 4.3, it is observed that the difference in absolute values for lifetimes are more pronounced at lower gas temperatures and seems to converge at higher temperatures. This is consistent with [O’Nea1, et al. [1968]] which shows the variation of acetone quantum efficiency (which is proportional to lifetime) with temperature for various acetone concentrations (Figure 4.4). Similar to Figure 4.3, Fig 4.4 shows a drop in quantum efficiency as the concentration increases. This is consistent with the acetone self-quenching analysis. Further the curves seem to converge at higher temperatures as observed in our data. 92 __+__. I I l IIIIIIIII‘ rIlllIIIIIA IIIIIIIIIA Error propogation for MTT .llllllllll —+—-I————+-——I—— If the curves obtained in Figure 4.3 are fitted to a higher order polynomial; based on the polynomial obtained we can determine the accuracy expected in being able to predict the measurement. It is in essence a plot of the local slope of the fitted polynomial. This is temperature of an unknown flow for a given inaccuracy in the corresponding lifetime shown below in Figure 4.5. 4.3 ANALYSIS Rtpbvkteb: ”om—8:083 Sum mo 23% 200 0 5 1 ) T ( C O .m m 03 0a 1m e T 0 5 Figure 4. 52: Error propagation in MTT. 93 The ordinate in Figure 4.5 represents the fractional error in temperature per fractional error in lifetime measured. Mathematically, Where 0' T and 0'2- represents the deviation in lifetime measured and temperature inferred. The curves obtained, qualitatively makes sense since the value of ‘y’ is always negative. Physically when the lifetime goes up, it implies that the temperature has gone down. Hence a positive error in lifetime measurement leads to a negative error in temperature measurement and vice versa. From the plots, the percentage error in the temperature measured in smaller compared to the corresponding error in lifetime measurements for temperatures below 60°C. Apart from this there are certain other potential sources of error that need to be discussed as well. For example, the concentration of acetone in our case is calculated inside the seeding chamber. Since in our case the flow velocities are so low (<3m/s), and the total loss coefficient for the section of the flow circuit from the seeding chamber to the exit of the nozzle is around 6-10 (based on the sum of loss coefficients due to a sudden contraction, sudden expansion, loss around four gradual bends and other minor flow losses), this means that the pressure drop from the inside of the seeding chamber to the exit is only around lO-25Pa (this can be calculated very easily using Bemoulli’s equation). Compared to 105 Pa (atm pressure), this number is very small. This was 94 measured physically due by putting a pressure gage to measure the pressure inside the seeding chamber which showed no detectable gage pressure. Therefore the difference in pressure at the exit of the seeding chamber is not very different from atmospheric pressure. This in turn implies that the concentration of acetone inside the seeding chamber is not very different from the concentration of the acetone in the stream exiting the nozzle. In case of very high flow speeds, the pressure drop needs to be accounted for and the drop in concentration will be inversely proportional to the ratio of the pressure inside and outside the seeding chamber. Assuming that the flow speeds are small, the limit for this explanation to hold is the point when the acetone partial pressure inside the seeding chamber is lbar (at 22°C, the partial pressure is around 0.3 bar).This coincides with a seeding chamber temperature of 47°C. At this point, all the nitrogen will be replaced by acetone vapor and the flow will be driven by the vapor pressure of acetone itself. Increasing the temperature of the seeding chamber will speed up the flow. In the case where the acetone seeded flow exits the jet slowly, but at a temperature higher than that inside the seeding chamber, the flow would actually expand as it comes out (ideal gas law). The density of the exiting flow would be lower than that inside the seeding chamber. Higher the exit temperature, lower is the density. 95 Based on the Ideal Gas Law one can say: P56 = psc x 9? x T SC (SC =seeding chamber) and Pexit = pexit x 93 x Text! ' Given that Pseedmg chamber z Pm, , we can say that: psc T exit pexit Tsc This in turn would imply that there is a reduction in the concentration of acetone in the exiting jet by a factor proportional to the absolute temperature ratio. Since lifetime depends on local acetone concentration, the extent of self-quenching would go down and this in turn would increase the lifetime at elevated temperatures as compared to the lifetime one would expect, if the flow did not expand due to heating, at that temperature. This effect is a lot more pronounced at higher temperatures (the expansion factor is 1.57 for 170°C) than for lower temperatures (the expansion factor is 1.07 for 40°C). From all of these discussions we can conclude that MTT for gas flows based on acetone’s phosphorescence lifetime is a non-intrusive tool that has significant merit and with some more refinement can be used effectively to make temperature measurements in gas flows. 96 CHAPTER 5 FUTURE DEVELOPMENTS The objective of this thesis was to determine the feasibility of using MTM as a non- intrusive pressure measurement technique; which has been accomplished. The next logical step is to improve its accuracy and usability as a practical diagnostic tool. This chapter deals with some of these aspects and possible methods by which MTM as a technique can be improved and used in conjunction with MTT to make simultaneous pressure and temperature measurements in gas flows. 5.1 PROVISION FOR MAKING MIXTURES OF DIFFERENT COMPOSITIONS One of the limitations faced was the manner in which the mixture used in the static cell was made. It gave very little flexibility in changing the mixture composition apart from changing the number of purge cycles. The biggest advantage this would provide is the ability to adjust the position of the response curve into a range of pressures that a specific problem would encounter. One of the first steps would be to develop a technique for making very precise mixtures of a desired composition. One way this can be done is to fill two tanks connected to each other with nitrogen and acetone (see APPENDIX E and Figure E.2). Of the two tanks one is small and the other big. The option of disconnecting the two tanks from one another should be available. Once the two tanks are filled to the desired pressure and have equalized, they should be disconnected from each other. Then a high sensitivity differential pressure transducer the ends of which are connected to each of the two tanks will be turned on. At this point it should show zero differential pressure. Then pure oxygen will be injected into the larger one. The differential pressure now 97 indicated by the pressure transducer will reflect the partial pressure of the oxygen added to the larger tank and can be used to calculate the concentration of oxygen in the mixture. This in theory should provide substantial flexibility in being able to measure precisely the concentration of oxygen in the larger tank. A more detailed description has been included in APPENDIX E. The different components of the system are already present in the laboratory where this research was conducted and testing this technique will be the next course of action. 5.2 ADJUSTABLE SENSITIVITY OF THE SIGNAL RESPONSE BY VARYING INTER-FRAME DELAY The analysis shown in Chapter 3 (Section 3.1.2) talked about trying to adjust the sensitivity of the signal response curves by varying parameters such as the gate period(s), gating schemes or mixture composition (see Eqn. 3.4). Another parameter that can be varied and should be considered is the delay ‘ td ’. The variation in signal response with pressure is illustrated in Figure. 5.1 and Figure. 5.2. Varying td does two things: the first is that it moves around the position of the signal response curve with the different pressure ranges (as is obvious from Figure 5.1), and second is it varies the dynamic range of the signal response curve (the log version of Figure 5.1 shown in Figure 5.2 shows this). This gives another control parameter to optimize the experiment. In fact the delay is one of the most effective methods of adjusting the sensitivity of the signal response curve if a particular pressure range is of importance. For given gate periods, by plotting the variation of the signal response with delay for a particular pressure of a mixture, the delay which produces the required curve can be obtained. 98 62:; wins 52% ”Bambi 5:5 own—0&8 3:me .3 c2398» 3 .m .cuauE Ev “was :3 E 83?? mo 238.5 2 .2 .2 .-S II_,I..._ . :Hflflflflfi “ :fififiwflq H _ ....... ..-..-H-” ..--..s.“ ..... m. ........ ..... U. ........ .... ...... m ...... ..... m ......... III-I . Ih-TL-L--.--+---.----_ ......... ..-.-T-T-.--L---L ..... T--------.,T.-.-.-L--L---n-----. ......... m.~ 1-“--u--n--r-r--_. ..... u ........ _3_-_.-_ -..-..--..:-.. ..... u ......... #121: :8. ESQIIN ”u u n H u n u I” n _ n n u u H: u u v- H H No. __ . _ _ _ _ _ ___ _ . _ . . _ ___ _ _ “mafia“ HQ“ 0 -._-_r-u:n-4:#:_. ..... u ....... -_.._-..-”--_.-_ -..:L. ..... n .......... 1-4-..-..-- qegeas.-m~ m ..... ..... ...... .. m u” n n u u n u I" H u u u _ u ”I u u . I m. L-"--u--n--T-T--H. ..... u ......... Seawozzwftnm ..-., II"_" n 11".“ n _ "In” 39:12 am I._-.I-”-"-L-L-I_. ..... n ......... ....-.w-ruwifii- .. IIIII u ..... ---..-"-."-_.L- coup—“80:00.1? ”u _ _ _ . _ _ ... _ _ _ H _ " ___ u " combo I. 35"..“ .3mt§§~”mogmwfiscoaDILH-I.” ----- .... ..... .. ......... mé I-n-“mlmgufl .mnmcgmifi”mogmw_m=uo:DI..H-l-u ..... ”I-------..-”-"-".L-u--”---“-----“. --------- m : ...-......s.........ml m m m x: m m m m m ... H ” W1m.~"vaficma~”mouwm_mswmI H u H W“ n u u H u u n H. H n n :I.” H u n H e E2 63...; fl ca >23. 3 9:53.:— 53 8.896.. 3:3... he E5355 99 .83 w2 ”85> mfion >220 ”Samoa 5:5 398%“: Rama no countm> "fin .95»?— Ev ”was :3. E “:3me mo 233.5 A H H u u u T... H M " X H H :H H “ H M fl “ IIII_II_IIHIIHIII_I IIIII .IIIIIIIIII.I_II_I_III.II.III_IIII_IIIIII_ IIIIIIIII _II_I_II_II_III_III_IIII. IIIII .I IIIIIIII I IIIUIIHIIHIIHIIIP lllll . IIIIIIIII kl.lLlrl-.—ll_lll"llll_ IIIII I“ IIIIIIIII ”IL—IFII_IIF.ILIILIIIL_ lllll _F lllllllll _ _ _ . _ . ___ . . _ _ _ _ ___ _ -”HUHUHHHHHHHHHHHHHHHmHHHHHH“Hwy”MHHHHHHmHHmHHHmHHHHmHHHHHHHHHHHummmumflWHHWHHHWHHHHHWHHHHHHHHH l|I_-II_II+.II.».|II>. IIIII _ lllllllll FILILIfILllrll—IIIIl. lllll L lllllllll rLlr|.||rlLllLIItL lllll T IIIIIIII I HHUHUHHHHHHHHHHHHHHUHHHHHHHHHHUHHHHHHHUHHHHHHHfl-I-HHHHHH-I.HHHHHHHHHflHflHflHuHHUHHHUHHHHHHHHHHHHHHH HHH_HH.HHH.IIHHHHHNHHHHH.“HHHH-MHHHHUHHHHHan-LHHHHHHH_II.H.IIHHUHHHHHHHHHDHHHHHHHHHHHUHHHUHHHHHHHHHHHHHHII. _ _ _ _ . ___ _ _ _ _ _ _ _ _ . _ _ U a . _._ _ _ _ _ _ _ ._--“--"--...-J--.T-- maul”. .0126me .mouawfiscocb --.I_-_T”-_T._---.--l_ ..... .- ........ - WM HHHHHHHHH _u HHHHHHH---.---..---.b.------..H VIM-...-HHHHHHHHHH mimén a .mimwsm 9: ”mouawdwswoca-I “NEHHHH H Ik|_l l_ll.rllhlllrllll IIrILlrl_l|—llt I1 mmmm-hflmmmmmmmmmmmm u . mmmmmmmmmmm“ . WI” ” u “ mix-H chmafibouawisva " n u u " SoEYSxmmw-I- __ _ ___ _ . .I_II_ IIIIIIII_IIIII . U . II.II.I.II_II_II- anfiummuomH-I I ”NH 3:” “333.833“:ch HUHHHH- sougfioaoon -“ . --------------- “I-“ M “Iain---. ----- u --------- H w ocouoo22. 3 «Sumo-E .53 92:93.. :3me .«e nemesis-tr 3 0mm Iwfits 6‘s) 100 5.3 DUAL TRACER BASED SYSTEM FOR COMBINED THERMOMETRY AND MANOMETRY MEASUREMENTS Chapters 3 and 4 mainly dealt with the pressure measurement and thermometry aspects independent of each other. However for realistic compressible flow problems, pressure and temperature do not remain independent of one another. Hence part of the effort in developing MTM will be to devise a method by which both can be measured together. Recalling Eqn 3.3 we have, —1 r=[;1;+cf[1<10—3 mol/l: since the concentration of acetone at 13.2 bar 13.2 was 10.94x10’3 mol/l. 116 Based on this we obtain the factor by which the laser beam attenuates from the entry to the imaging window. This is summarized in Table B.l 13.2 bar “'16 7.7 bar 4.4 bar 1 bar bar Factor by which laser beam energy drops 4.24 3.39 2.32 1.62 1.12 over 135mm. Table B.1: Drop in laser power in static cell due to attenuation. Figure B.1: Absolute location of imaging region location with respect to the test cell entry window. 117 I I I I l I I I I I I I WI I T I I T I I I rT I I I I- N 6 -— 'I E .. _ 15 ° 9 A _ - o 8 .- E is: I 3 5 ' - E b 4 I. ._ 10 £5 a - O .. ‘° .2 - t:- 8 - " £3 7‘ - " é a .1 8 2 - o t- - 5 8 U P I I: b S O: k/ -I I: I: o e - 1 05 E O llllIJlllillJleLlJLllll_ 0 a ‘3 20.0- 225 250 275 300 O 325 350 Wavelength in nm Figure 2. 6: Acetone absorption spectra. [Lozano, [1992]] Figure B.2 : Extinction co-efficient of acetone as a function of excitation wavelength. [A. Lozano [1992]] Initial intensity obtained fiom the imaging window based on experimental data and Eqn. 1.1, now corrected for attenuation is summarized in Table B.2 and plotted in FigureB.3. Since the overlap of acetone’s absorption and emission spectra is small, the absorption of acetone’s emission by itself is not significant and has been ignored. There is an obvious discrepancy in the expected and observed values. A possible explanation is the fact that the lifetimes used to calculate the initial intensity 10 were 118 13.2 bar 11.16 bar 7.7 bar 4.4 bar 1 bar Initial intensity calculated at imaging window before accounting for attenuation. 256 271.5 235.7 206.6 105.3 Initial intensity calculated at imaging window after accounting for attenuation. 1085.5 920.5 547 334.5 118 Measured ratio of corrected initial intensity at any pressure with respect to that at 1 bar. 9.2 7.8 4.64 2.84 Expected ratio of initial intensity at any pressure with respect to that at 1 bar. 13.2 11.16 7.7 4.4 Table B.2: Expected and calculated variation in initial intensity with attenuation taken into account. improve the experimental data obtained. 119 The ideal situation would be if one could obtain a larger number of signal ratios at a given pressure by changing the delay (td ) between the image pair. The signal ratios then obtained if fitted to a single exponential using a least squares fit, would give a more accurate lifetime Figure However that would mean that the mixture at that pressure would be irradiated by the beam longer leading to photo-dissociation of the acetone introducing another error. This can be minimized by placing a magnetic stirrer inside the test cell. This is one of the improvements that can be made to the test cell and would test cell. This is one of the improvements that can be made to the test cell and would improve the experimental data obtained. Variation of initial intensity with pressure : MT M experimental results 15 I I I I T r r Y f | I I I l l l l L - ._ B—Et Observed values of IO ” e—e Expected values of IO ‘ “ »—*=—_du-—bfl_—I___—a-—___ >._.___________._______._ p—I O _-- LII Inrtral Intensrty normalised to initial intensity at 1 bar: (IO@P/I@ lbar) Pressure in bar abs: (P) FigureB.3: Variation of initial intensity with pressure: MTM experimental results. Another more important source of error is the calculation of the magnitude of attenuation itself. The difference in the expected and observed values is z 30-35%. N ent ry - 8X1 x C If one refers to Eqn.B.4, ___-=10 ( ) N image and modifies it as shown below, 120 M :10_(EXIXC)Observed N- (Eqn.B.6) lmag e Observed then taking the ratio of the above two equations (assuming a 30% difference in the expected and observed values) we get: 10‘“ x 1" C)Expected 1.3 = 10_(gxIXC)0bserved 2) 10gm [L3]: (5 x I X C)0bserved —(5 x l x C)Expected = 0114 (8 x l x C) Expected = 0.627 for the 13.2 bar case. This means that an approximately 18% change in the value of (5x1 x C) Expected would account for this difference. There are variations of upto 20% in the reported values of the extinction co-efficient from one publication to another which adds to the error. Further Beer’s Law that was used to calculate the attenuation of the laser beam also needs to be adapted for high intensity of incident radiation. Most of the Figures quoted for e are for low intensity excitation. Recent work [Klingbeil, et al. [2006]] has shown that the absorption cross-section of certain hydrocarbons with nitrogen as a buffer gas has a dependence on pressure. Since we are dealing with a static cell in which the pressure varies, this could be a reason why there is a discrepancy in the expected and observed attenuation values. 121 APPENDIX C 122 APPENDIX C: PRE-EXPERIMENTAL CHECKS: SEEDING CHAMBER TEMPERATURE. Before the actual MTT experiments were carried out, some experimental aspects likely to introduce errors in the measurements made needed to be understood and quantified. The largest one among these was the variation in the seeding chamber temperature; particularly due to its cooling when nitrogen was bubbled was flowing through it and acetone was evaporating. Before embarking on why this was important, knowing some of the details of the set-up used will be beneficial when trying to explain the finer details. C.1 SET-UP The following block diagram (Figure C.1) will shows the layout of the test system and the flow circuit used to produce the stream of acetone seeded nitrogen. As shown, the nitrogen flows out from a standard high purity nitrogen cylinder (AIRGAS, high purity nitrogen, <5ppm of oxygen) through a dual stage pressure regulator (Reg) which is the first flow rate control device. This provides the upper limit and bulk control of the flow rate. The flow then proceeds to a flow control valve (F C V) which provides finer control of the flow rate. From there the nitrogen moves to an on-off type solenoid valve (SC). The outlet of the solenoid valve leads to a damper, which is essentially a tank with a large enough volume to smoothen out small scale pressure fluctuations and helps maintain a stable flow rate. The outlet of the damper leads into the seeding chamber containing the acetone. Inside, the nitrogen is directed to the bottom of the seeding chamber and bubbles through the acetone, coming out from another outlet which then flows out from a copper pipe (7.5mm ID, 1mm wall thickness and 2m in length). 123 The tracer in the flow is excited using a 20ns pulse of a 308nm, XeCl excimer laser (Lambda Physik LPX 200) and imaged using an intensified CCD camera (12bit, P46, DiCAM Pro, Cooke Corp.). Typical energy level per pulse was 125-145mJ. All temperatures were measured using a T type Cu Constantan Omega thermocouple with a maximum rated temperature of 260°C .All measurements (temperature and lifetime) were made at the core of the jet and within the first diameter. SV : Solenoid valve F CC! Flow control circuit Reg.: Pressure regulator Sd. Cham.: Seeding chamber Acetone seeded L nitrogen flow — Acetone -—- Coolant “UnseededNitrogen Nitrogen Reg. I | FCC Damper '5... Molecular tagging thermometry flow circuit Constant temperature water bath I l I f": : 3.2:: ::‘_‘..".‘_:;; f;;‘;,v NESLAB Figure C. 3: MTT flow circuit 124 C.2 MOTIVATION As described above, the acetone mixes into the nitrogen stream as nitrogen bubbles through it. As nitrogen carries away acetone vapor, the acetone vapor removed gets replaced by liquid acetone which evaporates in order for acetone to maintain its vapor pressure (corresponding to the temperature in the seeding chamber at that that time) in the seeding chamber. This process leads to lowering of the temperature inside the seeding chamber as heat/energy is lost in evaporative cooling. This becomes important while making measurements since changes in the temperature of the seeding chamber leads to changes in the vapor pressure of the acetone which then leads to lowering of the concentration of acetone in the flow since its concentration in the flow depends on the vapor pressure. Non-uniform acetone concentration in the flow affects the lifetimes measured at the outlet since the lifetime is dependent on acetone concentration based on the self-quenching analysis spoken of earlier. This greatly impacts the accuracy of our results, the magnitude of which will be shown subsequently. It is therefore imperative that the seeding chamber bath temperature is held fixed (at least to some reasonable degree while the experiments are being conducted) in order to minimize errors due to varying acetone concentration. The following procedure was used to maintain a stable seeding chamber bath temperature. C.3 PROCEDURE 1. In order to stabilize the temperature inside the seeding chamber, the seeding chamber was immersed in a constant temperature bath filled with water (see Figure Cl). 125 . This bath was typically maintained 5°C above the temperature desired inside the seeding chamber. The bath itself had copper coils immersed in it with ethylene glycol solution circulating through them. The flow rate and the temperature of the ethylene glycol were maintained very precisely by a constant temperature heating and cooling system (NESLAB) rated at 600W. . With all systems running, it would take typically 3-5 mins before all temperatures (inside seeding chamber, constant temperature bath and nozzle outlet.) stabilized so the temperature variations would be minimal. . This was when measurements were made and the time required to take measurements was typically 40secs to 1 min. . The readings were always taken at around 3-4 minutes after flow was started and lasted a maximum of 1 minute. . For lifetime measurements, 6-7 delays were needed in order to get enough points for a good fit. In this case the time required was around 3 minutes. . Figure C.2 clearly shows that the flow system was capable of keeping the seeding chamber temperature fairly constant during this time. The following graphs will illustrate the effectiveness of the system just described in maintaining a fixed temperature. C.4: RESULTS Figure C.2 shows the normalized intensity ratio of a pair of phosphorescence images taken at a time‘t’ after a laser pulse and how it changes in time while nitrogen was flowing through the seeding chamber. The DiCAM Pro was operated in the dual frame 126 mode and the image intensity ratios were normalized with respect to the phosphorescence intensity ratio of the first image pair that was acquired. The two curves shown in figure C.2 represent this variation before and after the seeding chamber bath was put in. Figure C.3 shows the changes in temperature at the tubing outlet and the water bath before and afier the temperature bath was put in. Drop in intensity value of first frame over the duration of the experiment ‘ 5 p—. O O .9 \O 00 J .0 \o o 0.94 - 0.92 I 0 Case]: Intensity of first frame with constant temperature bath in place. L'J Case2: Intensity of first frame with no constant temperature bath. m - 0 100 200 300 400 Time in experiment after the first image was taken in seconds: (t) Intensrty of an average image taken at 't', normalised to that of the first average image taken: (I 0/It) L .0 o o Figure C.4: Variation in intensity of image before and afier the constant temperature bath was put in. Nozzle implies gas exiting the copper pipe, i.e. temperature at the core of the jet. Casel: Change in temperature for the nozzle: 23.7 °C to 23.7 °C and 23.6 °C to 23.4 °C for the Seeding Chamber. Case2: Change in temperature: 20.5 0C to 17.6 °C and for the nozzle 16.6 °C to 11.1 °C for the Seeding Chamber. It is obvious that for the time needed to make measurements, this system is fairly good in keeping the seeding chamber temperature fixed. 127 Temperature versus time 0 Seeding chamber bath temperature: Without constant temperature bath. V Nozzle temperature: Without constant temperature bath El Seeding chamber bath temperature: 25 if: With constant temperature bath 5.120 ’ O Nozzle temperature: 9 With constant temperature bath .5 23. El 1 53 D 5‘ £15 ' ‘ lO“““ ‘4‘ ”Mr O 100 200 300 400 500 Time in experiment after the first image was taken in seconds: (t) 4 A 4* L Figure C. 5: Variation in temperature with time for nozzle and seeding chamber bath before and after the constant temperature bath was put in. Even though the flow system was designed to maintain a constant temperature in the seeding chamber, as shown in Figure C.4, there was still some change in the seeding chamber temperature over the duration of the experiment. It therefore becomes important to quantify the error introduced due to this. While data was being acquired, the temperature changes for the bath, inside the seeding chamber were fixed to within i 1-2°C (corresponding to a 4-8% change in concentration and comparable changes in lifetime). If these conditions were not met, the reading was 128 taken again. The following calculations (similar to ones shown in the MTM analysis) will show the variation in lifetimes obtained due to this. Assuming the following data for the calculations: 2'0 =200,us Kq =80001/mol.,us Ka =121/mol.,us [act] = Concentration of acetone at 22°C. : l l x 10—3 mol / l. [Lozano, et al.(992)] [02 ] = Concentration of oxygen from nitrogen cylinder at N T P conditions = 2.067 x 10_7 mol / I. So based on the Stern-Volmer relation we have: —-1 r=[—l—+[Kq[02]+ Ka[act]]:| w 70 For the above data we have: 2' = 7.25 ,us. If we were to change the seeding chamber temperature by 321°C, we get approximately i 4w 5% change in concentration levels which corresponds to a z i 5% change in lifetime. Table C.1 shows the changes. 0.95x[act] 1.0 ><[act] 1.05 ><[act] 7.61 ,us 7.25 as 6.92m Table C.l: Example of variations in lifetime corresponding to changes in seeding chamber temperatures. Concentration of acetone based on its vapor pressure at 22°C at room pressure. The analysis just shown clearly demonstrates that variations in the seeding chamber concentration can introduce substantial error and needs to be dealt with suitably before an experiment of this nature can be conducted. APPENDIX A shows the typical changes one 129 can expect in terms of baseline lifetimes for different seeding chamber temperatures/acetone concentrations(and an oxygen free environment likely) to be used for experiments of this nature. It should be noted though that the temperature of the bath will always be higher than the temperature inside the seeding chamber. Readings were taken for a fixed nitrogen cylinder regulator setting which fixed the flowrate. This setting was maintained for all experiments since changes in nitrogen flowrate would change the cooling rate. The following (Table C.2 and Figure C.4) show data taken during a 7 minute run during which temperature readings were taken both inside the seeding chamber and of the constant temperature bath which shows that the temperature inside the seeding chamber is lower than the bath (as expected) and that it drops faster than the temperature of the bath. Constant temperature Seeding chamber Time elapsed after bath temperature: (°C) temperature(inside): (°C) flow was started: (mins) 31.5 29.7 0 31.4 27.7 1 31.3 26.6 2 31 25.9 3 30.8 25.8 4 30.6 25.5 5 30.4 25.4 6 30.2 25.] 7 Table C.2: Temperature variations inside the seeding chamber corresponding to changes in temperature of the bath. 130 Temperature vs. time: Seeding chamber and bath temperatures v v I I I f I I 7 1 i I L _ “I _ ‘1 — —:_ G—-0 Temperature ofthe bath. _ ._ —— 1 — — ,— [3——-El Temperature inside seeding chamber. I— —- —— — — —- —- — — — — —— — — — -— — -— — — — -— — —— — — — —-1 Temperature in °C: (T) _— _ _ _ _ ._ _ ._ ___. _ _. _. _. _ — _ ._ —1 25 __|_.r.____ Time elapsed after the flow was started in minutes: (t) Figure C.6: Temperature change inside seeding chamber over time compared to changes in bath temperature. Keeping this in mind, for all data reported in this thesis, the seeding chamber was recorded and used to determine the acetone concentration. 131 APPENDIX D 132 APPENDIX D: D] CHANGES IN PHOSPHORESCENCE LIFETIME MEASUREMENTS DUE TO VARIATIONS IN LASER INTENSITY An aspect of the experimental setup that has the potential to introduce errors is the local laser intensity. The typical issue one would face is due to fluctuations in the laser power, laser jitter or its non-uniform cross-section. Since all data was obtained from an average of 40-60 instantaneous images, most of these errors can be neglected. The following (Table D.1) will indicate the order of magnitude of each. Parameter Variation Error introduced in the intensity of an instantaneous image 1. Variation in laser energy Typically i4mJ for a 2.85% fiom pulse-to-pulse mean value of z l40mJ/pulse 2. Laser jitter. i0.1us about mean This error can be calculated using: d] dt — = — — .For a lifetime of I r lOus, error is around 1% and 10% for a lus lifetime Table D.1: Sources of error in an average image due to variations in the laser and their respective magnitudes. It should be noted that these errors go down by a factor of «[6—0 for an average image since the average image is obtained by averaging 60 instantaneous images. The reason why the number of images was restricted to 60 was the amount of time that it would take to acquire and store the additional images during would result in larger changes in the flow conditions and thus introduce another error while trying to minimize one. Using a camera in a dual flame mode and taking a ratio of the 2 images accounts also for non-unifonnities across the laser beam or non-uniformities introduced by the optics 133 (lenses, mirrors, etc.). However one needs to be careful that the amount of signal or light integrated in both frames is sufficiently high to minimize effects of noise by maintaining a sufficiently high signal to noise ratio. It is obvious that using a large number of images and averaging them helps to almost eliminate most of these issues. However a more important issue arises when considering the dependence of lifetime on local laser intensity. It has been shown [Kaskan, et al. 1949] that the changes are observed in acetones’ phosphorescence lifetime with changes in the local laser intensity. Figure D.l illustrates the changes observed. ~——~— (D V O) 1 - -1 —3 /1_1nsec x10 0! Lwl JI, o ‘ ‘50 :00 ISO 1 in arbitrary units. A Figure D.l: Change in lifetime with change in local laser intensity. [Kaskan, et al., 1949] In order to verify this effect and get an idea of the effect it has on our measurements, the following experiments were done. 134 D.2 SET-UP AND PROCEDURE. The set-up used is exactly the same as shown in Figure C.l. The laser intensity was changed independent of the pulsing rate or the duration of the pulse (20ns). The flow system was run in a manner so as to keep the temperatures at the outlet and seeding bath constant and the flow was imaged based on which the lifetime would be calculated. The only parameter varied was the laser intensity. The laser power wasn’t focused into a beam but rather a sheet in order to prevent any photolysis driven behaviour though this effect isn’t typical of a flowing system. The lifetime was calculated using a section of the illuminated area, typically a 5x5 pixel window, within the first diameter of the jet where one diameter of the jet occupies 2530 pixels. D.3 RESULTS Based on the lifetimes measured, a very minor change in lifetime was observed (z2% drops in lifetime when the local laser intensity increases by for a factor of 2). Figures D.2 and D3 shows the observed changes. Figure D.2 shows the normalized exponential decay curves. These data shown on these curves were fitted to an exponentially decaying function and used to obtain the phosphorescence lifetime. The lifetime obtained from these curves are shown in Figure D3. 135 NORMALISED INTENSITY vs DELAY 1.0 I I I I I T I I I I I I | I I I I I | I I I I I I—-— |___I__I___|___1__I___I___I__I _____ I____I_._I.__I___4 I I I 1 1 1 i | I | I I I -LASERINTENSITY:120mJ I I I I ”7:08 I I I . LASERINTENSITY2160mJ I I I I 39 . , I - LASER INTENSTIY: 195m] , . . . 5: .--. -4-.-“ oLASERINTENSITY:227mJ.—————I——+——I———I——« I" I I I . . . I I I I a I I I | I I I I I I I I 20.6 I l I I I I I I I I U [E]. I I | I I I I I I I I | I I I I | I I I I I I E “‘1“? r‘1“—“"T“r‘7“r “““ I‘—T‘“r‘1“‘ a I I I I I | I I I I I (004 l l l L 1 1’1 1 l l l l S ' I I I I I I I I I I I <5 I I I I I I I I I I I I ~——I——--+——I—--I~——— —I———I—-+--——-I—-+——I———I-—--< g I I I I I I I I l I I I I I | I I I I | I 20.2 I I I I I F I T I I I | I I I I I | | I I | I I I I I I | l I I P.-_I__—I“—r—_l—-—I‘_I_—_I‘__I—_I ————— I—_—I—_I_——l__ I | I | I | I I I I I I 0 l l l l l J. l l l l l l 0 10 20 30 DELAY IN us AFTER LASER: (t d) FigureD.2: Signal ratio versus delay: exponential decay curves. As expected, the changes were small and for almost all of the work that will be done, the region in the flow where the lifetime was calculated, remained fixed thus mitigating this problem even more. Also the center of the laser beam used has a flat topped profile. Hence when a sheet is generated, the central portion of the sheet is almost invariant in terms of its power. The non-uniformity comes into play if one uses the edges of the sheet which was never done here. Typically the laser sheet was adjusted so the region in the flow best illuminated coincided with the first diameter of the jet and was used to obtain the lifetime since that also helped us make best use of the dynamic range of the detection devices. 136 8 LIFETIME vs LASER INTENSITY .___I___b’ I _____..I___.I___l___l.___.__L__L__I_-_|___. L___I___I___I___I ____ ‘3 _____ r __J-____-.__:__LT9_L__'___I I I I I I I I I I I I I >———I———I——-I-——I —————— I--—I-—-t——-t--'I’——+——t‘—-‘t-——I——— r__I___I___I_._I ______ I___l__.l___l ______ L__L__I___I___ I I I I I I I I I I I I 7 I I I I I I I I I I I I A -—I---I-—-I--—I ------ I---I---I——-r--~I-—r--r——I---I--- 3; [__I.__I_-_I__-I ______ I___I___l.__._l ______ L__L__I__..L__._. En' ,_ I I I I I I I I I I I I I 1 T’F—W-‘W—"W ______ I___I_—‘I___I__”__T—_I__—I_—'I_—_ E ”—-I_"I"—‘I““‘I —————— I““I""I"_'T“I‘_T-_T'—-I‘——I'"‘_ 6 I r 1 If I r l r I l r l E ,__'___'___'___I ______ '___'__.'___'__-.--L__'___L__'___ I: h-_I__l_-_I___I ______ I__I__I_-I__‘__I_-I__I__I__ IL] I— I I I I ‘I 'I 7 I T I‘ I— I" a >~—-—I———I———|———I —————— I———I———I—-—I——d———I———I—-—-I—-——I———-I ...] >___I___I____I___I ______ I___I____I___I.__d___l___l___I___I___‘ I I I I I I I I I I I I 5 I I I I I I I I I f I I ~——I——-I——-I———I——————I———I———I——+—— --+-——+---I-——I-—— .___I_--|___I___|___,__._|___J___|___I-_-t_.-l___L-_L__|__-. I I I I I I I I I I I I ’""F_—I_’_I___I _____ I"‘7—‘7—‘T’_""T"_T"_I‘_—I'-_‘ r—-I———I—-—I———I-'-----I-—-I-—"I‘--t——-I*--+-—I---t--—I——— 4 J l l l l l l l l l l l 100 150 200 250 LOCAL LASER INTENSITY in m]: Figure D.3: Variations in lifetime with variations in local laser intensity. D.4 ADDITIONAL SOURCES OF ERROR FOR MTT Apart from the aforementioned causes, there are certain other experimental parameters which also affect the accuracy of our measurements. These are discussed below. ERRORS CONNECTED T 0 T EMPERA T URE MEASURING DEVICES. Inaccuracy connected with the thermocouple is also something that needs to be accounted for. The T-Type copper constantan thermocouple used typically has at least a i 1 °C error in reading [OMEGA engineering: source for thermocouple wires]. This was confirmed using a more accurate RTD sensor. 137 LIFETIME vs LASER INTENSITY 8 f . f , . . . . ,___L_.___E_' |....___|___I___I__._l-_.4.__L__L__L___.|__.__ I I I I fl? I I I I I I "“I“"I‘“’I““I _____ I‘__I’—-I‘——I‘_‘"’“T“#F‘“I‘”"I I‘“-'I-“—-I——-‘I———I —————— I—-—I—-—-I——+———«I——+——I———I———~I—-—- ,___I___I____I___I ______ I__.J___I__I__,__-L__._L__L__.L___ I I I I I I I I I I I I 7 I I I I I I I I I I I I A "“_I"‘-I“——I-“‘I ------ I““‘"I“—‘I-“'I“*"T“‘t““t“‘—I—*‘_‘ 3: .___I_._._I___I__._I ______ I_.__I__u__4_._..__t__r__._.I___I___. £13 I I I I I I I I I I I I 1 F—flI—“_Im——I—_—I ””””” I““’I"“I"T“”"“_T‘_T_‘F_“I"“"‘ E I'“_I__‘I’""I___I__‘—"I“‘_“I‘_"T’*'T—“ “*r——r-—r‘—-r*-— 6 I I l I I I l r l I l l g ____I___I___I____I ______ |___I____I___I__‘q____l___I____l__._I___‘ ; P_—I__I__‘I_-_I ______ I_*I__I__I~_ ~_|_~I__I“_I‘_ IL] I— I I I I ‘I “I T T r I‘ I“ a >——-—I—~—I———I——-——I ______ I———«I—-———I——-—I——-— -——I—~—-I—~——I——-—I——— _I ,___I____I_____I___I ______ I____I_____I____I___]__E_I___]___I____I_____‘ I I I I I I I I I I I I 5 I I I I I I I I I I I I I-——I———I———I———I —————— I———I—~——I-———+—~— -—+-——I———I———I——— r___|___|___l____l ______ I____I____I __I__I__L__L__I___I___I I I I I I I I I I I I I 7""I—__I__—I‘“_I """""" I“‘7“’I"‘7““”‘7T"_T__I__”I“'-‘ *——I—-"I—-—I“"—I———-——I———I———i_~#——‘I’--’+—~’r—*t’*—I——— 4 l l l 1 l l 1 Jr 1 l l l 100 150 200 LOCAL LASER INTENSITY in m]: Figure D.3: Variations in lifetime with variations in local laser intensity. D.4 ADDITIONAL SOURCES OF ERROR FOR MTT Apart from the aforementioned causes, there are certain other experimental parameters which also affect the accuracy of our measurements. These are discussed below. ERRORS CONNECTED T O T EMPERA T URE MEASURING DEVICES. Inaccuracy connected with the thermocouple is also something that needs to be accounted for. The T-Type copper constantan thermocouple used typically has at least a i 1 °C error in reading [OMEGA engineering: source for thermocouple wires]. This was confirmed using a more accurate RTD sensor. 137 EXTENT OF PURGING OF OXYGEN FROM THE FLOWLINES The oxygen content in the nitrogen cylinder at room pressure was 2.2 x10—7 mol/ 1 when expanded to room pressure. If this value doesn’t remain fixed due to either a change in the source of gas, non uniformity in the quality of nitrogen or the level of purging of the system, an estimate of its effect on the resultant lifetimes becomes important. A change in the oxygen content by a factor of 10 leads to the following changes in lifetimes as given in (Table D.2). The example shown is for acetone concentration corresponding to its vapor pressure at 22°C and room pressure. 0.1X[02] 1X[02] 10X[02] 7.32us 7.25us 6.54us TableD.2: Example of variation in lifetime with change in oxygen concentration for MTT. From the above table once can conclude that the changes are a lot more adverse when the oxygen content goes up than down. This is because the acetone self-quenching still remains dominant compared to oxygen quenching at low oxygen levels. Typically given the short lengths and the simple layout of the tubing used, keeping the flow running for two minutes was more than enough time to purge the lines and start taking readings. 138 VARIABILI T Y IN THE MA GNITUDE OF EXPANSION COOLING OF NITROGEN The source of nitrogen for the experiments is a standard industrial nitrogen cylinder which is 183bar when full and has <5ppm by volume of oxygen. As shown in Figure Cl and described before, the flowrate is controlled by a dual stage regulator mounted on the cylinder outlet valve. However since the pressure inside the cylinder (183bar) is a lot more than the pressure just downstream of the regulator (2bar), a large pressure drops occurs across the pressure regulator and as a result there is some amount of expansion cooling. The magnitude of this depends on how full or empty the nitrogen cylinder is. The higher the pressure of the cylinder, lower is the temperature downstream of the regulator since the pressure drop is bigger. As a result, when a new cylinder feeds the seeding chamber, the temperature drop inside the seeding chamber is more than when a relatively less filled cylinder is connected. This was dealt with by recording the actual temperatures inside the seeding chamber for the experiments whose data is presented. This would ensure that the concentration of acetone in the exiting flow is known as accurately as possible. 139 APPENDIX E 140 APPENDIX E: FLEXIBLE MIXTURE COMPOSITION SYSTEM FOR MTM EXPERIMENTS A general layout of the proposed system (see Figure E.2) with a detailed description of each component is given. This is followed by a description of the operating procedure that will be used. Apart from the test cell described in Chapter 3, Section 3.2.1., the test rig will comprise of the following main parts: AUXILIARY/REFERENCE TANK: The basic use of the auxiliary tank (Figure E.1) is to use it as a reference tank using which the acetone, nitrogen and acetone mixture of required composition will be created. The amount of oxygen filled into the larger tank will estimated by comparing the pressure differential across between the reference tank and the larger tank. "if Figure E.l: Auxiliary/ reference tank 141 .Eoumzm coEmomEoo 23ch 2&ch no 33828 2.52 "Nam .ouawE T $ 02!, 1:58 .35 O 0 3:8 gong—Sag JO autonomagévug 82g ggzgméoqn bah—353532 Bu tom 03.5 uooacag 253.. 839 65315 33> 1288 mg «mam £32m 2:5 gecko—930% ..=J Spa «25 2.303 was .8352 G 2:!» Beau—8 main 31> 1238 he: 1 ”Baum—ho 8908 :3 _l 82m 83 aowg 32 he tom owns.— b 2:? 3:52.: as?“ , z 4 I. I. , 98 ea £9.25 :2. omen 33> 138. .35 83 Has magic—Quasa— uongo 2508 g :8 3 89¢ 60? 0508 omen 0.3895 8 “ago 3308 v5 3&ng v8 :0»ng 33> 280—8 .55 3:8 EEO—53°33— 83 :8 88 2 «03 3208 v8 nowgmz gm A SO>