Simultaneous OH- and Formaldehyde-LIF Measurements in an HCCI Engine

Size: px

Start display at page:

Download "Simultaneous OH- and Formaldehyde-LIF Measurements in an HCCI Engine"

Georgiana Phelps
5 years ago
Views:

Simultneous OH- nd Formldehyde-LIF Mesurements in n HCCI Engine Collin, Roert; Nygren, Jenny; Richter, Mttis; Aldén, Mrcus; Hildingsson, Leif; Johnsson, Bengt Pulished in: SAE Trnsctions, Journl of

1 Simultneous OH- nd Formldehyde-LIF Mesurements in n HCCI Engine Collin, Roert; Nygren, Jenny; Richter, Mttis; Aldén, Mrcus; Hildingsson, Leif; Johnsson, Bengt Pulished in: SAE Trnsctions, Journl of Fuels nd Luricnts 23 Link to puliction Cittion for pulished version (APA): Collin, R., Nygren, J., Richter, M., Aldén, M., Hildingsson, L., & Johnsson, B. (23). Simultneous OH- nd Formldehyde-LIF Mesurements in n HCCI Engine. SAE Trnsctions, Journl of Fuels nd Luricnts, 2(4), Generl rights Copyright nd morl rights for the pulictions mde ccessile in the pulic portl re retined y the uthors nd/or other copyright owners nd it is condition of ccessing pulictions tht users recognise nd ide y the legl requirements ssocited with these rights. Users my downlod nd print one copy of ny puliction from the pulic portl for the purpose of privte study or reserch. You my not further distriute the mteril or use it for ny profit-mking ctivity or commercil gin You my freely distriute the URL identifying the puliction in the pulic portl Tke down policy If you elieve tht this document reches copyright plese contct us providing detils, nd we will remove ccess to the work immeditely nd investigte your clim. L UNDUNI VERS I TY PO Box7 22L und

2 Simultneous OH- nd Formldehyde-LIF Mesurements in n HCCI Engine Roert Collin, Jenny Nygren, Mttis Richter nd Mrcus Aldén Division of Comustion Physics, Lund Institute of Technology Leif Hildingsson nd Bengt Johnsson Division of Comustion Engines, Lund Institute of Technology Copyright 23 SAE Interntionl ABSTRACT Simultneous OH- nd formldehyde LIF mesurements hve een performed in n HCCI engine using two lser sources working on 283 nd 355 nm, respectively. Two ICCD cmer systems, equipped with long-pss filters, were used to collect the LIF signls. The simultneous imges of OH nd formldehyde were compred with het-relese clculted from the pressure-trce mtching the cycle for the LIF mesurements. The mesurements were performed on.5 l singlecylinder opticl engine equipped with port-fuel injection system. A lend of iso-octne nd n-heptne ws used s fuel nd the compression rtio ws set to 2:. The width of the lser sheet ws 4 mm nd hence covered pproximtely hlf of the cylinder ore. At some 2 CAD BTDC low temperture rections is present nd formldehyde is formed. The formldehyde signl is then rther constnt until the min het-relese strts just efore TDC, where the signl decreses rpidly to low vlues. From some 5 CAD to 5 CAD BTDC the formldehyde is uniformly distriuted in the imged re. As formldehyde decreses, OH increses nd follows the min rte of het relese curve, though with slight lg in phse. Therefter OH is formed in the res from which the formldehyde hs disppered nd the OH signl is present to some 2 CAD ATDC. INTRODUCTION When running n engine in HCCI-mode (HCCI=Homogeneous Chrge Compression Ignition) the fuel nd ir re llowed to mix in the intke port, or in the cylinder for DI-engines, in order to crete homogeneous chrge. The chrge is then compressed until uto-ignition occurs. A chrcteristic feture with HCCI tht distinguish it from more conventionl comustions modes, i.e. Otto or Diesel, is tht ignition occurs simultneously t multiple sites []. This results in comustion tht occurs through distriuted rections throughout the ulk volume. As the whole ulk urns lmost simultneously, the overll comustion rte is very high. Consequently, highly diluted mixtures hve to e used in order to limit the rte of het relese. The mjor dvntge with HCCI compred to the diesel concept is low NOX emissions in comintion with significntly less soot formtion [2]. In conventionl diesel engines, reduction of NOx most often results in n increse of soot volume frction nd vice vers. An dvntge for the HCCI, when compred to n SI engine, is higher efficiency t prt lod. In ddition, the cycle-to-cycle vritions in Indicted Men Effective Pressure (IMEP) re very smll for HCCI comustion [3]. A mjor chllenge ssocited with HCCI is the difficulty to control the timing of the ignition so to tht the comustion occurs close to TDC (TDC=Top Ded Center). Since the ignition process relies on spontneous uto-ignition there re no direct mens to control the onset of comustion. However, y djusting remote operting prmeters, like inlet ir temperture, fuel properties or exhust gs recircultion (EGR) rte, ignition timing cn e controlled [4, 5]. The ignition nd comustion processes cn e studied y visulizing the distriutions of formldehyde nd OH rdicls. The formldehyde molecule is formed s n intermedite species when comusting hydrocrons. The formtion occurs through low temperture oxidtion in n erly phse of the ignition process. Formldehyde

3 is then eing consumed lter in the comustion process. Formldehyde is lso ssocited with the low temperture rections tht occur when certin mixtures of hydrocron fuels nd ir re close to the explosion limit. Hence, formldehyde is n indictor of the first stge of ignition nd mrker for zones with low temperture rections. Plnr lser induced fluorescence from formldehyde hs een used erlier for visuliztion of self-ignition centres in SI-engines [6], for studying the ignition of Diesel sprys [7] nd for chrcteriztion of the Controlled Auto Ignition (CAI) comustion concept [8]. The OH rdicl is n importnt intermedite tht is formed during comustion. OH is formed in flme regions with high temperture nd there is strong coupling etween mximum temperture nd mximum OH concentrtion. For cses where conventionl flme fronts exist, the OH rdicl is often used s mrker of oth flme fronts nd of urnt regions. Becuse of the comustion temperture in n HCCI engine eing lower thn in n Otto engine, the OH concentrtion in n HCCI engine is significntly lower compred to wht is found in n Otto engine. However, OH cn still e used s mrker of regions where HCCI comustion is ongoing [9]. In this work the ignition nd comustion phses for HCCI engine re investigted with lser-induced fluorescence, LIF, on oth formldehyde nd OH simultneously. For this purpose two lser sources comined with two ICCD detector systems were used. EXPERIMENTAL ENGINE SETUP The work presented ws performed on.5 l singlecylinder opticl HCCI engine, the specifictions of which re summrized in tle. The engine ws equipped with port-fuel injection system, which genertes principlly homogeneous chrge, nd lend of 5% iso-octne nd 5% n-heptne ws used s fuel. The Type: Engine speed: Displcement (one cylinder): Bore: Stroke: compression rtio ws set to 2: nd the engine ws run with lmd etween 2.9 nd 3.3. A wide-nd lmd sensor, ETAS LA3, ws used to mesure the ir/fuel-rtio. To mintin 5% het relesed t 5 6 CAD ATDC the inlet ir temperture hd to e constntly chnged. The engine ws strted t n inlet ir temperture of 2 C nd during the mesurement it hd to e constntly decresed nd the temperture t the end of the mesurement ws etween 8 C nd 9 C. Horizontl opticl ccess to the upper prt of the cylinder liner ws otined through full qurtz ring with height of 25 mm. An elongted piston nd 45-degree mirror provided for opticl ccess from elow through piston window with dimeter of 58 mm. OPTICAL SETUP Four-vlve diesel engine 2 rpm.5 l 8 mm 93 mm Compression rtio: 2: Fuel: Tle. Specifictions of the test engine. 5% isooctne 5% n-heptne Simultneous OH- nd formldehyde-lif mesurements were performed in the HCCI engine using two lser sources fired with time seprtion of s in order not to distur ech other. For the OH-LIF frequency douled Nd:YAG ws used to pump dye lser with Rhodmin 59. The dye lser output ws further frequency douled using single BBO crystl nd the resulting pulse energy ws pproximtely 3 mj ner 283 nm. OH ws excited using the Q(8) trnsition in the v =,v = nd of the A2 +-X2 + system. For the formldehyde-lif the third hrmonic (355 nm) of Nd:YAG ws used with pulse energy of pproximtely 75 mj. The technique for formldehyde LIF is descried elsewhere []. The lser ems from the two lser Figure. Schemtic overview of the experimentl setup. Left: Lser nd detector lignment through the engine. Right: Position of the imged re reltive to the comustion chmer. 2

4 sources were lined on top of ech other nd therefter shped, with the sme optics, into horizontl lser sheet with width of 4 mm. The sheet ws focused with n f=3 mm cylindricl lens into the comustion chmer, 3 mm elow the roof. The lser sheets were lined in order to cover hlf the cylinder, see figure. Two ICCD-cmer systems (Princeton Instruments PI-MAX), equipped with UV-lenses (Nikon, f =5 mm, f#=2.5) were used to collect the OH- nd formldehyde-lif signls nd high-reflective mirror centered t 38 nm nd with trnsmittnce of pproximte 7% t 355 nm ws used s em splitter to seprte the two LIF signls from ech other. LIF from OH ws detected in the (,) nd (,) nds ner 39 nm using two 283-nm long-pss filters nd one UG filter in front of the detector. LIF from formldehyde ws detected using two GG395 Shott-filters. SPECTRAL INVESTIGATION The formldehyde signl ws recorded spectrlly in order to confirm tht the signl detected y the formldehyde-detector ws otined from formldehyde. This investigtion ws performed y replcing the ICCDsystem collecting the formldehyde-lif with spectrogrph. The otined spectrum, shown in figure 2, shows the LIF signl from formldehyde nd is similr to emission spectr of formldehyde otined in heted cell. The spectrum hs rod structure compred with similr spectr otined t lower tempertures. To some extent, this cn e explined y rodening of the nds since more vironic nd ro-vironic levels would e populted t higher tempertures []. The pek shown t 532 nm in figure 2 origintes from the frequency-douled component used for generting the lser rdition t 355 nm. Intensity (.u.),8,6, lmd (nm) Figure 2. Spectrlly resolved formldehyde signl from the HCCIcycle. RESULTS For every mesured crnk-ngle position ckground imges, from the corresponding motored cse, were collected. These ckground imges were sutrcted from the mesured LIF imges nd the remining noise level were set to zero in the imges. By this, only informtion out the LIF signl intensity ws remining in the OH- nd formldehyde-imges. For clrity the intensity imges shown in the pper re flttened in order to see the sptil structures in the imges more esily. This procedure is shown in figure 3 where the intensity imges (to the left in figure 3) re flttened to one single color efore the OH- nd formldehyde imges re comined to one single imge (to the right in figure 3). Mesurements were performed in wide time rnge, strting efore the onset of the low temperture region (LTR), through ll comustion nd until lte in the expnsion stroke. For every mesured crnk-ngle position 2 imge-pirs were collected nd the corresponding pressure-trces were recorded. Those temporl positions of ech imge cquisition were Formldehyde OH Figure 3. Procedure for representtion of OH- nd formldehyde-lif imges. To the left re intensity imges of OH nd formldehyde shown. The second column represents the flttened imges. The comined OH- nd formldehyde imge is shown to the right. 3

5 mrked in the pressure-trces. In figure 4 selected series of LIF imges otined from different crnkngles is shown to exemplify the formtion nd consumption of formldehyde nd OH during the comustion when the engine is run with lmd of Rther thn presenting verge imges, however, single-shot imges representtive for ech position re presented. This is ecuse verging smers out the shpe of the sptil distriution of the OH- nd formldehyde signls. In figure 5 the verge rte of het-relese for individul cycles collected in connection with the LIF imges otined t CAD ATDC is shown. Also shown in this figure re the men vlues of the totl sum-up intensity for the 2 single-shot imges otined t ech mesured crnk-ngle position. These men vlues re normlized to the mximum vlue of the OH- nd formldehyde signls, respectively. To cover the intervl from the onset of low temperture rections until lte in the expnsion stroke the mesurement series hd to e divided into two prts (indicted y nd 2 in figure 5) in order not to overhet the engine Figure 4. Single-shot imges from onset of LTR comustion until the end of the min comustion. Formldehyde is shown in green/light gry nd OH is shown in red/drk gry OH- nd Formldehydesignl (.u.),8,6,4 Form OH Form 2 OH 2 RoHR CAD Figure 5. Averge rte of het-relese nd verge OH- nd formldehyde-signls for the mesurement series shown in figure 4. From figure 4 nd 5 it cn e seen tht low temperture region is present t some 2 CAD BTDC nd formldehyde is formed. From some 5 CAD to 5 CAD BTDC the formldehyde is uniformly distriuted in the imged re nd this uniformly distriuted formldehyde signl remins until the min het relese strts just efore TDC. Here, the signl decreses rpidly to low vlues nd s formldehyde decreses OH increses nd follows the min rte of het-relese curve, though with slight lg in phse. Therefter OH is formed in the res from which the formldehyde hs disppered nd the OH signl is present to some 2 CAD ATDC. OH signl do not fill the entire comustion chmer ecuse the signl is prompt nd dispper when the comustion is over. Everything in the comustion chmer does not urn t the sme time. Lter in the expnsion stroke, round 5 CAD ATDC, signl strts to pper gin t the formldehydedetector. This signl my come from formldehyde ut this is not confirmed since no spectrl investigtion of the signl ws mde this lte in the cycle. Therefore it is possile tht this signl my come from other species tht fluoresce when excited y 355 nm lser rdition. Figure 6 is zoomed-in version of the grph in figure 5. Here only the first hlf of the mesurement series is shown, ut with the stndrd devition of the mesurements tht uild ech verged point included. From this it cn e seen tht the stndrd devition is smll up to the point where the min comustion tkes plce, especilly for the OH formtion. The stndrd devition for the undefined signl lter in the cycle (not seen in figure 6) is of the sme mgnitude s for formldehyde during the min comustion RoHR (J/CAD) 4

6 OH- nd Formldehyde- Signl (.u.),2,8,6,4 Formldehyde OH RoHR RoHR (J/CAD) CAD Figure 6. Averge OH- nd formldehyde signls with the stndrd devition of the mesurement points included. In figure 5 nd 6 the LIF mesurements re represented y the verge sum-up intensity of the imges for ech mesurement position. Another wy to get n overview of the mesurement series is to represent ech LIF imge-pir y the surfce tht is otined y the signl of formldehyde nd OH, respectively, s in figure 4, nd uild the verge surfce for ech mesurement position. This representtion ends-up in grph similr to those in figure 5 nd 6 with the difference tht, since the imge is flttened, the LIF signl dependence of lser power, pressure, numer density, etc. is removed. This representtion is shown in figure 7 where the verge surfce of the signls is normlized to the totl imged re tht cn e occupied with signl. Formldehyde Surfce,2,8,6,4 Formldehyde OH CAD Figure 7. Averge surfce otined with signl for OH nd formldehyde for ech mesurement position Shown trends of formldehyde nd OH in figure 7 re similr with those shown in figure 6, including similr stndrd devitions. It cn e seen tht formldehyde fills the totl imged re etween the low temperture region nd the min comustion (high temperture region, HTR) which in this region result in constnt level with very smll stndrd devition. In verge OH fills lmost 3 % of the imged re efore it strts to dispper.,35,3 5,5,,5 Figure 8. Percentge het relesed s function of CAD. Figure 8 (lener) ove; 8 (richer) elow. By plotting the percentge of het relesed, s shown in figure 8, it is seen tht there is some spred of the dt for ech CAD. For ech CAD 2 imges were recorded nd thus lso 2 individul pressure trces. Ech one is represented y n x. Due to the rpid comustion phse of HCCI there re some CAD-positions where the het relesed is spred over s much s 25% of the verticl xis. During the period efore TDC nd the period fter 2 ATDC the comustion is much more stle nd doesn t hve such lrge spred, see figure 8. Due to rounding of the results there seem to e only two x for the two lst CAD-positions in figure 8, ut there re 2; they re on top of ech other. The LTR/pre-rections cn e seen contriuting with out 3% of the totl mount of het relesed. It should e noted tht het relese of % mens tht the comustion is finished, ut this doesn t imply tht ll of the fuel tht ws injected hs een consumed; on the contrry there is proly considerle mount of unurnt fuel trpped in the inevitle crevices tht occur when using n engine with opticl components. Figure 8 shows plot for one cse with slightly lener mixture, lmd = 3.3, thn the mixture used for figure 6 nd 7, which ws 3. to 3.; 8 is for cse with somewht richer mixture, out lmd 2.9. To mintin 5% het relesed t 5 6 CAD ATDC the inlet ir temperture hd to e constntly chnged. 5

7 Figure 9. Normlized surfce of formldehyde nd OH s function of CAD. Figure 9 (lener) ove; 9 (richer) elow. Squres represent formldehyde nd tringles represent OH. In figure 9 the normlized surfce percentge filled y formldehyde nd OH is plotted s function of CAD, for the 2 cycles per CAD position. The mximum formldehyde surfce present in ech mesured cse ws determined nd ll mesured surfces were susequently normlized to the gretest of the mximum formldehyde surfces. A lrge spred in signl level cn e seen for given CAD, especilly for the OH signl. As oth the percentge het relesed nd PLIF signl intensity vry for given CAD it is interesting to plot PLIF signl vs. percentge het relesed insted. Figure shows this. Figure. Normlized surfce of formldehyde nd OH s function of percentge het relesed. Figure (lener) ove; (richer) elow. Squres represent formldehyde nd tringles represent OH As seen in figure, where the normlized surfce re of formldehyde nd OH is plotted s function of percentge het relesed, the formldehyde is lwys present in lmost the whole viewed re during the time from % to 3% het relesed, which trnsltes to somewht efore 5 BTDC to round TDC when compring figure with figure 9; this is were the LTR occur. The normliztion is the sme s for figure 9. In the plot, squres represent formldehyde surfce nd tringles represent OH surfce. The decresing trend of the formldehyde surfce is clerly seen, with lrger spred for the lener cse (figure ) thn for the cse with slightly richer mixture. The fluctution of the OHsurfce is considerle, which is lso seen in figure 6 y studying the stndrd devitions for the cse with lmd = One notle oservtion is tht there is no OH signl until round 6% het relesed for figure, nd erlier, t out 4% het relesed for figure. 6

8 Figure. Normlized formldehyde nd OH intensities s function of CAD. Figure (lener) ove; (richer) elow. Squres represent formldehyde nd tringles represent OH. In figure it cn clerly e seen tht the intensities of formldehyde nd OH re lowe in the cse of the lener mixture compred with the richer mixture. The intensity of formldehyde ws normlized to the highest formldehyde intensity in ny of the studied cses; the sme ws done for the OH-intensities. Figure 2 nd shows the normlized formldehyde nd OH intensities s function of percentge het relesed. The normliztion is the sme s for figure. Figure 2. Normlized formldehyde nd OH intensities s function of percentge het relesed for the lener cse. Squres represent formldehyde nd tringles represent OH. Figure 2 (lener) ove; 2 (richer) elow. CONCLUSIONS The concentrtion nd loction of formldehyde nd OH were mesured in n opticl engine using two seprte PLIF systems. The engine ws operted with 5/5 mixture of iso-octne nd n-heptne. This generted low relesed some 2 CAD efore the high temperture rections. At the strt of the low temperture rections clouds of formldehyde were detected. More formldehyde ws detected s the LTR progressed. The formldehyde filled the entire mesured region fter the LTR ended. 7

9 At the strt of the HTR holes in the homogeneous formldehyde signl were detected. At pek het relese rte the signl hs gone down to less thn 5 of its mximum. The PLIF signl of OH is detected only in regions were formldehyde is sent. Under reltively long period OH nd formldehyde is detected simultneously ut never in the sme loction. The PLIF signl strength of OH is lgging the rte of het relese y roughly 8 CAD. Mx OH is detected when most of the fuel is consumed nd thus close to pek in-cylinder temperture. Running the engine with len mixture mkes the OH signl pper t higher percentge het relesed compred to rich mixture. This could e due to the lower mount of het relesed with the lener mixture nd thus lower temperture during the cycle. Significnt cycle to cycle vritions exist during the HTR comustion process wheres the LTR is very stle. This gives very low cycle to cycle vrition of the formtion of formldehyde ut much more vritions in the destruction t the first phse of HTR. Both formtion nd destruction of OH exhiited lrge cycle to cycle vritions. ACKNOWLEDGMENTS This work ws finnced y The Green Cr project nd The Centre of Competence. The uthors re grteful for this support. CONTACT Corresponding uthor: Roert Collin, Division of Comustion Physics, Lund Institute of Technology, PO Box 8, S-22, LUND, Sweden, phone: , fx: , e-mil: Roert.collin@forrf.lth.se. REFERENCES. A. Hultqvist, M. Christensen, B. Johnsson, A. Frnke, M. Richter, M. Aldén: A Study of the Homogeneous Chrge Compression Ignition Comustion Process y Chemiluminescence Imging, SAE M. Christensen, A. Hultqvist, B. Johnsson: Demonstrting the Multi Fuel Cpility of Homogeneous Chrge Compression Ignition Engine with Vrile Compression Rtio, SAE S. Onishi, S. Hong Jo, K. Shod, P. Do Jo, S. Kto: Active Thermo-Atmosphere Comustion (ATAC) A New Comustion Process for Internl Comustion Engines, SAE J-O. Olsson, P. Tunestl, B. Johnsson: Closed-Loop Control of n HCCI Engine, SAE J-O. Olsson, P. Tunestl, G. Hrldsson, B. Johnsson: A Turo Chrged Dul Fuel HCCI Engine, SAE R. Schießl, A. Dreizler, U. Ms, A. Grnt, P. Ewrt: Doule-Pulse PLIF imging of selfignition centers in n SI engine, SAE H. Kosk, V. H. Drewes, L. Ctlfmo, A. A. Ardi, N. Iid nd T. Kmimoto: Two- Dimensionl Imging of Formldehyde Formed During the Ignition Process of Diesel Fuel Spry, SAE N. Grf, J. Gronki, C. Schulz, T. Britud, J. Cherel, P. Duret, nd J. Lvy: In-cylinder Comustion Visuliztion in n Auto-Igniting Engine using Fuel Trcer- nd Formldehyde- LIF Imging, SAE M. Richter, A. Frnke, M. Aldén, A. Hultqvist nd B. Johnsson: Opticl Dignostics Applied to Nturlly Aspirted Homogeneous Chrge Compression Ignition Engine, SAE C. Brckmnn, J. Nygren, X. Bi, Z. Li, H. Bldh, B. Axelsson, I. Denrtt, L. Koopmns, P.-E. Bengtsson nd M. Aldén: Lser-Induced Fluorescence of Formldehyde in Comustion using Third-Hrmonic Nd:YAG Lser Excittion, sumitted to Spectrochimic Act Prt A (23).. T. Metz, X. Bi, F. Ossler, nd M. Aldén: Fluorescence lifetimes of formldehyde (H 2 CO) in the Ã A 2!X A nd system t elevted tempertures nd pressures, sumitted to Spectrochimic Act Prt A (23). 8

Effect of RME Fuel on Combustion in Optical Engine Operating in HCCI Mode

Effect of RME Fuel on Combustion in Optical Engine Operating in HCCI Mode Effect of RME Fuel on Comustion in Opticl Engine Operting in HCCI Mode E. Mncruso nd B. M. Vglieco Istituto Motori - C.N.R., Npoli ITALY 1. Introduction In the lst thirty yers it hs een growing increse