![]() |
||||
|
|
|
|
||
| January 1995 Vol. 5 No. 1 ERC HAS NEW DIRECTOR On July 1, 1994Gary Borman retired after 30 years of service on the U.W. Faculty and DavidFoster took over as the new ERC Director. Professor Foster joined the U.W.faculty in 1978 after receiving his Ph.D. from MIT. He plans to keep activein research and teaching while serving as director. Dave, who has receivedmany awards for his teaching, will continue to teach the senior coursein I.C. engines and the graduate course in Chemical Kinetics. Much of hiscurrent research is concerned with emissions, particularly optical measurementsof soot in engines and models l for soot suitable for use in CFD codes. JAAL GHANDHI JOINSERC FACULTY Jaal Ghandhiwho is just completing his Ph.D. at Princeton has joined U.W.-Madison asan Assistant Professor of Mechanical Engineering this January. Jaal, whois a Wisconsin native and Badger, will be doing research in the ERC applyinglaser optics to a variety of engine combustion topics. One topic will bedirect-injectedtwo-stroke marine engines, a natural follow-up of his doctoralthesis as well as following his father's life time involvement with marinetwo-strokes. ENGINE LAB CO-FOUNDERWINS ASME AWARD Professor EmeritusOtto A. Uyehara, who with Phil Myers founded the Engine Lab at U.W.-Madison,recently received the ASME Internal Combustion Engine Award. This awardadds one more star to Otto's galaxy of awards received for his contributionsto engine research and teaching. Otto has been living in Anaheim, Californiasince shortly after his retirement in 1982, but has kept busy consulting,writing papers and organizing technical meeting sessions. RUTLAND WINS PAPERAWARD Professor Chris Rutland andhis co-author, S. Zhang, won the best paperaward at the Spring Meetingof the Central States Section of the Combustion Institute for their papertitled, "Heat Release Effects on Turbulence in a Premixed Flame". Thispresentation was a shortened version of a manuscript which has been submittedfor publication to Combustion and Flame. An abstract of that paper is presentedelsewhere in this newsletter. ERC RELATIONSHIPS GROW The ERC has recently extendedits Co- operative research relationship withIstuto Motori of Naples, Italyand formed new relationships with four other groups. The new relationshipsare:
The ERC currently has fourvisiting faculty, each contributing to our researchefforts. Professor Sang soon Hwangfrom U. Inchan, in Korea has been visiting since February, 1994. He isstudying droplet behavior in high-speed airflows using high resolutionphotography and CFD calculations. Professor Zhengbai Liu ofDalian U. in China has been visiting since February 1994. He is studyingCFD heat transfer modeling for liquid fuel jets impacting on surfaces. Professor Hidoyuki Ogawaof Hokkaido U. in Japan has been visiting since September, 1994. He isdoing a study of particulate oxidation using the ERC Cummins single cylinderengine. Professor biro Senda of DoshishaU. in Japan has been visiting since August 1994. He is continuing his workon detailed modeling of a diesel spray impinging on a piston bowl surface. NEW PAPERS FROM THEERC Fourteen papers to be presentedby the ERC this spring are briefly reviewed below. The first eleven willbe given at the SAE 1995 Congress. SAE 950083 "Nozzle Effectson High Pressure Diesel Injection" by TV. Su, P.V. Farrell and R.T. Nagarajan.This paper is the result of cooperation between Caterpillar Inc. and theERC during which Mr. Najarajan of Caterpillar spent many days in Madisonworking with Professor Farrell and his student TF. Su. In this work studiesof transient diesel spray characteristics at high injection pressures wereconducted in a constant volume chamber by utilizing a high speed photographyand light extinction optical diagnostic technique. Two different typesof nozzle hole entrances were investigated: a sharp- edged and a round-edgednozzle. The experimental results show that for the same injection delivery,the sharpedged inlet injector needed a higher injection pressure to overcomethe higher friction loss, but it produced longer spray tip penetrationlength, larger spray angle, smaller droplet sizes, and also lower particulateemission from a parallel engine test. For the round-edged and smooth edgedtips at the same injection pressure, the sharp-edged inlet tip took a longerinjection duration to deliver a fixed mass of fuel and produced largeroverall average Sauter Mean Diameter droplets. The sharwedged inlet alsoproduced shorter penetration, but almost the same spray angle for b othhigh and low fuel delivery cases. SAE 950162 "Single-SurfaceFlame Quenching Distance Dependence on Wall Temperature,Quenching Geometry, and Turbulence," by DJ. Cleary and P.V.Farrell. The effect of walltemperature on single- surface flame quenching distance was characterizedfor atmospheric, premixed methaneair flames. The studyincludes a comparison of the wall temperature dependent, singleQsurfaceflame quenching distance for laminar and turbulent flames. The laminar flame-wall interaction was studied for flames that were configured at anglesnear 0! and 45! relative to a temperature-controlled surface. For eachflame quenching configuration, the flame quenching distance was chosenas the locationfrom the surface for which a constant value on C2 concentration occurred;spatially resolved measurements of C2 concentration were obtainedwith the technique of laser-induced fluorescence. The results indicated that the single-surfaceflame quenching distance, for each flame configuration, decreasedwith increasing wall temperature.For the laminar flames, Rarnan based gas temperature measurementsindicated that the near-wall temperature gradient wasreduced when the wall temperature was increased. The wall heat flux wasestimated from the measurednear-wall temperature profile and gas thermal conductivity. These resultsindicated that the wall heat flux decreased for the laminarsidewall flame and increased for the laminar stagnation flame forincreasing wall temperaturefrom 250 C to 600 C. SAE 950217 "Reducing Particulateand NOX Using Multiple Injections and EGR in a D.I. Diesel," by D A. Pierpont,D.T. Montgomery, and RD. Reitz. An emissions and performance study wasconducted to explore the effects of EGR and multiple injections on particulate,NOX, and BSFC. EGR is known to be effective at reducing NOX, but at highloads there is usually a large increase in particulate. Recent work hasshown that multiple injections are effective at reducing particulate. Thus,it was of interest to examine the possibility of simultaneously reducingparticulate and NOX with the combined use of EGR and multiple injections.The tests were conducted on a fully instrumented single cylinder versionof the Caterpillar 3406 heavy duty truck engine. Tests were done at highload (75% of peak torque at 1600 RPM) where EGR has been shown to produceunacceptable increases in particulate emissions. The fuel system used wasan electronically controlled, common rail injector and supporting hardware.The fuel system was capable of up to four independent injections per cycle.Two noble spray angles were used with included spray angles of 125!and140! (i.e., with and without significant spray wall impingement). Particulateversus NOX trade-off curves were generated over a range of injection timingsfor each injection scheme and EGR rate studied. The results show that thecombined use of EGR and SAE 950278 "The Developmentand Application of a Diesel Ignition and Combustion Model for MultidimensionalEngine Simulations," by S.-C. Kong, Z. Han, and R.D. Reitz. An integratednumerical model has been developed for diesel engine computations basedon the KIVA-II code. The model incorporates a modified RNG k-e turbulencemodel, a 'wave' breakup spray model, the Shell ignition model, the laminarand-turbulentcharacteristic-time combustion model, a crevice flow model, a spray/wallimpingement model that includes rebounding and breaking-up drops, and otherimproved submodels in the KIVA code. The model was validated and appliedto model successfully different types of diesel engines under various operatingconditions. These engines include a Caterpillar engine with different injectionpressures at different injection timings, a small TACOM engine at differentloads, and a Cummins engine modified by Sandia for optical experiments.Good levels of agreement in cylinder pressures and heat release rate datawere obtained using the same computer model for all engine cases. Predictionsof engine emissions were also performed and a good quantitative agreementbetween measured and predicted NOx and soot emission data were obtainedwith the use of the present modified RNG turbulence model. The resultsindicate that the predicted flow structures have an important impact onthe prediction of NOX formation since it is very sensitive to the localtemperatures in the combustion chamber. SAE 950281 "An Applicationof the Coherent Flamelet Model to Diesel Engine Combustion," by M.P. Musculusand CJ. Rutland. A turbulent combustion model based on the coherent flameletmodel was developed in this study and applied to diesel engines. The combustionwas modeled in three distinct but overlapping phases: low temperature ignitionkinetics using the Shell ignition model, high temperature premixed burnusing a single step Arrhenius equation, and the flameletbased diffusionbum. Two criteria for transition based on temperature, heat release rate,and the local Damkohler number were developed for the progression of combustionbetween each of these phases. The model was implemented into the computationalcomputer code KIVA-II. Previous experiments on a Caterpillar model E 300,# lY0540 engine, a TACOM Labeco research engine, and a single cylinderversion of a Cumrnins N14 production engine were used to validate the cylinderaveraged predictions of the model. The characteristics of the modelingapproach were also addressed by examining the spatial resolution of themodel results inside the engine cylinder. The location and magnitude ofmodel heat releases, flame areas, and spray distributions were examined.The results of this approach to the modeling of diesel engine combustion SAE 950282 "Modeling of Effects ofIntake Flow Characteristics on Diesel Engine Combustion," by P.W. Stephensonand C.J. Rutland. The tlueedimensional CFD codes KIVAII and KIVA-3 havebeen used together to study the effects of intake generated in-cylinderflow strucD on fuel-air mixing and combustion in a direct injected (DI)Diesel engine. Simulation of the intake and compression strokes in a heavy-dutyfourstroke M Diesel engine has been carried out using KIVA-3. Flow quantitiesand thermodynamic field information were then mapped into a simpler computationalgrid in KIVA-II for use in the study of mixing and combustion. A laminarand turbulent time scale combustion model, as well as advanced spray models,including wave breakup atomization, dynamic drop drag, and spray-wall interactionhave been used in KIVA-II. Simulation of the mixing of the individual fueljets from the six hole injector shows significant differences in spraypenetration and droplet breakup, owing to sizable inhomogeneities in thevelocity and turbulence fields present at the time of fuel injection. Comparisonof a single fuel jet injected into the realistic in-cylinder flow fieldwith one injected into a uniform flow field which had equivalent averagevalues for all thermodynamic and velocity terms shows major differencesin peak heat release rate, and the timing of the peak itself. Comparisonof two fuel jets, injected into different parts of the same realistic flowfield having large differences in the local swirl number show that ignitiondelay is reduced in the case of the higher swirl number. The same comparison,made with the assumption of no turbulence, shows much less difference inthe heat release. SAE 950285 "Multidimensional Computationof Multicomponent Spray Vaporization and Combustion," by N.S. Ayoub andRD. Reitz. The threedimensional KIVAcode has been used to study the effectsof multicomponent fuel droplet vaporization on diesel engine performanceunder both standard and cold-starting conL ditions. The code has also beenupdated with stateof-the-art submodels including: a wave breakup atomizationmodel, drop drag with drop distortion, spray/wall interaction with sliding,rebounding, and breaking-up drops, multistep kinetics ignition and laminar-turbulentcharacteristic time combustion, wall heat transfer with unsteadiness andcompressibility, and crevice flow model. The baseline computational resultsare compared with experimental data from a single-cylinder Caterpillarresearch engine equipped with a high- pressure, electronically-controlledfuel injection system. The effect of multicomponent fuel droplet vaporizationis studied, with particular attention to the effects of the injection ambienttemperatures. The vaporization model, besides being able to describe bicomponentfuels, accounts for high pressure effects, variable properlies, and variableLewis number. The model was tested and compared with experimental datafor single droplets, and comparisons between the model results and theavailable data are satisfacl tory. The multicomponent fuel vaporizationmodel was extended to model diesel sprays under typical diesel conditions.Necessary modifications were also carried out in the atomization and collisionsubmodels. The spray results indicate the importance of multicomponentfuel effects in conditions where ambient temperatures are relatively low,such as in cold-starting. The model was successfully applied to enginecases. SAE950455 "In-Cylinder Diesel FlameImaging Compared with Numerical Computations," by S.-C. Kong, L.M. Ricartand R.D. Reitz. An image acquisition-and-processing camera system was developedfor in-cylinder diagnostics of a single-cylinder heavy duty diesel engine.The engine was equipped with an electronicallycontrolled common- rail fuelinjection system that allowed both single and split (multiple) injectionsto be studied. The imaging system uses an endoscope to acquire luminousflame images from the combustion chamber and ensures minimum modificationto the engine geometry. The system also includes an optical linkage, andimage intensifier, a CID camera, a frame grabber, control circuitry anda computer. Experiments include both single and split injection cases at90 MPa and 45 MPa injection pressures at 3/4 load and 1600 rev/ min withsimulated turbocharging. For the single injection at high injection pressure(90 MPa) the results show that the first luminous emissions from the ignitionzone occur very close to the injector decreases, ignition occurs furtherdownstream of the nozzle. When only a small amount of fuel is injected,ignition does not occur until the fuel reaches the piston crown. The secondfuel injection pulse is injected into a high temperature environment containingthe combustion products of the first injection and flame luminosity appearsclose to the injector exit with almost no detectable delay. Experimentalresults are compared to numerical computations using a version of KIVA-IIcode with improved spray, ignition, combustion and emissions models. Goodlevels of agreement were obtained with measured cylinder pressures, heatrelease rates, ignition timings and locations and flame shapes. However,the need for some improvement in the vaporization and mixing models isindicated. The present imaging and modeling results are also used to explorethe reasons for the significant soot and NOX emissions reductions withmultiple/split injections. In this case, a late second injection preventsthe high temperatures that promote NOX formation. The combustion productsof the first injected fuel set up a sufficiently high temperate for therapid ignition of the second injection. The second injection may quenchNOX being formed from the combustion of the first injection and also entrainair into the first combustion region. This helps in the oxidation of soot. SAE 950458 investigation of DieselSprays Using Diffraction-based Droplet Sizing," by M A. Coil and P.V. Farrell.The study of combustion in direct injection Diesel engines demands detailedunderstanding of the behavior of the injection. Understanding the injectioninvolves characterizing the distribution of fuel particle sizes throughoutthe spray. This work studied the size distributions of sprays from commercialDiesel injectors under a series of conditions. A diffraction based diagnosticobtained maps of local fuel droplet size information over the full sprayfield. Most quantitative techniques currently used in spray research providequantitative time-ranging data at a single point in the spray field. Spatiallyresolved information proves more useful in studying transient sprays. Thespatially resolved maps of particle size obtained in this experiment showedthe reliability of the diagnostic, exhibited the transience of the finestructure of these sprays, and demonstrated the evolution of the sprayswith time. SAE 950604 "Effects of Injection Pressureand Nozzle Geometry on D.I. Diesel Emissions and Performance," by D.A.Pierpont and R.D. Reitz. An emissions and performance study was performedto show the effects of injection pressure, nozzle hole inlet condition(sharp and rounded edge) and nozzle included spray angle on particulate,NOX, and BSFC. The tests were conducted on a fully instrumented single-cylinderversion of the Caterpillar 3406 heavy duty engine at 75% and 25% load at1600 RPM. The fuel system consisted of an electronically controlled, hydraulicallyactuated, unit injector capable of injection pressures up to 160 MPa. Particulateversus NOX trade- off curves were generated for each case by varying theinjection timing. The 75% load results showed the expected decrease inparticulate and flattening of the trade-off curve with increased injectionpressure. However, in going from 90 to 160 MPa, the timing had to be retardedto maintain the same NOX level, and this resulted in a 1 to 2% increasein BSFC. The rounded edged nozzles were found to have an increased dischargecoefficientL By adjusting the injection pressure, it was possible to comparethe perfonnance of the rounded and sharp edged nozzles with the same massrate of injection profiles. Interestingly, the sharp edged nozzle gavesignificantly lower particulate emissions and lower B SFC at lower injectionpressures. However, as the injection pressure was increased the differencein particulate became smaller and the rounded edged nobles gave lower BSFC.Two nozzle spray angles with included angles of 125 and 140 degrees werestudied. The effects of spray angle on particulate and NOX emissions wasfound to be small at high load, but differences were seen at light load.These results are interesting because the s pray in the 125 degree caseis directed so as to give significant spray impingement on the piston bowlwall, while the 140 degree nozzle has minimal wall impingement. SAE950814 "Cycle-by-Cycle Variationsin Combustion and Mixture Concentration in the Vicinity of the Spark PlugGap," by K.-H. Lee and D.E. Foster. The correlation's between IMEP andpressures at referenced crank angles have different trends for differentequivalence ratios. A fiber optic spark plug was used to detect the initialflame development which was then used to analyze the combustion cyclicvariation. Rayleigh scattering measurements were applied to detect theair-fuel mixture fluctuations in the vicinity of spark plug gap for bothhomogeneous and in homogeneous mixture preparations in a spark ignitionengine. The variation in mixture concentration in the vicinity of sparkplug gap was not confined as a major contributor to cycle-by-cycle variationin combustion for any of the homogeneous mixture cases or for the stoichiometricand lean mixtures of port injection. However, a leaner mixture (F0.8) ofport injection did correlate with the cyclic variation in combustion. Atomization and Sprays (in press) "Near-NozzleCharacteristics of a Transient Fuel Spray," by J.-Y. Koo and J.K. Martin.The near nozzle characteristics of a transient fuel spray have been investigatedwith the measurement of drop sizes and velocities, and microphotographsof the near-nozzle region for a range of gas-to-liquid density ratios.In addition, a steady-state hydrodynamic simulation of the internal nozzleflow was performed to observe the effects of needle position on the internalflow. Measurement of droplet size and velocity near the nozzle on the edgeof the spray showed that the average droplet velocity peaked during needleopening and needle closing, and changed throughout the spray evenL Dropsizes tended to be small on the spray edge. Microphotographs of the nearnozzleregion showed that the spray was most widely dispersed immediately afterinjection began, narrowing rapidly to a constant spray angle. The samebehavior was observed even for injection into nearvacuum conditions. However,once the spray was established, aerodynamic interactions were necessaryfor near-nozzle atomization. The internal flow calculation showedthat the magnitude of the turbulence intensity in the fluid was relatedto needle position. The potential for cavitation didn't however seem tobe related to needle position. Combustion Science and Technology (inpress) "Coherent Flamelet Modeling of Diesel Engine Combustion," by M.P.Musculus and CJ. Rutland. A turbulent combustion model based on the coherentflamelet model was developed in this study and applied to diesel engines.The important physics involved in each phase of diesel engine combustionwere defined and modeled as directly as possible. The combustion eventwas broken into tluee phases: low temperature ignition kinetics, high temperaturepremixed burn kinetics, and diffusion burn. Two transition steps were developedto model the progression of combustion between each of these phases. Theignition phase was accomplished using the Shell ignition model. The transitionto the high temperature premixed burn kinetics was accomplished using acriteria based on heat release rate and temperature. The high temperaturepremixed burn kinetics were modeled using a global Arrhenius equation forthe rate of reaction. The transition to the diffusion burn was based ona critical Damkohler number. Finally, the coherent flamelet model was usedas a foundation for the diffusion burn portion of the model. The modelwas implemented into the multidimensional computational computer code KIVA-II.The sensitivity of the combustion events to several of the model parameterswas examined and the results were used to pick the optimum settings forthe model. Previous experiments on a Caterpillar model E 300, # 1 Y0540engine, a TACOM Labeco research engine, and a single cylinder version ofa Cummins N14 production engine were used to validate the cylinder averagedpredictions of the model. The characteristics of the modeling approachwere also addressed by examining the spatial resolution of the model resultsinside the engine cylinder. The location and magnitude of model heat releases,flame areas, and equivalence ratios were examined. The results of thisapproach to the modeling of diesel engine combustion could be used to enhanceto modeling of engine emissions. Combustion and Flame (in press) "PremixedFlame Effects on Turbulence and Pressure Related Terms," by S. Zhang andCJ. Rutland. Direct Numerical Simulations (DNS) were carried out for premixed,planar, turbulent flames. Heat release effects are accounted for by inclusionof variable density. The simulated flames are thin in the sense that thereaction progress variable is bi-modal and consistent with BML theory.The DNS data was used for detailed study of flame effects on turbulencewithin the turbulent flame brush by examining the turbulent kinetic energybudget. The flame effects on turbulent kinetic energy were found to dependstrongly on the heat release. Both mean and fluctuating pressure termswere found to be the main factors responsible for increases in turbulentkinetic energy. The main sinks for turbulence are dissipation and meandilatation. Pressure diffusion was found to dominate the other turbulentkinetic energy diffusion terms. A model was developed for pressure dilatationthat matches the DNS results very closely. The model indicates that pressuredilation will remain an important source of turbulence even as heat releaseincreases. SOOT MODEL IN PROGRESS Professor Foster, visiting scientistA. Fuseo of C.N.R., and Post-Doc A. Knox Kelexy have published preliminaryfindings for a phenomenological soot model at the 1994 COMODIA meeting. "Application of a PhenomenologicalSoot Model to Diesel Engine Combustion," by A. Fusco, A.L. Knox-Kelecyand D.E. Foster. A phenomological soot model has been used to simulatesoot production for combustion conditions typical of a diesel engine. Themodel accounts for the number of carbon atoms in the fuel and incorporatesthe physical processes of inception, surface growth, coagulation, and oxidationusing global rate expressions. It consists of four differential equationsbalancing the rates of change of particle number density, soot precursorradicals, acetylene (assumed to be the soot growth species), and soot volumefraction. Arrhenius type global kinetic rate expressions have been writtenfor most of the processes considered. For a few of the processes, suchas coagulation and oxidation, detailed rate expressions of non- Arrheniusfonn have been taken from the literature. The model has been coded in FORTRANand is operational as a stand-alone submodel. It has been used to simulatesoot production using inputs of fuel, oxygen, temperature and pressureprofiles based on results obtained from a KIVA II simulation of a Cumminsdiesel engine. The model has demonstrated the ability to handle a widerange of temperatures in both rich and lean mixtures with predictions thatbehave as expected. Both the shapes and the magnitudes of the profilesfor soot volume fraction, particle number density, soot precursor radicals,and acetylene growth species are predicted to behave consistently withthe trends published in the literature.
1500 Johnson Drive, Room 119 ERB, University of Wisconsin-Madison |
|
File last updated:
September 13, 2006
|