FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'08) Malta, September 11-13, 2008

Transcription

1 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 Injection Timing and Cone angle Behavior on a Heavy duty Diesel Engine M.Gorji-bandpy,D.D.Ganji, S.Soleimani Department of Mechanical Engineering, University of Mazandaran, P.O.Box 484,Babol,Iran zzsajad@yahoo.com Abstract:- In this paper the three dimensional computational fluid dynamics (CFD) analysis have been used to improve understanding of the formation of soot and NO during combustion in a heavy duty diesel engine. Six injection strategies were used as follows: start of injection at 1, 13 and 16 ree before top dead center (BTDC) and injection with cone angle of 16, 12 and 8 ree. The comparison between the injection strategies like rate of heat release pressure and temperature was investigated. The principal results show significant differences in soot and NO generation during combustion between these deferent strategies. Keywords: - Diesel engine, Heavy duty, Injection timing, Injection cone angle, NO, Soot 1 Introduction Despite their superior efficiency and resulting low fuel consumption and CO 2 emissions, diesel engines are notorious for their relatively levels of high emission of particulate matter (PM) and NOx. Optimizing the combustion process (i.e., reducing NOx and PM emissions without serious compromises in fuel consumption and emissions of CO and unburnt hydrocarbons) requires profound knowledge of all formation processes involved. Diesel combustion models need support from numerical and/or experimental data, so threedimensional numerical simulations of diesel combustion are very useful. The possibility of describing precisely the influence of the various volumes on injection permitted the simulation of new systems such as the compact injector-pump units [1]. The use of simplifying hypotheses (neglecting friction, assuming fluid velocity and density constant with pressure) permitted a theory of formulation of the propagation in pipes based on small acoustic perturbation. This is characterized by its simplicity and a sense of physics. Nevertheless, in the new systems the high pressures (more than 1MPa) contradict the hypothesis of a constant sound velocity and call attention to phenomena like blow by losses in the gap between plunger and needle, pipe elasticity, and variation of the flow characteristics with pressure. Afterwards, interest was directed to the spray and the interrelation between its characteristics and the geometric configuration of the system [2-4]. Here, we focus on the different injection strategies, presenting mean NO mass fraction and mean soot mass fraction in a DI diesel engine. These detailed results has allowed for the good judgment in a heavy-duty diesel engine. The rate of heat release, in-cylinder temperature and in-cylinder pressure are also presented for more discussion. 2 Model description Spray simulations involve multi-phase flow phenomena and as such requires the numerical solution of conservation equations for the gas and the liquid phase simultaneously. With respect to the liquid phase, practically all spray calculations in the engineering environment today are based on a statistical method referred to as the discrete droplet method (DDM) [5]. This operates by solving ordinary differential equations for the trajectory, momentum, heat and mass transfer of single droplets, each being a member of a group of identical non-interacting droplets termed a parcel. Thus one member of the group represents the behavior of the complete parcel. Droplet parcels are introduced in the flow domain with initial conditions of position, size, velocity, temperature and number of particles in the parcel. Introduction of droplets is emerging from a nozzle as a spray and entering the flow domain through the inlet areas as a gas/liquid mixture. The atomization process of sprays is accounted for with distinctive submodels. The droplet-gas momentum exchange, turbulent dispersion, evaporation of droplets, secondary break-up, droplet collision and dropletwall interaction are covered with a comprehensive set of models. The vapor of evaporating droplets is used as a source term of an additional transport equation for the vapor void fraction in Eulerian formulation. The droplets are tracked in a Lagrangian way ISSN: ISBN:

2 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 through the computational grid used for solving the gas phase partial differential equations. Full twoway coupling (interaction) between the gas and liquid phases is taken into account. In situations of negligible influence of the dispersed phase on the continuous one, the gas phase flow can be simulated in advance and the droplet simulation can be performed afterwards. Fig.1: Schematically view of injection For this work, computational fluid dynamics simulation is used as the modeling platform and the sub-models of NO and Soot emissions are described briefly. 3 NO and soot emissions NO emissions depend strongly on the history of the heat release rates and the major and minor species concentrations. In this work, NO predictions are determined using the extended Zeldovich mechanism [6-14]. Under high temperature and fuel rich conditions, as typically found in diesel combustion, hydrocarbon fuels exhibit a strong tendency to form carbonaceous particles - soot. Usually, under engine running conditions, most of the soot formed in the early stages of the combustion process is depleted due to oxidation. This takes place in oxygen rich areas of the combustion chamber later in the engine cycle. In diesel engines, it is the amount and completeness of the soot oxidation process that actually determines the engine particle emission characteristics. The formation of particulates involves a large number of different chemical and physical processes, like the formation and growth of large aromatic hydrocarbons, their subsequent conversion to particles, the coagulation of primary particles, and the growth of solid soot particles due to the accumulation of gaseous components [15,16]. The soot particle formation process is characterized by a gaseous-solid conversion, whereby the solid phase does not exhibit a uniform chemical and physical topology. It is evident that the formation of soot, i.e. the conversion of hydrocarbon rich, aliphatic compounds involving only a relatively small number of carbon atoms into an agglomerate comprising millions of them, is the result of a highly complex chemical process involving hundreds of reactions and as many intermediate and radical species. Particle oxidation mainly occurs due to the attack of atomic oxygen onto the carbonaceous particles under high temperature conditions. In spite of the great complexity of the underlying processes, the individual reactions contributing to the soot formation and oxidation rates can be related to known flame parameters, such as fuel mass fraction, partial pressure of oxygen, flame temperature and/or turbulent mixing intensity. Under ideal conditions, the combustion of hydrocarbon fuels forms CO 2 and H 2 O. The necessary amount of oxygen is the stoichiometric oxygen requirement O 2, st calculated from the following equation: C H + m m ( n+ ) O nco + H O n m (3) So for ideal combusting of 1 mole C n H m, m ( n+ ) moles O 2 is required. This ideal amount 4 of oxygen is called O 2, st. The real amount of oxygen, available for the combustion, is expressed by the air access ratio O λ = 2 or by the equivalence ratio φ: O 2, st O2, st 1 ϕ = = λ O2 (31) For conditions at equivalence ratios φ > 1, there is a big potential for soot formation. There are four major processes in soot formation: nucleation, coagulation, surface growth and oxidation [15,16]. Investigation of premixed flames shows that the fuel molecules are split into radicals, ISSN: ISBN:

3 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 mainly acetylene. Subsequently this 2D radical grows by chemical reactions, H abstraction and acetylene addition. This process forms large aromatic rings out of aliphatic species. In further steps the molecules become 3D and form carbonaceous particles by coagulation. Through gaseous-solid conversion, the soot particles grow afterwards (surface growth). A similar process is running in diffusion flames, but highly influenced by the inhomogeneous mixture and turbulent mixing. The most important parameters during the soot formation are the local air/fuel ratio (C/H-ratio and C/O-ratio), temperature, pressure, and residence time. Case Start of injection ( ATDC) R R R C C C2-1 8 Spray cone angle (ree) Figure 2 shows a schematically view of spray cone angle and figure 3 shows the bowl of the combustion chamber in two views. 4 Engine Simulation Details A heavy duty diesel engine combustion chamber with the bowl shape of figure 3 is modeled numerically. The engine geometry is provided in Table 1. The models of wave standard spray breakup model [17], k-ε turbulence model and a specified wall temperature for estimating the heat flux into the combustion chamber were used. To increase computational efficiency, a 6 -sector computational mesh with a single fuel spray 5 injecting kg in every cycle was used to model the six-hole injector. Periodic boundary conditions were assumed in the axial direction. The calculations were performed using a threedimensional sector that consisted of cells at top dead center, corresponding to a maximum cell wall dimension of approximately 2 mm. The simulation conditions considered in the study are listed in Table 2. A range of injection timings and injection cone angle were investigated. Figures 8-12 study the effects of changing the injection timing and figures represent a study of the effects of different injection cone angle while fixing all other engine conditions at constant values. Fig.2: Schematically view of nozzle and spray cone angle, spray cone angle=2α Table 1. Geometry of simulated engine. Bore [cm] 14,92 Stroke [cm] Connecting rod 33 length [cm] Compression ratio 16.5:1 Speed [rpm] 15 Table 2. Engine simulation conditions. Fig.3: Bowl of combustion chamber with its grids in ISSN: ISBN:

4 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 two views. In figure 4 a comparison between two different grid sizes results for (a) in-cylinder temperature, (b) rate of heat release and (c) in-cylinder pressure against various rees of crank angle were first conducted. This was done to ensure that the prediction made with this work is trustworthy. The results indicate that the comparisons between two grids of A (with cells at bottom dead center) and B (with 8322 cells at bottom dead center) are in a good agreement, so the combustion chamber with grid A is used for the calculations. te m p e ra tu re K ra te o f h e a t re le a s e J /d e g (a) Grid A Grid B Grid A Grid B p re s s u re b a r Grid A Grid B (c) Fig. 4: (a) In-cylinder temperature, (b) Rate of heat release and (c) In-cylinder pressure against various rees of crank angle for grid A and grid B. 5 Model Validity Because the target engine is new generation engine and it is under development, another heavy duty diesel engine which is developed by Ref. [18] was used.the specification of reference diesel engine are listed in table 1.Figure 5 shows the differences between reference and present results of in-cylinder pressure. Table3. Specification of reference engine Bore, mm 1 Stroke, mm 115 Compression 18:1 ratio Combustion ω type chamber Nozzle hole.29 diameter, mm Nozzle hole 4 numbers Speed, rpm 2 (b) ISSN: ISBN:

5 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, p r e s s u r e M P a present reference CA,ATDC rate of heat release J/ R R1 R Fig.5: Comparison between reference and present in-cylinder pressure. These results show good agreement between reference and present data and this model is reliable to predict the engine performance near real conditions. 6 Results and discussions There are three injection strategies, injection with different timings of R, R1 and R2 which are as follows, R: start of injection at crank angle of 35 ree, R1: start of injection at crank angle of 347 ree and R2: start of injection at crank angle of 344 ree. Figure 6 indicates the rate of heat release against various rees of crank angle. By beginning of the injection a severe increase in the rate of heat release is observed. This leads to a pressure increase suddenly which is called knock phenomena. As shown in this figure this is the highest amount of all for R1 and the least amount of all for R,so the injection strategy which applies injecting timing of R1suffer from the knock phenomena more than other strategies. Knock phenomena is a sudden pressure increase happens because of sudden increase of the rate of heat release. By continuing the combustion which leads to diffusion combustion phase the rate of heat release for R1 is the highest of all and for R is the least of all. After the end of injection (late combustion phase) the rate of heat release for R is the highest and for R1 is the least of all. Fig. 6: Rate of heat release against various rees of crank angle. Figures 7 and 8 indicate in-cylinder temperature and in-cylinder pressure against various rees of crank angle respectively. In figure 7 it is observed that in-cylinder temperature from the beginning of the injection till crank angle of 4 is the highest of all for R1and the least of all for R.After the crank angle of 4 this relation reverses however the differences of temperature between the cases of R,R1 and R2 is negligible. Regarding to figure 8 which indicate in-cylinder pressure, it is seen that in-cylinder pressure for the case of R1 is the highest of all while in the case of R experiences the least amount of all. As the figure shows in-cylinder pressure for R1 is 15 bar more than R2 and for R2 is 15 bar more than R1 at incylinder peak pressure. Figure 9 shows the produced NO in combustion chamber. The observations of this figure is that, the produced NO for R1 is the highest and for R is the least of all whereas from the crank angle of 39 the produced NO in the case of R is lower than R2 by the amount of.5 and the produced NO in the case of R2 is lower than R1 by the amount of.7. Figure 1 shows the mean soot mass fraction for three cases of R, R1 and R2.From the beginning of injection to the crank angle of 37 the mean soot mass fraction for R1 is the highest of all and for R is the least of all. The mean soot mass fraction for R2 in this period of crank angle is higher than R with only a little difference. But from the crank angle of after 37 the mean soot mass fraction for R is the highest of all and for R1 is the least of all. The mean soot mass fraction of the case of R2 is between R and R1 in this period of crank angle. Overall, the produced soot for R1 and R2 is almost ISSN: ISBN:

6 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 the same in the whole crank angles. Fig. 9: Mean NO mass fraction against various rees of crank angle tem perau ture K R R1 R Fig. 7: Mean in-cylinder temperature against various rees of crank angle. p ressure b ar R R1 R2 Fig. 8: Mean in-cylinder pressure against various rees of crank angle. m ean N O m ass fraction R R1 R m ean soot mass fraction R R1 R2 Fig. 1: Mean soot mass fraction against various rees of crank angle. Figure 11 is the representation of the rate of heat release against various rees of crank angle. At the beginning of injection the sudden increase of the rate of heat release is observed as for the case of C is more than C1 and that is more than C2.But by continuing the injection the rate of heat release is almost the same for all three cases, however it is a little higher in case C2. Figures 12 and 13 are the representations of incylinder temperature and pressure. As figures show those are almost the same for all three cases. r a te o f h e a t r e le a s e C C1 C2 Fig. 11: Rate of heat release against various rees of crank angle. ISSN: ISBN:

7 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 te m p e r a t u r e K C C1 C m e a n N O m a s s f r a c t i o n C C1 C Fig. 12: Mean in-cylinder temperature against various rees of crank angle. Fig. 14: Mean NO mass fraction against various rees of crank angle. p r e s s u re b a r C C1 C2 Fig. 13: Mean in-cylinder pressure against various rees of crank angle. Figure 14 is the indications of the variations of mean NO mass fraction against various rees of crank angle. As shown in the figure the production of NO in case C1, and C2, is more than the case of C by the amount of 12% and 23% respectively. Figure 15 indicates the mean soot mass fraction against various rees of crank angle. As shown in the figure, by decreasing the spray cone angle by the amount of 2 ree and 4 ree, the mean soot mass fraction decreases about 31% and 46% at the peak zone respectively. m e a n s o o t m a s s f r a c t i o n C C1 C2 Fig. 15: Mean soot mass fraction against various rees of crank angle. 7 Conclusions The effect of different spray cone angle and injection timing on the performance and pollution of a heavy duty diesel engine was investigated. The results show that the produced pressure for R1 is the highest and for R is the least of all. The amount of produced NO for R1 is highest and for R is the least of all. But produced soot for the crank angles of before 37 is highest for R1 and least of all for R and R2.However for the crank angles after 37 it is the case of R1 which produce the least mean soot mass fraction of all. Finally, it is obvious that the injection strategy of R is not a proper choice. As the total produced soot for R1 and R2 is almost the same, one should mention to in-cylinder pressure and the amount of NO produced by each cases of R1 and R2 to find the ISSN: ISBN:

8 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, 28 advantages and disadvantages of them for choosing the best injection strategy. By choosing R2 the lower in-cylinder pressure is gained which is with the advantage of lower NO while by choosing R1, the higher in-cylinder pressure is gained with the penalty of higher NO production. It was observed that the different spray cone angle have not any significant effect on the pressure and temperature in the combustion chamber, but by decreasing the spray cone angle the produced NO increased and the produced soot decreased,as, by decreasing the spray cone angle by the amount of 2 and 4 ree the mean soot mass fraction decreased 31% and 46% respectively, so it is obvious that the variations of spray cone angle have significant effect on producing the pollutions. Nomenclature: T NO N 2 O O 2 OH H N CO 2 C n H m O H 2 a Subscripts st m Temperature nitrogen monoxide nitrogen atomic oxygen oxygen hydroxyl radical hydrogen atom atomic nitrogen carbon dioxide hydrocarbon fuel water stoichiometric relations stoichiometric Number of [ K ] n atoms in chemical combination Number of atoms in chemical combination References: 1. Schuen, R. and Hames, R., 'Computer simulation of the GM unit injector', SAE Paper No , Marcic, M. and Kovacic, Z., 'Computer simulation of the diesel fuel injection system', SAE Paper No , Gibson, D. H., 'A flexible fuel injection simulation', SAE Paper No , Kumar, K., Gajendra Babu, M. K., Gaur, R. R. and Garg, R. D., 'A finite difference scheme for the simulation of a fuel injection system', SAE Paper No , Dukowicz, J.K. "A Particle-Fluid Numerical Model for Liquid Sprays", J. Comp.Physics, 35,198,pp Jones, W.P. and Lindstedt, R.P. "Global Reaction Schemes for Hydrocarbon Combustion." Combustion and Flame 73,1988,pp Lam, S. H. "Reduced Chemistry Modeling and Sensitivity Analysis". Lecture Notes for Aerothermochemistry for Hypersonic Technology, Lecture Series Program, at the VonKarman Institute For Fluid Dynamics, April 24-28, Brussels, Belgium, Maas, U. and Pope, S. B. "Simplifying Chemical Kinetics; Intrinsic low-dimensional Manifolds in Composition Space." Combustion and Flame 88,1992,pp Bowman, C. T. "Chemistry of Gaseous Pollutant Formation and Destruction, in Fossil Fuel Combustion. A Source Book,1991,pp Hanson, R.U. and Salimian, S. "Survey of Rate Constants in the N/H/O-System." Combustion Chemistry. William C. Gardiner Jr. Springer Verlag, 1984,pp Polifke, W. "Fundamental and Practical Limitations of NOx, Reduction in Lean-Premixed Combustion", Notes for the Euro Conference RWTH about "Premixed Turbulent Combustion: Introduction of the State of the ART." Aachen, Germany, June 8-9, 1995,pp Zeldovich, Y. B., Sadovnikov, P. Y. and Frank- Kamenetskii, D. A. Oxidation of Nitrogen in Combustion. Translation by M. Shelef, Academy of Sciences of USSR, Institute of Chemical Physics, Moscow-Leningrad, ISSN: ISBN:

9 FINITE DIFFERENCES - FINITE ELEMENTS - FINITE VOLUMES - BOUNDARY ELEMENTS (F-and-B'8) Malta, September 11-13, Bogensperger, M. "A Comparative Study of Different Calculation Approaches for the Numerical Simulation of Thermal NO Formation." Diss. U. Graz, Heywood, J. B. Internal Combustion Engine Fundamentals. McGrawHill Book Company,Second Series, Bockhorn, H. "Soot Formation in Combustion: Mechanisms and Models.", Springer, Bockhorn, H., Fetting, F., Heddrich, A. and Wannemacher, G. "Investigation of the Surface Growth of Soot in Flat Low Pressure Hydrocarbon Oxygen Flames." Twentieth International Symposium on Combustion. Pittsburgh: The Combustion Institute, 1985,pp Liu, A.B. and Reitz, R.D. "Modeling the Effects of Drop Drag and Break-up on Fuel Sprays", SAE Yanfeng,G., Shenghua,L., Hejun,G., Tiegang,H. and Longbao,Z., "A new diesel oxygenate additive and its effects on engine combustion and emissions",applied Thermal Engineering, 27,27,pp ISSN: ISBN: