Mechanism of Oxygen Concentration Effects on Combustion Process

Oct 5, 2009 - *To whom correspondence should be addressed. Phone: 86-23-65102473. E-mail: [email protected]. Cite this:Energy Fuels ...
0 downloads 0 Views 3MB Size
Energy Fuels 2009, 23, 5835–5845 Published on Web 10/05/2009

: DOI:10.1021/ef900676h

Mechanism of Oxygen Concentration Effects on Combustion Process and Emissions of Diesel Engine Zhaolei Zheng*,† and Mingfa Yao‡ †

College of Power Engineering, Chongqing University, Chongqing, 400030, China, and ‡State Key Laboratory of Engines, Tianjin University, Tianjin, 300072, China Received July 1, 2009. Revised Manuscript Received September 8, 2009

The mechanism of oxygen concentration effects on the combustion process of diesel combustion and the effects of injection pressure and intake pressure on low-temperature combustion are investigated by CFD simulation. The oxygen concentration is modulated by CO2. The results indicate: with the decrease of oxygen concentration, the peak of the average in-cylinder pressure decreases and the ignition delay becomes long. Decreasing oxygen concentration is the most effective method to control NOx emissions. With the decrease of oxygen concentration, soot emissions increase first and then decrease. It is the essential reason of low soot emissions at low oxygen concentrations that the low in-cylinder combustion temperature leads to the inhibition of soot formation. With the increase of injection pressure, the flow velocity and turbulent kinetic energy in the cylinder increase, which are beneficial to improving the mixing of fuel and air; soot emissions decrease; and NOx emissions increase. The enhancement of intake pressure improves the condition of oxygen lack at lower oxygen concentration. With the increase of intake pressure, ignition delay becomes long. Soot and NOx emissions decrease.

and air has an important role on the HCCI combustion process. In fact, it is impossible to get an absolutely homogeneous mixture in the operation of a practical HCCI engine.4 A little inhomogeneity in fuel concentration and temperature in mixing can produce significant effects on autoignition and combustion processes. Utilizing the inhomogeneity (charge stratification) is an important path to achieve clean and highefficiency combustion in engines.5-8 It can be concluded that stratification combustion has the potential to extend the highload limits for HCCI operation and reducing UHC and CO emissions. Although the stratification strategy has the potential to extend HCCI operation range, the rapid increase of NOx emissions limits operating stratified HCCI engines at high loads. A new combustion mode is needed for clean and highefficiency combustion of diesel engines at high loads. In fact, the amount of both soot and NOx emissions from a diesel engine is dependent on local temperature and mixture in the cylinder. Alriksson and Denbratt calculated the φ-T map for soot and NO concentrations using the Senkin code with a surrogate diesel fuel consisting of n-heptane and toluene.9 The

1. Introduction The diesel engine has a considerable advantage in regards to engine power, fuel economy, and durability compared with other types of internal combustion engines. Unfortunately, the diesel engine is also a main source of particle matter (PM) and nitrogen oxides (NOx) emissions, both of which are subject to legislative limits because of their adverse effects on the environment and human health. To meet the requirements of future legislation, researchers across the world have been seeking methods to achieve clean and high-efficiency combustion. Post-treatment technology of exhaust gas is an effective approach to decrease NOx and soot emissions simultaneously, but there are many disadvantages such as cost, assemblage, security, shortage of noble metal resources, etc. With deeper and deeper understanding of the combustion process, new combustion modes based on improvement of combustion process have received more and more attention in recent years. It is well-known that homogeneous charge compression ignition (HCCI) combustion has the potential to simultaneously reduce NOx and particulate emissions, while achieving high thermal efficiency at partial load.1-3 However, HCCI engines tend to have high unburned hydrocarbon (UHC) and carbon monoxide (CO) emissions, high rates of heat release, and narrow operation ranges. After a great deal of HCCI research, researchers realized that the mixing process of fuel

(4) Dec, J. E.; Hwang, W.; Sj€ oberg, M. An Investigation of Thermal Stratification in HCCI Engines Using Chemiluminescence Imaging. SAE Tech. Pap. Ser. 2006-01-1518; 2006. (5) Thirouard, B.; Cherel, J.; Knop, V. Investigation of Mixture Quality Effect on CAI Combustion. SAE Tech. Pap. Ser. 2005-01-0141; 2005. (6) Dec, J. E.; Sj€ oberg, M. Isolating the Effects of Fuel Chemistry on Combustion Phasing in an HCCI Engine and the Potential of Fuel Stratification for Ignition Control. SAE Tech. Pap. Ser. 2004-01-0557; 2004. (7) Sj€ oberg, M.; Dec, J. E. Smoothing HCCI Heat-Release Rates Using Partial Fuel Stratification with Two-Stage Ignition Fuels. SAE Tech. Pap. Ser. 2006-01-0629; 2006. (8) Zheng, Z. L; Yao, M. F Energy Fuels 2007, 21 (4), 2018–2026. (9) Alriksson, M.; Denbratt, I. Low Temperature Combustion in a Heavy Duty Diesel Engine Using High Levels of EGR. SAE Tech. Pap. Ser. 2006-01-0075; 2006.

*To whom correspondence should be addressed. Phone: 86-2365102473. E-mail: [email protected]. (1) Onishi, S.; Jo, S. H.; Shoda, K.; Jo, P. D.; Kato, S. Active thermoatmosphere combustion: A new combustion progress for internal combustion engines. SAE Tech. Pap. Ser. 790501; 1979. (2) Thring, R. H. Homogeneous-Charge Compression Ignition (HCCI) Engines; SAE Tech. Pap. Ser. 892068; 1989. (3) Lshibashi, Y. Basic Understanding of Activated Radical Combustion and Its Two-Stroke Engine Application and Benefits. SAE Tech. Pap. Ser. 2000-01-1836; 2000. r 2009 American Chemical Society

5835

pubs.acs.org/EF

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

results suggest that the local combustion temperature should be kept below approximately 2200 K to avoid high NO concentrations for low equivalence ratios. The maximum soot concentration can be found in intermediate flame temperature and high equivalence ratio regions, which are ideal for both the formation of polycyclic aromatic hydrocarbons (PAHs) and their transformation to soot particles. In low temperature flames, both NOx and soot emissions are very low. In conventional diesel combustion engines, one way to reduce the temperature is to use exhaust gas recirculation (EGR). Unfortunately, the reductions in NOx emissions cause an increase in particulate emissions. If large amounts of cooled EGR can be used, the combustion temperature can be reduced to low enough to simultaneously suppress NOx and soot formation. It is rather remarkable that advanced control strategies are necessary if large amounts of cooled EGR are used. This concept is referred to as EGR-diluted low temperature combustion (LTC). Simultaneous soot and NOx reduction is observed as EGR rates are increased to very high levels (approximately 60% or greater).10,11 The possibilities of using low temperature combustion have also been studied by Alriksson et al.12 At 25% engine load, the use of very high EGR levels resulted in low soot and NOx emissions. However, at these operating conditions, fuel consumption and emissions of CO and HC were increased. In the following research, they investigated the possibilities for extending the range of engine loads in which soot and NOx emissions can be minimized by using LTC.9 Very low levels of both soot and NOx emissions can be achieved at engine loads up to 50% by reducing the compression ratio to 14 and applying high levels of EGR (up to approximately 60%). Therefore, EGR-diluted LTC mode has the potential to extend the operation range to higher loads by proper control strategies, including optimization of fuel injection strategy, high injection pressure, proper swirl and boost, etc. Ogawa et al. also obtained simultaneous soot and NOx reductions by using 55% EGR while fuel injection timings were adjusted such that combustion started at top dead center (TDC).13 Hardy et al. proposed that sufficient mixing time (interval between the end of injection and the start of combustion) was necessary to achieve low soot and NOx emissions.14 Their study also used high EGR (70%) to properly phase combustion with very early injection timings (60° before top dead center, BTDC). Furthermore, they used various multiple injection configurations at high loads and high-speed operation for a heavy-duty diesel engine.15 Henein et al. examined the effects of injection pressure and the swirl motion on engine-out emissions over a wide range of EGR

Table 1. Submodels Used in the Calculation turbulence model spray models

ignition model combustion model NOx model soot model

RNG k-ε nozzle atomization droplet breakup droplet wall interaction shell autoignition LATCT (laminar-and-turbulent characteristic time) EBU extend Zeldovich mechanism I. Magnusson

MPI-2 Huh Reitz/ Diwakar Bai

rates from the conventional to the low-temperature combustion regime.16,17 At the high EGR rates in the two regimes, soot can be reduced by applying high injection pressures and a moderate increase in the swirl ratio. Aoyagi et al. investigated studied LTC by inducting much air into the cylinder and the high EGR rate up to 30-40% at the full load of the engine using high boost and turbo intercooled technologies.18 It was proved experimentally that the both NOx and PM emissions reduce effectively with little fuel consumption penalty by the wide range, high boost, and cooled EGR system. From the researches mentioned above, the control strategies improving the mixing of fuel and air are of great benefit to overcoming the disadvantages of LTC, such as low combustion efficiency and high CO emissions. This study focused on the mechanism of low-temperature combustion. The mechanism of oxygen concentration effects on combustion and emission of diesel engine was investigated to explain the characteristic of LTC by simulation study. Furthermore, the mechanism of injection pressure and intake pressure effects on LTC was studied to overcome the disadvantages of LTC. 2. Computational Model 2.1. Computational Model. The CFD code STAR-CD was utilized to show the complex physical and chemical processes involved in diesel spray combustion process. The submodels used in the calculation are shown in Table 1. The product mechanism of NOx in combustion consists of thermal, prompt, and fuel NOx formation. In the combustion of diesel engines, NO is dominant in the components of NOx emissions. Its formation obeys extended Zeldovich mechanism, strongly temperature-dependent, which contains the following three reactions:19,20 k1

N2 þ O S NO þ N

ð1Þ

N þ O2 S NO þ O

k2

ð2Þ

k3

ð3Þ

k -1

k -2

N þ OH S NO þ H k -3

(10) Akihama, K.; Takatori, Y.; Inagaki, K.; Sasaki, S.; Dean, A. M. Mechanism of the Smokeless Rich Diesel Combustion by Reducing Temperature. SAE Tech. Pap. Ser. 2001-01-0655; 2001. (11) Weissback, M.; Csat o, J.; Glensvig, M.; Sams, T.; Herzog, P. MTZ 2003, 64, 718–727. (12) Alriksson, M.; Rente, T.; Denbratt, I. Low Soot, Low NOx in a Heavy Duty Diesel Engine Using High Levels of EGR. SAE Tech. Pap. Ser. 2005-01-3836; 2005. (13) Ogawa, H.; Miyamoto, N.; Shimizu, H.; Kido, S. Characteristics of Diesel Combustion in Low Oxygen Mixtures with Ultra-High EGR. SAE Tech. Pap. Ser. 2006-01-1147; 2006. (14) Hardy, W. L.; Reitz, R. D. A Study of the Effects of High EGR, High Equivalence Ratio, and Mixing Time on Emissions Levels in a Heavy-Duty Diesel Engine for PCCI Combustion. SAE Tech. Pap. Ser. 2006-01-0026; 2006. (15) Hardy, W. L.; Reitz, R., D.An Experimental Investigation of Partially Premixed Combustion Strategies Using Multiple Injections in a Heavy-Duty Diesel Engine. SAE Tech. Pap. Ser. 2006-01-0917; 2006.

(16) Henein, N. A.; Bhattacharyya, B.; Schipper, J.; Kastury, A.; Bryzik, W. Effect of Injection Pressure and Swirl Motion on Diesel Engine-out Emissions in Conventional and Advanced Combustion Regimes. SAE Tech. Pap. Ser. 2006-01-0076; 2006. (17) Natti, K. C. PM Characterization in an HSDI Diesel Engine under Conventional and LTC Regimes. SAE Tech. Pap. Ser. 2008-01-1086; 2008. (18) Aoyagi, Y.; Osada, H.; Misawa, M.; Goto, Y.; Ishii, H. Advanced Diesel Combustion Using of Wide Range, High boosted and Cooled EGR System by Single cylinder engine. SAE Tech. Pap. Ser. 2006-01-0077; 2006. (19) Aberg, P.; Smith, G. P; Jeffries, J. B. Proc. Combust. Inst. 1998, 27, 1377–1384. (20) Monat, J. P.; Hanson, R. K.; Kruger, C. H. Proc. Combust. Inst. 1979, 17, 543–552.

5836

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao Table 2. Engine and Injector Specifications bore stroke connecting rod length compression ratio combustion chamber shape intake valve closure timing number of nozzle holes nozzle hole diameter included spray angle

105 mm 125 mm 210 mm 17.5:1 ω -137° ATDC 7 0.17 mm 155°

The computations are started at the time of intake valve closure (IVC = -137° after top dead center, ATDC). The average in-cylinder pressure and temperature at the beginning of the calculation are determined by running a cycle analysis program. Initial turbulent kinetic energy at the timing of intake valve closure is 74% of the square of the mean piston speed, whereas the initial turbulence length scale is 7% of the cylinder bore.27 The surface temperatures of the piston, cylinder head, and linear are estimated as 525, 500, and 475 K, respectively. The computations use n-dodecane as the fuel because its properties are diesel fuel’s. As for computational time, it takes about 6 h for one case. 2.2. Comparisons between Experiment and Calculation Results. Experiments were conducted on a modified singlecylinder, 4-valve, four-stroke-cycle, water-cooled diesel engine equipped with a common rail fuel injection system. Table 2 shows the engine specifications. The schematic diagram of engine setup and experimental approach were introduced in ref 28. CO2 is used to simulate EGR modulating oxygen concentration. An oxygen sensor is installed on the intake pipe. By adjusting the valve opening, intake mass of CO2 can be changed, so the intake oxygen concentration is changed. Fuel injection is controlled by an electric control system. The parameters of injection mass, injection timing, common rail pressure, and so on can be flexibly adjusted. Figure 2 shows the comparison of in-cylinder pressure and heat release rate between experiment and simulation at the injection pressure of 1400 bar and natural aspiration condition. The injection timings are -20 and -25° ATDC, respectively. It can be seen from Figure 2 that the calculated in-cylinder pressures agree well with the measured results. For heat release rate, peaks of the calculated results are a little higher than those of the measured results, however the starts and trends of heat release agree well between the calculated and measured results. Figure 3a shows the comparison of NOx emissions between experiment and simulation at different injection pressures. It can be seen that the trends of calculated results agree well with the measured results. Figure 3b shows the comparison of soot production and consumption process between experiment and simulation. The experimental data were measured by total cylinder dumping system in a diesel engine with a bore diameter of 102 mm. The configuration and principle of total cylinder dumping system and experimental approach were introduced in ref 29. The experiments cited here were carried out at an engine speed of 1000 rpm, an injection pressure of 900 bar, and an injection timing of -8° ATDC. The fuel delivery per cycle is 45 mg. It can be seen from Figure 3b that the calculated soot production and oxidation trends agree well with the measured results.

Figure 1. Engine combustion chamber geometry and computational mesh.

where k1, k2, and k3 are rate constants; and k-1, k-2, and k-3 are the corresponding reverse rate constants. Soot model is primarily based on that of Magnusson.21 This model tries to describe the interaction between the formation and oxidation of soot, which is modeled by the flamelet library approach in the framework of prescribed probability density functions (PDFs). Two additional scalars are to be solved for, namely, the soot mass fraction and the variance. The numerical solution of the Navier-Stokes equations becomes more and more accurate if the grid is refined. Abraham22 has shown that the jet cross-sectional area has to be resolved by at least four grid cells near the nozzle in the case of a gas jet being injected into a gas atmosphere. On the other hand, because it is impractical to follow each individual drop inside a spray, the combination of Monte Carlo method and stochastic parcel technique is used in order to reduce the number of individual drops, the behavior of which has to be directly calculated. The more parcels are used, the better the behavior of the dispersed liquid phase is resolved, and the better the statistical convergence. If the grid size is too small, the parcels in the cell are not enough to ensure numerical stabilities. Therefore, appropriate grid size should meet numerical accuracies and stabilities. On the basis of the results of a mass of computation, it is summarized that a grid size between 1 and 2 mm and a time step of 0.1 °CA can obtain good numerical accuracies and stabilities in the condition of bore diameter and engine speed of current diesel engines. Furthermore, the computational time can be acceptable. The validation of submodels in Table 1 and the selection of model parameters could be found in the literatures.23,24 Figure 1 shows the engine combustion chamber geometry and computational mesh. To reduce the computation time, only a sector of 51° was used in the simulation with single injection in the 7-hole injector. There are 33 524 cells in the computational mesh. Because of the inherent restriction of DDM spray model, the simulations are sensitive to mesh refinement. The applied mesh resolution has therefore been chosen such that it is close to the spray model optimized mesh resolution suggested by Baumgarten25 and Abraham et al.22,26 (21) Karlsson, A.; Magnusson, I.; Balthasar, M. Simulation of soot formation under Diesel engine conditions using a detailed kinetic soot model. SAE Tech. Pap. Ser. 981022; 1998. (22) Abraham, J. What is Adequate Resolution in the Numerical Computations of Transient Jets? SAE Tech. Pap. Ser. 970051; 1997. (23) Zhang X. Y. Research on Interaction of Physical and Chemical Factors in MULINBUMP Compound Combustion. Ph.D. Dissertation, Tianjin University: 2007 (in Chinese). (24) CD adapco Group, Methodology, STAR-CD, Version 3.24; 2004 (25) Baumgarten, C., Mixture Formation in Internal Combustion Engine; Springer-Verlag: New York, 2006. (26) Abraham, J.; Magi, V. A Virtual Liquid Source (VLS) Model for Vaporizing Diesel Sprays, SAE Tech. Pap. Ser. 1999-01-0911; 1999.

3. Results and Discussion In the present study, CO2 is used to simulate EGR modulating oxygen concentration. Engine speed is 1400 rpm. Fuel (27) Kidoguchi, Y.; Sanda, M.; Miwa, K. J. Eng. Gas Turbines Power 2003, 125, 351–357. (28) Yao, M. F.; Zhang, Q. C.; Zheng, Z. Q.; Zhang, P. Sci China Ser., E 2009, 52 (6), 1527–1534. (29) Pei, Y. Q.; Dong, S. R.; Song, C. L. Combust. Sci. Technol. 2006, 12 (2), 115–120.

5837

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 2. Comparison of in-cylinder pressure and heat release rate between experiment and simulation.

Figure 3. Comparison of NOx and soot emissions at different injection pressures.

rapidly with the drop of oxygen concentration, and NOx emissions are very low when the oxygen concentration is below 15%. In addition, NOx emissions significantly increase with the advancing of the injection timing. The local amplified figure indicates the oxygen concentration can decrease to 13% at the injection timings of -20 and -25° ATDC, while misfire occurs when the oxygen concentration is below 16% at injection timing of -5° ATDC. That is to say, earlier injection timing can make the combustion process suffer from higher EGR rates (lower oxygen concentration), then obtain lower NOx emissions. Therefore, the following studies in the present study are all carried out at an earlier injection timing of -20° ATDC.

delivery per cycle per cylinder is 50 mg, and initial overall equivalence ratio at 21% O2 is 0.58. 3.1. Mechanism of Oxygen Concentration Effects on Combustion Processes and Emissions. In this part, the effects of oxygen concentration on combustion processes were investigated. The engine was operated in the condition of injection pressure of 1400 bar and intake pressure of 1.0 bar. The identical injection profile was used in the simulation cases of different oxygen concentrations, and it was measured from experiment. Figure 4 shows NOx emissions at different oxygen concentrations and different injection timings from the experiment studies.28 It can be seen that NOx emissions decrease 5838

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 5 shows the calculated in-cylinder pressures and temperatures at different oxygen concentrations. The fall of oxygen concentration is equivalent to the rise of CO2 in the mixture, which leads to the increase of specific heat capacity. Then, the in-cylinder temperatures decrease. Therefore, the peak of average in-cylinder pressure decrease and the ignition delay becomes long (the ignition delay is defined as the time between the onset of fuel injection and the pressure jump caused by combustion). On one hand, the decrease of in-cylinder temperature inhibits fuel evaporation. On the other hand, the increase of ignition delay provides more time for the evaporation of fuel droplets. The competition results of these two aspects are that the evaporated percentage of fuel before ignition increases with the decrease of oxygen concentration as shown in Figure 6. That is to say, the improvement of long ignition delay in fuel evaporation is more significant than the inhibition of low in-cylinder temperature in fuel evaporation. Figure 7 shows calculated in-cylinder distribution of air/fuel mixture concentration before ignition at different oxygen concentrations. It can be seen that the premixed combustible mixture increases with the decrease of oxygen concentration because of the improvement of fuel evaporation (both the concentration of the mixture and the range of higher concentration increase with the decrease of oxygen concentration). The more the premixed combustible mixture is, the stronger the premixed combustion is. This can explain why the heat release proportion of premixed combustion increases with the decrease of oxygen concentration as shown in the experimental studies.28 The experimental studies show that NOx emissions rapidly decrease with the decrease of oxygen concentration. At lower

oxygen concentrations, NOx emissions also increase with the advancing of the injection timing, but the increasing degree is greatly low compared with higher oxygen concentration. At oxygen concentration below 16%, the NOx emissions are close to zero for all injection timings.28 As mentioned in introduction, investigation has shown that NOx emissions rapidly increase when the temperature exceeds 2200 K, and the maximum soot concentration can be found at intermediate flame temperatures.9 Figure 8a presents the calculated formation processes of NOx emissions at different oxygen concentrations. NOx emissions are quickly produced when the in-cylinder temperature increases to some degree. After combustion, NOx emissions get to the maximal values and almost keep constant. When the oxygen concentration is relatively higher (20.5-17.5% O2), NOx emissions rapidly decrease with the decrease of oxygen concentration; when the oxygen concentration is relatively lower, NOx emissions are all very low, and the decrease of oxygen concentration has a little effect on them. The trends of NOx emissions can be explained by Figure 9, which shows the calculated incylinder temperature distributions when the maximal local temperature appears at different oxygen concentrations. The in-cylinder temperature obviously decreases with the decrease of oxygen concentration. When the oxygen concentration is relatively higher (20.5-17.5% O2), the decrease of the maximal temperature and the shrinkage of high-temperature regions lead to the rapid decrease of NOx emissions. At lower oxygen concentration conditions, the highest temperatures are all lower than 2200 K. Therefore, NOx emissions are all very low and the decrease of oxygen concentration has a little effect on them.

Figure 4. Measured NOx emissions at different oxygen concentrations and different injection timings.28

Figure 6. Calculated evaporated fuel percentage before ignition at different oxygen concentrations.

Figure 5. Calculated in-cylinder pressures and temperatures at different oxygen concentrations (injection timing of -20° ATDC).

5839

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 7. Calculated in-cylinder distribution of mixture concentration before ignition at different oxygen concentrations.

Figure 8. Calculated formation processes of emissions at different oxygen concentrations.

Figure 9. Calculated in-cylinder temperature distributions when the maximal local temperature appears at different oxygen concentrations.

The experimental studies show that soot emissions increase first and then decrease with the decrease of oxygen concentration.28 Figure 8b presents calculated production and oxidation processes of soot emissions at different oxygen concentrations. It can be seen that soot emissions are formed by the competition of production and oxidation processes

both in conventional diesel combustion (20.5%O2) and in low-temperature combustion (low oxygen concentration). At lower oxygen concentration (such as 15 and 13% O2 in Figure 8b), both the peaks and the oxidation rates of soot are smaller than those at 20.5% O2. At 15% O2, lower incylinder temperature leads to smaller peak of soot (Figure 9). 5840

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 10. Calculated φ-T distributions when the maximal local temperature appears at different oxygen concentration.

sions, the high in-cylinder temperature leads to very much fuel entering NOx formation regions. When the oxygen concentration decreases to 17.5%, the in-cylinder temperature decreases and the fuel entering NOx formation regions reduces. When the oxygen concentration decreases to 15%, combustion process avoids NOx formation regions, but still crosses soot formation regions. This is because the low in-cylinder temperature decreases the oxidation rate of soot as analyzed in the previous paragraph. When the oxygen concentration decreases to 13%, combustion can completely avoid NOx and soot generation regions. In addition, at this oxygen concentration, UHC and CO emissions rapidly increase because low in-cylinder temperature regions leads to more fuel entering CO/UHC formation regions. 3.2. Effects of Injection Pressure on Combustion and Emissions at Lower Oxygen Concentration. In a conventional diesel combustion engine, the combustion process is dominated by the mixing rates of fuel and air. The needed energy of mixing comes from the kinetic energy of injected fuel and air. When the injected fuel mass is fixed, the increase of injection pressure can lead to the increase of fuel flow rates, then the mixing rates of fuel and air can be improved. In a low oxygen concentration, a large proportion of exhaust gases exist in the cylinder instead of fresh air. Therefore, the mixing of fuel and air becomes more important. For different injection pressures, the injection profiles are different and the fuel deliveries per cycle are identical.

However, lower in-cylinder temperature also brings the low oxidation rate, which leads to the higher final value of soot than that at 20.5% O2. At 13% O2, very little production results in the lower final value of soot than that at 20.5% O2, although the oxidation rate continues to reduce. Therefore, it is the essential reason of low soot emissions at low oxygen concentrations that the low incylinder combustion temperature leads to the inhibition of soot formation. In addition, CO emissions increase rapidly with the decrease of oxygen concentration, and high CO emissions are one of the challenges of LTC as shown in the experimental studies.28 CO emissions have great coherence with in-cylinder temperature. With the decrease of oxygen concentration, in-cylinder temperature decrease (Figure 9), then CO emissions increase. For HC emissions, the long ignition delay leads to the improvement of the mixing of fuel and air (Figure 7) from 20.5 to 14% O2, Therefore, HC emissions decrease first. At very low oxygen concentration (13% O2 in the present study), HC emissions rapidly increase because the temperature is too low (Figure 9). These results are consistent with the experimental results that HC emissions decrease first, but rapidly increase at very low oxygen concentration.28 Figure 10 shows the calculated φ-T distributions when the maximal local temperature appears at different oxygen concentration. At 20.5% O2, there is much fuel located in NOx and soot generation regions. Especially for NOx emis5841

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 11 shows calculated in-cylinder pressure at different injection pressures. It can be seen that the peak of pressure increases with the enhancement of injection pressure. Furthermore, the pressure rise rate slightly increases with the enhancement of injection pressure. In addition, Figure 11 also shows injection pressure almost has little effect on ignition delay.

Figure 12 shows the calculated evaporated fuel percentage before ignition at different injection pressures. The increase of injection pressure leads to the increase of evaporated fuel percentage before ignition. Figure 13 shows comparison of in-cylinder flow field of different injection pressures when the maximal velocity appears. With the increase of injection pressure, the flow velocity in the cylinder increases. Figure 14 indicates that turbulent kinetic energy in the cylinder also increases with the increase of injection pressure. All of the above factors are of benefit to improving the mixing of fuel and air, which improves the combustion process. It is because the improvement of the mixing of fuel and air that the increase injection pressure leads to the increase of incylinder pressure (as shown in Figure 11) and heat release rate (as shown in the experimental studies28). Figure 15a presents the calculated formation processes of NOx emissions at different injection pressures at lower oxygen concentration (15% O2). It can be found that the formation of NOx almost starts at the same time at three injection pressures. In fact, NOx emissions still increase with the increase of injection pressure at lower oxygen concentration. As mentioned in the Introduction, high NOx emissions appear in the regions of high temperature (above 2200 K) and lower mixture concentration. Because the mixing of fuel and air improves, the maximal mixture concentration becomes lower and the concentration difference between lean and rich mixture regions decreases. Furthermore, the in-cylinder temperature becomes higher because of the improvement of combustion. These factors lead to the increase of NOx emissions. Figure 15b presents the calculated production and oxidation processes of soot emissions at different injection pressures at lower oxygen concentration (15% O2). It can be seen that both the peak and final values of soot greatly decrease with the increase of injection pressure. This is also because of the improvement of the mixing of fuel and air. Figure 16 presents the calculated concentration distributions before ignition at three injection pressures. The higher the injection

Figure 11. Calculated in-cylinder pressure at different injection pressures (injection timing of -20° ATDC).

Figure 12. Calculated evaporated fuel percentage before ignition (14 and 15% O2).

Figure 13. Calculated in-cylinder flow field at different injection pressures when the maximal velocity appears (15% O2).

Figure 14. Calculated turbulent kinetic energy at different injection pressures when the maximal value appears (15% O2).

5842

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 15. Calculated formation processes of emissions at different injection pressures (15% O2).

Figure 16. Calculated concentration distributions before ignition at three injection pressures (15% O2).

pressure is, the lower maximal concentration the mixture has. The fuel located in higher-concentration regions becomes less at higher injection pressure in the combustion process. Therefore, high injection pressure is the most effective measure in the reducing of soot emissions. In addition, the improvement of the mixing of fuel and air then improving the combustion process is of benefit to decreasing UHC/CO emissions. The final effects of injection pressure on emissions are consistent with experimental results. 3.3. Effects of Intake Pressure on Combustion and Emissions at Lower Oxygen Concentration. For low oxygen concentration conditions, large amounts of exhaust gases exist in the cylinder instead of fresh air. The greatly decreased fresh air leads to the deterioration of combustion, especially for high loads. Therefore, providing enough fresh air by increasing intake pressure is an effective measure to improve combustion at high loads and low oxygen concentrations. Figure 17 shows the calculated in-cylinder pressure at different injection pressures. The injection pressures of these conditions are all 1600 bar. With the increase of intake pressure, the peak of in-cylinder pressure rapidly increases. Therefore, reducing compression and improving mechanical strength of the engine are necessary when the high level EGR and high intake pressure are simultaneously used in LTC. With the increase of intake pressure, the increase of fresh air leads to the decrease of fuel/air equivalence ratio. Figure 18 shows the calculated concentration distributions at three intake pressures when the maximal concentration appears at lower oxygen concentration (15% O2). It can be seen that the maximal concentration of fuel air mixture drops with the enhancement of intake pressure. Therefore, the

Figure 17. Calculated in-cylinder pressure at different injection pressures (injection timing of -20° ATDC).

enhancement of intake pressure improves the condition of oxygen lack in the cylinder at lower oxygen concentration. Appropriate increase of fresh air can improve the mixing of fuel and air. However, too low fuel/air equivalence ratio will deteriorate combustion. This can explain why the heat release rate is the highest at the intake pressure of 1.5 bar as shown in the experimental studies.28 In addition, Figure 17 also indicates ignition delay becomes long with the increase of the intake pressure. The combustion theory demonstrates: at the same condition, higher in-cylinder temperature can lead to higher rate and kinetic energy of molecule, which can increase the chemical reaction rate and reduce ignition delay; higher pressure can lead to higher molecule density, therefore the distance of molecule reduces then collision frequency enhances, which can also increase the chemical reaction rate and reduce ignition delay. In the present study, both the inhibiting role from the 5843

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 18. Calculated concentration distributions at three intake pressures when the maximal concentration appears (15% O2).

Figure 19. Calculated in-cylinder temperature distributions at three intake pressures when the maximal temperature appears (15% O2).

Figure 20. Calculated formation processes of emissions at different intake pressures (15% O2).

decrease of in-cylinder temperature (Figure 19) and the promoting role from the increase of intake pressure exist simultaneously. Therefore, the change of ignition delay depends on the competition of these two factors. Obviously, the effects of temperature are greater in the present study. Experimental studies show that NOx emissions decrease with the increase of intake pressure.28 While at lower oxygen concentration, intake pressure has little effect on NOx emissions at lower oxygen concentration. Figure 19 shows the calculated in-cylinder temperature distributions at three intake pressures when the maximal temperature appears (15% O2). The maximal temperature temperatures decrease with the increase of intake pressure, but they are all lower than 2200 K at three intake pressures. Therefore, NOx emissions are all very low at lower oxygen concentrations. Figure 20a presents the calculated formation processes of NOx emissions at different intake pressures at lower oxygen concentration (15% O2). NOx emissions still decrease with the increase of intake pressure for the decrease of the maximal local temperature.

Soot emissions also decrease with increase of intake pressure. At higher intake pressure, it is because the improvement of mixing that less fuel enters into soot generation regions (Figure 18). Then, soot emissions decrease. Figure 20b presents the calculated production and oxidation processes of soot emissions at different intake pressures at lower oxygen concentration (15% O2). It can be found that not only the final value, but also the peak value, of soot decreases with the increase of intake pressure. Figure 21 shows the calculated φ-T distributions when the maximal local temperature appears at different intake pressures (15% O2). At three intake pressures, the combustion process is all far away from NOx generation regions. With the increase of intake pressure, the fuel located in high equivalence ratio regions (φ-T) becomes less and less, which is of benefit to the decrease of soot emission. In addition, the comparison of the three figures also indicates that the combustion process is the farthest away from the UHC/CO formation regions at the intake pressure of 1.5 bar, which is benefit to decrease UHC/CO emissions. The final effects of intake pressure on emissions are consistent with experimental results.28 5844

Energy Fuels 2009, 23, 5835–5845

: DOI:10.1021/ef900676h

Zheng and Yao

Figure 21. Calculated φ-T distributions when the maximal local temperature appears at different intake pressures (15% O2).

flow velocity, turbulent kinetic energy in the cylinder, and the evaporated fuel percentage before ignition increase, which are of benefit to improving the mixing of fuel and air. Therefore, the heat release rate slightly increases, NOx emissions increase, and soot emissions significantly decrease. (4) With the increase of intake pressure: ignition delay becomes long because the decrease of fuel/air equivalence ratio. Soot emissions greatly decrease because less fuel located in rich fuel concentration regions. The decrease of in-cylinder temperature leads to the decrease of NOx emissions with the increase of intake pressure.

4. Conclusions In the current study, a simulation study is adopted to investigate the mechanism of low-temperature combustion and the effects of injection pressure and intake pressure on low-temperature combustion. The results can be summarized as follows: (1) With the decrease of oxygen concentration, the in-cylinder temperature decreases, which then leads to the decrease of the peak of average in-cylinder pressure and the increase of ignition delay. The improvement of long ignition delay in fuel evaporation is more significant than the inhibition of low in-cylinder temperature in fuel evaporation. Therefore, premixed combustible mixture increases and the proportion of heat release from premixed combustion increases. (2) Decreasing oxygen concentration is the most effective method to control NOx emissions. With the decrease of oxygen concentration, soot emissions increase first and then decrease. Soot emissions are formed by the competition of production and oxidation processes at either higher or lower oxygen concentration. It is the essential reason of low soot emissions at low oxygen concentrations that the low incylinder combustion temperature leads to the inhibition of soot formation. (3) With the increase of injection pressure, the

Abbreviations PM = particle matter UHC = unburned hydrocarbon CO = carbon monoxide PAHs = polycyclic aromatic hydrocarbons LTC = low temperature combustion Acknowledgment. The research is supported by State Key Laboratory of Engines (SKLE 200901).

5845