Hydrogen homogeneous charge compression ignition (HCCI) engine with DME as an ignition promoter

Hydrogen homogeneous charge compression ignition (HCCI) engine with DME as an ignition promoter J. Jeon, H. Yoon, C. Bae Korea Advanced Institute of Science and Technology, Korea ABSTRACT Hydrogen combustion with dimethyl-ether (DME) was investigated in a single cylinder engine with a homogeneous charge compression ignition strategy. The hydrogen and DME injection quantities were varied respectively with a fixed sum of the total heat release value of the fuels. The injection timing of the DME was varied from -35 crank angle degrees to the top dead centre. The combustion temperature was lowered, and the emissions of carbon monoxide and hydrocarbon were increased as the DME injection timing was advanced. When the DME was injected near the top dead centre, carbon monoxide and hydrocarbon emissions were decreased while the nitric oxides emission was increased. 1 INTRODUCTION A homogeneous charge compression ignition (HCCI) engine is a new engine concept for the future. Its main features are the ability to breathe a premixed charge, as in spark ignition (SI) engines, and to ignite without a spark, as in compression ignition (CI) engines. With HCCI strategies, lean burn combustion is possible and the combustion temperature is lower than that of conventional SI and CI engines (1). Low temperature combustion reduces the emission of nitrogen oxide (NOx) and soot. However, the HCCI concept has many problems, such as the emission of large amounts of hydrocarbon (HC) and carbon monoxide (CO) and a limited operational range due to knocking from a high rise in pressure. Other researchers have investigated ways of reducing

the knocking effect by controlling the combustion phase with the use of high-octane number fuel and high-cetane number fuel (1-4). Hydrogen is considered to be the most promising alternative fuel for the next-generation energy source due to its carbon-free emissions and excellent combustion characteristics. The use of hydrogen in an internal combustion engine is less expensive than hydrogen fuel-cell vehicles. Many researchers have investigated the application of hydrogen in SI engines (5-7). As a high-octane number fuel, hydrogen can be easily applied to a conventional SI engine. However, hydrogen SI engines have problems such as backfiring and knocking (8). A hydrogen engine with CI operation has great advantages, particularly high efficiency, variety in engine sizes, and reduced abnormal combustion (9). In spite of these advantages, the application of hydrogen to CI engines is fraught with difficulty due to the high auto-ignition temperature (9). An ignition-assist fuel is required to achieve stable hydrogen CI operation. DME is a promising substitute fuel for diesel. It enables the use of the developed liquefied petroleum (LPG) gas infrastructure because its physical properties are similar to those of LPG. It also has good auto-ignitability owing to its low auto-ignition temperature (10). Furthermore, because it has a low boiling point, DME can be instantaneously vaporized after being injected. The presence of oxygen in the DME molecule also means there is no C- C bond; hence, soot emission is negligible (11). However, because DME s enthalpy of formation is lower than that of diesel, a greater volume of DME needs to be injected. In addition, the lower viscosity of DME leads to more wear of moving parts; and the sealing material must be carefully selected on account of DME s rubber corrosiveness characteristics. Note also that DME has a two-stage ignition involving a low-temperature reaction (LTR) and a high-temperature reaction (HTR). The DME LTR includes several chain reaction steps and creates radicals such as OH, HCHO, and H 2 O 2 (12). Moreover, formaldehyde has been shown to have an ignition-promoting effect on hydrogen combustion (9). For this study, hydrogen was applied to a HCCI engine and, for stable hydrogen ignition, DME was injected as an ignition promoter. The way the quantity and timing of the DME injection affects a hydrogen HCCI engine without a spark-assist was investigated. 2 EXPERIMENT 2.1 Experimental apparatus The test engine is a single-cylinder direct CI engine with a common-rail fuel injection system and a port fuel injection system. The cylinder has a bore of 83 mm and a stroke of 92 mm. The

displacement is 498 cm 3 for a cylinder with a compression ratio of 14.8. The engine speed was controlled by a direct current (DC) dynamometer (37 kw). Figure 1 shows a schematic of the experimental setup, and Table 1 gives the engine specifications. Figure 1. Experimental setup of single cylinder direct injection diesel engine Table 1. Engine specifications Research engine type Single-cylinder direct injection compression ignition engine Valves per cylinder 4 (2 intake and 2 exhaust) Bore x Stroke 83 mm x 92 mm Connecting rod length 145 mm Displacement per cylinder 498 cm 3 Intake valve timing Open at 23 BTDC / Close at 65 ABDC Exhaust valve timing Open at 75 BBDC / Close at 45 ATDC The hydrogen was injected at the intake manifold at a pressure of 0.5 MPa with a modified commercial CNG injector (Bosch). An injection controller (Zenobalti, IDU 5000B) controlled the injection by changing the main parameters, namely the injection pressure, the timing, and the duration. The DME was supplied through a commercial common-rail diesel injection system modified for DME. DME was supplied from a fuel tank at 2 MPa to maintain the liquid phase; it was then pressurized to 30 MPa by a pneumatic pump (Haskel Inc., MS-188). The pressurized DME was supplied to the common-rail injection system. The test injector was a solenoidtype commercial diesel injector (Bosch) with seven holes. The injector was driven by a solenoid injector driver for a diesel common-rail fuel injection system (TEMS, 3200H). For combustion analysis, the in-cylinder pressure was measured with a piezoelectric pressure transducer (Kistler, 6052A) and the intake pressure was measured with a piezo-resistive pressure

sensor (Kistler, 4045A5). The exhaust emissions were measured with an exhaust gas analyzer (HORIBA; MEXA 1500D). A data acquisition board (Iotech, Wavebook516E) was used to measure data for 100 consecutive cycles at increments with a crank angle degree (CAD) of 0.2. Table 2. Engine operating conditions Fuel Intake air temperature Coolant temperature Engine speed Injection pressure H2 DME(with lubricity additive, 500 ppm) 303 ± 1 K 353 ± 2 K 1200 rpm H2 0.5MPa DME 30MPa Table 3. Fuel injection timing and quantity Fuel Hydrogen DME 5 mg Injection adjusted to total injection 7 mg quantity quantity of 1000J 9 mg T0 T5 Injection timing -360 CAD ATDC (Intake TDC) T10 T20 T30 T35 TDC -5 CAD ATDC -10 CAD ATDC -20 CAD ATDC -30 CAD ATDC -35 CAD ATDC 2.2 Operating conditions The engine operated at 1200 rpm. The coolant temperature was set to 353K±2K and the intake temperature was kept at 303K±1K. Table 2 shows the operating conditions of the engine. Hydrogen was used as the main fuel and DME was used to assist the ignition of hydrogen. The viscosity of the DME was enhanced by the addition of 500 ppm of a lubricity additive (Infineum R655). The sum total injection quantity of each fuel was fixed at 1000 J, and the quantity of injected DME varied from 5 mg to 9 mg. The timing of the hydrogen injection was fixed at the intake top dead centre (TDC) and the injection pressure is 0.5 MPa. The timing of the DME injection varied from -35 CAD ATDC to the TDC; and the injection pressure was fixed at 30 MPa. The timing of the DME injection was denoted as T0 to T35. Table 3 shows the timing and quantity of the fuel injection. The heat release rate (HRR) was calculated from an ideal gas equation, and the in-cylinder pressure on the basis of constant gas

properties (13). The Woschini model for a direct injection diesel engine was used to determine the heat transfer to the cylinder wall (14). 3 RESULTS AND DISCUSSIONS 3.1 Effect of DME injection timing The aim of applying the DME injection was to improve the hydrogen ignition. The DME injection was expected to promote the ignition quality of the hydrogen HCCI combustion. The DME injection timing was varied to determine the optimal point of the hydrogen combustion. 3.1.1 In-cylinder pressure, Heat release rate and indicated mean effective pressure Figure 2(a) shows how the DME injection timing affects the incylinder pressure. The DME injection quantity was fixed to 9 mg and the injection timing was varied from T0 to T35. The hydrogen injection quantity was adjusted so that the sum of each injected fuel quantity was 1000 J; it was then injected at the intake TDC. The in-cylinder pressure and HRR curves could be divided into two domains: the HCCI combustion domain and the conventional CI combustion domain. The HCCI combustion domain, from T35 to T20, shows the LTR and the HTR (15-16). In this domain, the incylinder pressure was increased and the HRR curve was sharper; furthermore, the starting time of the LTR and HTR was delayed on account of the delay in the DME injection. An advanced DME injection produced a leaner and homogeneous mixture, whereas a retarded DME injection produced a locally rich mixture. Because of the small quantity of DME, the relatively rich region of the fuel reacted faster than the lean homogeneous region; hence, after the DME auto-ignition, the hydrogen was oxidized more under higher temperature conditions. When HCCI combustion occurs as a result of an early DME injection, a lower temperature can reduce the HRR curve. The retarded DME injection, from T10 to T0, produced only a HTR. Thus, the LTR and HTR occurred at the same time due to the relatively high temperature and pressure, thereby causing a DME auto-ignition condition near the TDC. However, a delay in the DME injection near the TDC led to a lower in-cylinder pressure and HRR curve; it also caused combustion during the expansion stroke and led to a smoother and lower in-cylinder pressure and HRR curve. Figure 2(b) shows the indicated mean effective pressure (IMEP) in relation to the DME injection timing. The IMEP improved when DME was injected at T30 to T5. With DME injections at T35 and T0, the IMEP was low due to inefficient heat release of the DME.

60 Incyl. Pressure[bar] HRR[J/deg] 40 20 60 40 20 DM E injection tim ing T0 T5 T10 IMEP T20 T30 T35 motor -40-35 -30-25 -20-15 -10-5 0 5 2.2 2.1 2.0 1.9 1.8 1.7 CAD IMEP [bar] 0 340 350 360 370 380 390 Crank angle degree (a) DM E injection tim ing [CAD] (b) Figure 1. Effect of DME injection timing on in-cylinder pressure, HRR(a), and IMEP(b), 9mg DME injection, 1000J. Figure 3. Effect of DME injection timing on emissions, 9 mg DME injection, 1000J 3.1.2 Emissions Figure 3 shows how the DME injection timing affects engine emissions. The NOx emission and CO 2 emission had a similar tendency. More NOx was emitted when the peak HRR reached its highest value; that is, at T10. The region that is rich in DME due to a late injection tended to create a high combustion temperature, which in turn led to higher NOx emissions. However, relatively lean homogeneous mixture that was created with advanced DME injections produced a lower temperature combustion and a negligible NOx emission. The CO 2 emission increased when the

DME injection is retarded. Conceivably, more DME was completely oxidized with the retarded DME injection than with the advanced DME injection. HC emissions increased with advanced DME injections. Quenching at the crevice volume and cylinder wall was one of the main sources of HC (17). With advanced DME injections, there was an increase of HC emissions (up to 400 ppm) in spite of the small amount of DME. This phenomenon occurred for two reasons: the lower combustion temperature from the lean homogenous charge and the greater quenching effect of the crevice volume and wall wetting. When the DME injection was retarded near the TDC, the HC emission increased slightly again. The retarded DME injection caused the start of the combustion after the TDC and lower incylinder pressure and temperature were created during the expansion stroke. As a result, the HC content was increased as a result of the incomplete combustion. The characteristics of the CO emission curve and the HC emission curve were similar. When the DME injection was delayed, the CO emissions decreased dramatically but increased slightly near the TDC. 3.2 Effect of DME injection quantity The way the DME quantity affects hydrogen ignition was investigated in this chapter. The quantity of injected DME varied from 5 mg to 9 mg, and the quantity of injected hydrogen was adjusted to the fixed total heat release. 3.2.1 In-cylinder pressure, Heat release rate and indicated mean effective pressure Figure 4 shows how the quantity of injected DME affects the incylinder pressure, the HRR, and the IMEP. Only P35 and P10 were shown, and the total injection quantity was fixed in the figure. The injected DME was first oxidized and then the released heat helped oxidize the hydrogen. The more DME injected, the greater the LTR heat. The heat-assisted hydrogen ignition and the start of the DME LTR produced higher values for the in-cylinder pressure, the HRR curve, and the IMEP. When the DME injection occurred earlier and in small quantities, the in-cylinder pressure increased and the HRR curve was almost zero. Specifically, the injection of 5 mg and 7 mg of DME at T35 produced almost zero work. However, the injection of 9 mg of DME at T35 produced much better results, particularly in terms of the LTR, the HTR, and the IMEP. An overly lean mixture was created when a small amount of DME was injected at an early stage; the mixture not only ignited the DME itself but also assisted the hydrogen oxidation.

Incyl. Pressure[bar] HRR[J/deg] 60 50 40 30 20 60 40 20 0 DME Injection quantity and tim ging 5m g T20 7m g T20 9m g T20 5m g T10 7m g T10 9m g T10 m otor 340 350 360 370 380 390 CAD [crank angle degree] (a) 5m g 7m g 9m g -40-35 -30-25 -20-15 -10-5 0 5 DME injection timing[cad] (b) Figure 4. Effect of DME injection quantity on in-cylinder pressure, HRR(a), and IMEP(b), at 1000J. 2.0 1.5 1.0 0.5 0.0 IMEP [bar] CAD 250 NOx Emssions with NOx [ppm] THC [ppm C1] 200 150 100 50 0 3000 2000 1000 DME injection quantity 5mg 7mg 9mg THC Emissions with DME injection quantty 5mg 7mg 9mg CO2 Emissions with DME injection quantity 5mg 7mg 9mg CO Emissions with DME injection quantity 5mg 7mg 9mg 1.5 1.0 0.5 4000 3000 2000 1000 CO [ppm] CO2 [%] 0-35 -30-25 -20-15 -10-5 DME injection tim ing [CAD] -35-30 -25-20 -15-10 -5 DME injection tim ing [CAD] 0 Figure 5. Effect of DME injection quantity on emissions, 1000J 3.2.2 Emissions Figure 5 shows the emission tendencies in relation to various quantities of injected DME. The NOx and CO 2 emissions increased with larger quantities of injected DME. The heat released by the DME assisted the combustion of the remaining DME and hydrogen. As a result, the higher combustion temperature yielded higher emission levels of NOx and CO 2. In spite of the increase in carbon, the HC and CO emissions decreased when the quantity of injected

DME was increased. Moreover, the decreased HC and CO emissions increased the CO 2 emission. 4 CONCLUSIONS This study investigated a hydrogen-fueled HCCI strategy in which DME was used as an ignition promoter in a single-cylinder CI engine. The engine was equipped with a common rail injection system for DME injection, and hydrogen as injected at the intake manifold. The engine speed was 1200 rpm. In addition, the total injected heat release was fixed whereas the hydrogen and DME injection quantities were varied. The hydrogen was injected at - 360 CAD, and the DME injection timings were varied from -35 CAD to 0 CAD. The results are summarized as follows: DME had potential for promoting hydrogen HCCI combustion. Advanced injection of DME yielded the premixed combustion characteristics of high-octane number fuel. The IMEP, the rise of the in-cylinder pressure, and the HRR were low due to the formation of a lean homogeneous mixture. The fact that a retarded injection yielded higher values for the IMEP, the pressure rise, and the HRR could be attributed to a partially rich region which helped DME and hydrogen combustion. Larger quantities of injected DME yielded lower emission levels of CO and HC and higher emission levels of NOx and CO 2. The IMEP was improved and the larger quantities of injected DME released more heat for hydrogen ignition. ACKNOWLEDGEMENT This study was sponsored by the office of KAIST EEWS institute (EEWS: Energy, environment, water, and sustainability) REFERENCES 1. K. Yeom, J. Jang, and C. Bae, Homogeneous charge compression ignition of LPG and gasoline using variable valve timing in an engine, Fuel, Vol. 86, No. 4, pp. 494 503 (2007). 2. K. Yeom and C. Bae, Gasoline di-methyl ether homogeneous charge compression ignition engine, Energy and fuels, Vol. 21, No. 4, pp.1942~1949, (2007). 3. Y. Narioka, T. Yokoyama, S. lio, and Y. Takagi, HCCI combustion characteristics of hydrogen and hydrogen-rich natural gas reformate supported by DME supplement, SAE paper 2006-01-0628 (2006). 4. T. Miyamoto, B. Kobayshi, M. Mikami, N. Kojima, H. Kabashima, Y. Shimasaki, Exhaust emission characteristics of

a diesel engine with small amounts of hydrogen added to the intake air, COMODIA, FL1-1 (2008). 5. X. Tang, D. M. Kabat, R. J. Natkin, and W. F. Stockhausen, Ford P2000 hydrogen engine dynamometer development, SAE paper 2002-01-0242 (2002). 6. S. Kaiser and C. M. White, PIV and PLIF to evaluate mixture formation in a direct injection hydrogen fuelled engine SAE paper 2008-01-1034 (2008). 7. W. Thiel and K. Hartmann, Equations and Methods for Testing Hydrogen Fuel Consumption using Exhaust Emissions SAE paper 2008-01-1036 (2008). 8. S. Verhelst, J. D. Landtsheere, F. D. smet, C. Billiouw, A. Trenson, and R. Sierens, Effects of superchargning, EGR and variable valve timing on power emissions of hydrogen internal combustion engines, SAE paper 2008-01-1033 (2008). 9. T. Imuta, R. Hata, T. Tsujimura, Y. Tokunaga, J. Senda, and H. Fujimoto, Combustion method of diesel engine fuelled with hydrogen for high efficiency, COMODIA, FL1-2 (2008). 10. C. Arcoumanis, C. Bae, R. Crookes, and E. Kinoshita, The potential of di-methyl ether (DME) as an alternative fuel for compression-ignition engines: A review, Fuel, Vol. 87, No. 7, pp. 1014-1030 (2008). 11. T. Kitamura, Detailed chemical kinetic modelling of smokeless diesel combustion with oxygenated fuels, Engine Technology, pp.64-71, 6, 5 12. H.J. Curran, W. J. Pitz, C.K. Westbrook, P. Dagaut, J-C Boettner and M. Cathonnet, A wide range modeling study of dimethyl ether oxidation, International Journal Chemical Kinetics, 30-3, pp.229-241, 1998 13. J. B. Heywood, Internal combustion engine fundamentals, McGraw-Hill (1998). 14. G. Woschni, A universally applicable equation for the instantaneous heat transfer coefficient in the internal combustion engine, SAE Paper 670931 (1967). 15. S. S. Nathan, J. M. Mallikarjuna, and A. Ramesh, Effect of mixture preparation in an diesel HCCI engine using early incylinder injection during the suction stroke International journal of automotive technology, Vol. 8, NO. 5, pp.543-553 (2007) 16. G. S. Jung, Y. H. Sung, B. C. Choi, and M. T. Lim, Effect of mixture stratification on HCCI combustion of DME in a rapid compression and expansion machine, International journal of automotive technology, Vol. 10, NO. 1, pp.1-7 (2009). 17. M. Sjöberg, and J. E. Dec, "Smoothing HCCI heat-release rates using partial fuel stratification with two-stage ignition fuels," SAE Paper 2006-01-0629 (2006).