Jurnal Teknologi NUMERICAL STUDY OF HYDROGEN FUEL COMBUSTION IN COMPRESSION IGNITION ENGINE UNDER ARGON-OXYGEN ATMOSPHERE.

Transcription

1 Jurnal Teknologi NUMERICAL STUDY OF HYDROGEN FUEL COMBUSTION IN COMPRESSION IGNITION ENGINE UNDER ARGON-OXYGEN ATMOSPHERE Nik Muhammad Hafiz a, Mohd Radzi Abu Mansor a, b*, Wan Mohd Faizal Wan Mahmood a,b, Fadzli Ibrahim a, Shahrir Abdullah a,b, Kamaruzzaman Sopian a Department of Mehanial and Materials Engineering, Faulty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, UKM Bangi, Selangor, Malaysia b Centre for Automotive Researh, Faulty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, UKM Bangi, Selangor, Malaysia Solar Energy Researh Institute (SERI), Universiti Kebangsaan Malaysia, UKM Bangi, Selangor, Malaysia Full Paper Artile history Reeived 18 Deember 2015 Reeived in revised form 10 Marh 2016 Aepted 25 April 2016 *Corresponding author radzi@ukm.edu.my Abstrat Gas emissions from automobiles are one of the major auses of air pollution in our environment today. In fat, emissions of arbon dioxide (CO2), a produt of omplete ombustion, has beome a signifiant fator of the global warming effet. Hydrogen, whih is a renewable energy, is regarded as a promising energy to solve this problem sine the final produt of hydrogen (H2) ombustion, is water (H2O). However, the reation of hydrogen fuels in the air under high temperature onditions produes a high volume of harmful nitrogen oxide (NOx). Furthermore, the high auto-ignition temperature of H2 makes it diffiult to ignite in a ompression ignition engine in normal air. In this researh, argon (Ar) is used to replae nitrogen (N2), in order to eliminate NOx and enhane ombustion. Simulation for this researh was onduted using Converge, omputational fluid dynamis software that is based on Yanmar TF90M ompression ignition engine parameters. The simulation proess was initially onduted with normal air (N2-O2) as the medium of ombustion; but later it was replaed with an argon-oxygen (Ar-O2) atmosphere to investigate the ignition possibility of hydrogen fuel. Hydrogen was injeted at 9.95 MPa at the start of injetion (SOI) at 18º BTDC. The results show that, by employing the same parameters for both simulations in normal air and argon-oxygen mediums, the ombustion of hydrogen only ourred in the argon-oxygen medium. However, no ombustion took plae in normal air. It is therefore onluded that an argon-oxygen medium is appliable for diret hydrogen injetion in a ompression ignition engine. Keywords: Energy effiient vehile, hydrogen, argon, CFD, diret injetion ompression ignition Abstrak Pengeluaran gas daripada ekzos kenderaan merupakan puna utama kepada penemaran alam sekitar. Malah, hasil daripada pembakaran lengkap di dalam enjin juga turut mengeluarkan gas CO2 yang diklasifikasikan sebagai gas rumah hijau. Gas ini akan memberi kesan kepada pemanasan global. Hydrogen, sebagai puna tenaga yang boleh diperbaharui dianggap sebagai penyelesaian kepada masalah ini berdasarkan hasil utama pembakaran adalah wap air (H2O). Walau bagaimanapun, tindak balas pembakaran gas hidrogen di dalam udara turut menghasilkan gas yang berbahaya iaitu nitrogen oksida (NOx). Selain itu, sifat hidrogen yang mempunyai suhu auto-penuuhan yang tinggi menyukarkan pembakaran gas ini di dalam enjin pembakaran dalaman jenis penuuhan mampatan. Di dalam kajian ini, gas argon (Ar) dipilih untuk menggantikan gas Nitrogen (N2) di dalam udara bagi menyingkirkan NOx dan sebagai pemangkin pembakaran di dalam enjin. Converge, perisian dinamik bendalir komputeran telah digunakan dalam kajian ini berlandaskan kepada parameter enjin penuuhan mampatan Yanmar TF90M. Bagi mengkaji nyalaan bahan api hidrogen, simulasi pembakaran hidrogen dijalankan dengan menggunakan udara (N2-O2) sebagai medium dan kemudiannya ditukar kepada argon-oksigen (Ar-O2). Di dalam simulasi ini, gas hidrogen disuntik pada tekanan 9.95 MPa pada 18 BTDC. Keputusan kajian menunjukkan pembakaran gas hidrogen berlaku di dalam medium argon-oksigen manakala tiada pembakaran berlaku di dalam medium udara. Dapat disimpulkan bahawa dengan 78: 6 10 (2016) eissn

2 78 Nik Muhammad Hafiz et al. / Jurnal Teknologi (Sienes & Engineering) 78: 6 10 (2016) menggantikan medium pembakaran kepada argon-oksigen, pembakaran gas hidrogen seara penuuhan mampatan boleh dilaksanakan di dalam enjin. Kata kuni: Kenderaan ekap tenaga, hidrogen, argon, CFD, penuuhan mampatan 2016 Penerbit UTM Press. All rights reserved 1.0 INTRODUCTION The transportation industry depends primarily on petroleum fuels. Worldwide, a huge number of engines burn a vast amount of fuel. Indeed, 70% of the million barrels of rude oil that we onsume every day is used for internal ombustion engines [1]. Reiproating internal ombustion engines are used in a wide range of sizes; from small handheld model engines, to giganti marine engines whih an stand as high as four-storey building. Consequently, researh on engines (both petrol and diesel) has been performed, and small improvements in their effiieny have had major positive impats on eonomy and pollution. Important tasks related to petroleum fuels mainly fous on emissions. Petroleum ontains mostly hydroarbon hains, whih, during the burning proess, produe various exhaust pollutants that inlude nitrogen oxides (NOx), arbon monoxide (CO), hydroarbons (HC), partiulate matter (PM), among other produts. Even in a omplete ombustion produt, CO2 is believed to ontribute to the Greenhouse Effet or global warming. This has led to environmental health impliations; and thus, most governments have imposed stringent vehile emission regulations that are ontinually being tightened. In line with this issue, the International Energy Ageny road map s vision is to redue worldwide fuel use per kilometre by 30 50% in new road vehiles by 2030 and in all vehiles by The goal is to limit the global average temperature rise, whih some limatologists projet for 2050 [2]. In Malaysia, the government s target is to beome the regional hub for Energy Effiient Vehiles (EEV) through strategi investments and the adaptation of high tehnology for the domesti market; and to penetrate regional and global markets by Based on global praties, EEV is defined as vehiles that meet a set of speifiations in terms of arbon emission level (CO2/km) and fuel onsumption (L/km). EEV inludes fuel-effiient internal ombustion engine (ICE) vehiles, hybrid, eletri vehiles (EV) and alternative fuelled vehiles; suh as ompressed natural gas (CNG), liquefied petroleum gas (LPG), biodiesel, ethanol, hydrogen and fuel ell [3]. Further than these immediate issues, fossil fuels are a finite energy soure that annot be renewed. As time passes, these rude oil and natural gas reserves will be depleted. Huge efforts have been made to improve drilling tehnologies, whih will ertainly help to inrease fossil fuel reserves. However, fossil fuels do not appear to be a sustainable option at the urrent rapid rate that they are being onsumed. Muh work has been done in the searh for alternative fuel soures for powering vehiles and oping up with urrent regulatory demands. For this reason, renewable hydrogen energy is regarded as a promising energy storage form for vehiles. This ultimate arbon-free fuel has high energy ontent per unit mass, is easily ombustible due to low ignition energy, and has water as its major produt; thus making it one of the best potential fuels. Future demands for power units with high thermal effiieny and low pollutant emission levels have motivated the researher to develop hydrogen fuelled reiproating internal ombustion engines. In line with this researh, most efforts have been given to the types of engine spark ignition. It is widely aepted that an unthrottled lean-burn operation and redued NOx emissions in the exhaust may be attainable. However, a ommon problem with the use of hydrogen and air in an internal ombustion engine is the low amount of heat produed per unit volume. For example, in a stoihiometri mixture, the heat released per unit volume for a hydrogen-air mixture is only approximately 83% of that for a petrol-air system. A solution to this is to use diret injetion fuel gas into the ylinder. This eliminates the loss in sution, and improves the volumetri effiieny, thus inreasing the power output. It also provides additional advantages, in other respets, by removing the abnormalities of hydrogen ombustion [4-5]. However, the spark ignition engine of pure premixed type an only serve as a ompat power unit; it is not feasible for the requirements of larger output power, like heavy mahines, ships, and other stationary power supplies that are normally worked by diesel engines [5]. In order to ahieve high thermal and volumetri engine effiieny, inreased power output while simultaneously being able to eliminate ombustion abnormalities, ompression ignition (CI) engines operating with hydrogen fuel are strongly reognized [5]. An engine with the apability to operate in high ompression ratios will result in a higher power output than a spark ignition (SI) engine. However, the high auto-ignition of hydrogen fuel requires a high ompression ratio to burn. This high ompression ratio also leads to a higher ombustion temperature and enourages the formation of nitrogen oxide (NOx)

3 79 Nik Muhammad Hafiz et al. / Jurnal Teknologi (Sienes & Engineering) 78: 6 10 (2016) emissions [6-7]. Thus, to avoid suh problems, a monoatomi gas, argon, is suggested to replae the air, in order to inrease in-ylinder temperature for ignition purposes and simultaneously solve the NOx problem; sine the major produt of this reation is water [8]. Argon was seleted beause it is abundantly available, easy to obtain, and failitates the reation of gas tight seals [7]. In engine appliations, the storage and portability of an adequate mass of hydrogen for pratial appliations remains one of the most diffiult problems to overome. In this researh, argon with low heat apaity (Cp), is used to replae nitrogen in normal air, with the purpose of eliminating NOx and enhaning the ombustion of hydrogen that was identified with high auto-ignition temperature. The researh was onduted using Converge simulation software based on Yanmar TF90M ompression ignition engine. The first simulation used air as the ombustion medium and the next used an argonoxygen medium. Fuel injetion pressure 9.95 MPa Figure 1 Workflow 2.0 METHODOLOGY For this artile, Converge CFD software, based on Yanmar TF90M ompression ignition engine parameters, was used to investigate the ombustion proess. The software ame with Adaptive Mesh Refinement (AMR) that ould be applied to the veloity field or any salar field to automatially inrease the resolution, when or where needed. The maximum number of ells ould be speified by the user in order to maintain reasonable runtimes. The AMR algorithm prioritizes and adds resolution where it is needed the most [9]. Table 1 Engine speifiations for Yanmar TF90M The simulation runs on a diret injetion ompression ignition (DICI) engine, with an injetion pressure and start of injetion (SOI) of diesel fuel fixed at 9.95 MPa and 18 BTDC, respetively. Table 1 shows the speifiation details for the base model engine. Figure 1 shows the workflow of the study simulation. At the beginning, the engine geometry, whih was designed using 3D CAD software, was exported to Converge for the ase setup proess. Next, the simulation was run with different initial onditions to obtain the results from two different parameters, namely air and argon-oxygen mediums. Detailed simulation parameters for eah simulation set-up are shown in Table 2. Parameter Speifiations Engine type Horizontal Displaement l Bore x stroke 85 x 87 mm Compression ratio 18 Injetion Timing 18 CA BTDC Table 2 Simulation parameters for region and initialization in ase set-up Simulation Conditions Region Names Temperature [K] Pressure [kpa] Speies (Mass Fration) Cold Flow Air Cylinder O2 23% N2 77% Cold Flow Ar-O2 Cylinder O2 23% Ar 77% H2 injetion in Air Cylinder O2 23% Ar 77% H2 injetion in Ar-O2 Cylinder O2 23% Ar 77%

4 80 Nik Muhammad Hafiz et al. / Jurnal Teknologi (Sienes & Engineering) 78: 6 10 (2016) Gas Simulation Redlih-Kwong equation of state was seleted for the gas simulation model. Momentum and mass transport an be solved for both ompressible and inompressible flows. For ompressible flows, an equation of state is also required to ouple density, pressure and temperature. The equation for this model is shown in Equation (1). RT P v b v b v v , ; where v T P RT p 2 v p a T is ritial is the ritial is the ritial α is attrative fores between moleules, β is volumeof ; r 2 a bv volume, temperature, pressure, 2.2 Combustion Modelling the moleules SAGE ombustion model, whih is inluded in Converge, was seleted for this researh. The model has the apability to solve detailed hemial kinetis during the ombustion proess. SAGE an also determine kinetially-limited phenomena, inluding knok and emissions levels in ombustion. This ombustion model reads in a hemial reation mehanism in a Chemkin format, and solves the ODEs to identify the ell reation rate. In order to speed up the Converge software alulation, SAGE omes with options, suh as solving temperature, whih an be exluded from the hemistry solver (assumed onstant), and also allow the solver to pass an analytially-alulated Jaobian proess. 2.3 Turbulene Modelling In this researh, Reynolds-Averaged Navier Stokes (RANS) was seleted as the turbulene model. This model s approah is to solve the averaged quantities of turbulent motion. This approah does not require large omputing resoures and has been the bakbone of industrial CFD appliations for several deades, due to its modest omputing requirements [10-11]. (1) 3.0 RESULTS AND DISCUSSIONS In this researh, argon was studied to replae nitrogen in normal air for hydrogen fuel ombustion. A great deal of researh has been arried out that foused on spark ignition engine [5,7,12]. However, the potential of effiieny, power density and safety advantages of a diret injetion hydrogen fuelled diesel engine, over a premixed spark-ignited, provided the motivation to further study the ompression ignition engine. 3.1 Validation Of Simulation Results In order to obtain aurate results, the simulation was validated against theoretially thermodynami results. Sine the simulation was onduted from the ondition of intake valve lose (IVC) to exhaust valve open (EVO), the temperature and pressure obtained during ompression and expansion proesses an be alulated using equation (2). Where, T is the temperature, V is the volume and κ is the speifi heat ratio. The ombustion proess ourred under the isentropi ompression of an ideal gas, where speifi heats were onstant [13]. V1 T2 T1 V 2 V1 P2 P1 V 2 k1 k Cold flow simulation in an air medium was seleted for validation with a theoretially thermodynami alulation for this researh. Detailed boundary onditions are shown in Table 1 and Table 2. The simulation time started form IVC at 552 CA to EVO at 858 CA, and the simulation results for pressure and temperature were ompared with the thermodynami theory (as shown in Figure 2). From the graph, it an be observed that pressure and temperature showed a strong agreement between simulation and theory. The maximum value for pressure in the simulation was reorded as 5.4 MPa and temperature as 864K. However, the maximum value for temperature and pressure for the theoretially thermodynami alulations were 924 K and 5.7 MPa, respetively; slightly higher than the value obtained in the simulation. The disrepanies between maximum values for pressure and temperature were 5% and 6%, respetively. Due to this validation proess, it was deemed pratial to proeed to the next step of the simulation with hydrogen fuel diret injetion in air and argon-oxygen mediums. (2)

5 81 Nik Muhammad Hafiz et al. / Jurnal Teknologi (Sienes & Engineering) 78: 6 10 (2016) Figure 2 Pressure and temperature for simulation and theory results Figure 4 Temperature and heat release rate for H2 DI in air medium 3.2 Diret Injetion Of Hydrogen In Compression Ignition Engine With Air Medium The simulation proess ontinued with hydrogen fuel injeted into a ompression ignition engine at 702 CA to 712 CA. Figure 3 shows the pressure and temperature results from the simulation. Pressure inreased and dereased proportionally with the derease and inrease of in-ylinder volumes. The same trend was also observed for temperature. The maximum pressure and temperature obtained was 847 K and 5.5 MPa, respetively. Figure 4 shows a very small value of heat release rate with a maximum value of 1.54 X 10-5 [J/CA]. It an be onluded that no ombustion ourred throughout the proess. The maximum temperature of 847 K (as shown in Figure 4) prevented the hydrogen fuel from igniting, beause the auto-ignition temperature for hydrogen is 858K [14]. Figure 3 Pressure and temperature for H2 diret injetion in air atmosphere 3.3 Diret Injetion Of Hydrogen In Compression Ignition Engine With Argon-Oxygen Medium Figure 5 shows the values of temperature between three different onditions, whih are old flow simulation with air, old flow simulation with argonoxygen, and hydrogen diret injetion with argonoxygen medium. From the graph, it is understood that temperature inrease in the old flow simulation with argon-oxygen medium was higher than in the air medium beause the low speifi value of argon generated a higher temperature during the ompression stroke in the engine. The maximum temperature obtained for the old flow simulation in argon-oxygen medium was 1,286 K; instead of 847 K for the simulation in air medium. This ondition is suitable for hydrogen to ignite in a ompression ignition engine; whereby the temperature ahieves 1,245 K at 702 CA while the hydrogen is being injeted during the simulation proess. As a result, the maximum temperature obtained for hydrogen diret injetion in argon-oxygen medium was1,545 K. Figure 6 shows the heat release rate value for hydrogen fuel ombustion in argon-oxygen medium. Even though the injetion began at 702 CA, the fuel started to ignite with a premixed ombustion at 710 CA. The maximum heat release rate was reorded at 908 J/degree and an ignition delay (τ) of 8.8 CA (9.17 ms). Mansor et al. [15] performed an experiment using a onstant volume ombustion hamber with ambient temperature 1,200 K. They obtained a value of ignition delay of 0.3 ms; whih is muh shorter than the duration obtained in the simulation. This disrepany was due to the differenes of oxygen onentration and equivalene ratio between the experiment and the simulation [15-16]. From the simulation results, the value of heat release rate shown in Figure 6 is very high ompared with the value shown in Figure 4. Longer ignition delay provides suffiient time for the hydrogen fuel and the oxygen to mix properly and allow the gas mixture to burn under premixed ombustion [17]. Moreover, high fuel injetion pressure, at 9 MPa, and the potential of

6 82 Nik Muhammad Hafiz et al. / Jurnal Teknologi (Sienes & Engineering) 78: 6 10 (2016) H2 gas with higher flame speed and faster ombustion veloity [18], indues the high premixed ombustion phase and produes a high heat release rate. simulation proess, at 552 CA and 662 CA, only small differenes in temperature values were obtained. Later, during the injetion period at 702 CA, the values of temperature in argon-oxygen medium had already reahed the hydrogen auto-ignition temperature; but a different situation ourred in the medium in normal air. At 792 CA, the temperature remained high in the argon-oxygen medium due the ombustion proess; however, the temperature in normal air had already dereased towards the ambient temperature. Detailed values for pressure and temperature between hydrogen diret injetion in argon-oxygen and normal air medium are shown in Table 3. Figure 5 Temperature for H2-DI in Ar-O2 ompared to old flow temperature in air and Ar-O2 Figure 7 Mass of ombustion produt Figure 6 Heat release rate for H2-DI in Ar-O2 atmosphere Figure 7 shows the speies mass of reatant and produt during the simulation proess from the beginning at 552 CA until the end at 858 CA. Major produts for hydrogen ombustion with argon-oxygen medium are oxygen (O2) and water (H2O). Besides that, other speies, suh as O, OH, HO2 and H2O2, ontribute only a very minimum mass value. It was observed that O2 remained high in volume and proved that the ombustion ourred under lean ombustion. Sine argon gas reats with no other gases during the simulation, it is proposed that this gas is reirulated bak to the engine when designing and produing a onept engine in the future [7]. Figure 8 shows the simulation results of temperature for hydrogen diret injetion in argon-oxygen medium, ompared with normal air. The result was obtained using EnSight software. Contour of temperature was ompared at 552 CA (IVC), 662 CA (before injetion), 702 CA (during injetion) and 792 CA (during ombustion). At the beginning of the Figure 8 Comparison temperature image for DI H2 in Ar-O2 and air atmosphere

7 83 Nik Muhammad Hafiz et al. / Jurnal Teknologi (Sienes & Engineering) 78: 6 10 (2016) Table 3 Temperature and pressure at 552 CA, 632 CA, 702 CA, and 792 CA Medium Ar-O2 Air Pressure [MPa] Temperatur e [K] Pressure [MPa] Temperatur e [K] 4.0 CONCLUSIONS CA 632 C A 702 C A 792 C A , In this researh, the possibility to replae nitrogen in normal air with the mono atomi noble gas argon, for diret injetion of hydrogen fuel in ompression ignition engine, is studied using numerial Converge software. The following onlusions were obtained from the results: 1. Hydrogen diret injetion in normal air medium with maximum temperature of 848 K does not ignite as shown in the simulation results, even though the obtained temperature was lose to the hydrogen autoignition temperature. 2. Replaing the ombustion medium from air to argon-oxygen aused an inreasing inylinder temperature and allowed the hydrogen fuel to ignite and burn. 3. The ignition delay for hydrogen ombustion at the start of injetion at 702 CA was 9.17 ms. 4. Produt of hydrogen fuel ombustion in an argon-oxygen medium is mainly oxygen and water. In this study, only the preliminary data was obtained. More detailed analysis, with more variables onfirmed by simulation and experimentation, is needed. Hene, in order to realize this mission, muh more work should be done by many researhers, in terms of gas portability for hydrogen, oxygen and argon, and the safety and delivery of these gasses. All of these issues will remain as new hallenges for the next researh. Aknowledgement The author would like to thank the Ministry of Eduation Malaysia for supporting the researh with grants GGPM and FRGS/2/2013/TK01/UKM/02/1. Referenes [1] Reitz, R. D Diretions in Internal Combustion Engine Researh. Combustion and Flame. 160(1): 1 8. [2] Energy Tehnology Perspetives, International Energy Ageny, Paris, Information on publiation/name, 31269, en.html. [3] National Automotive Poliy (NAP) 2014, Malaysia Automotive Assoiation. Information on [4] Furuhama, S., Yamane, K. and Yamaguhi, I Combustion Improvement in a Hydrogen Fueled Engine. International Journal of Hydrogen Energy. 2(3): [5] Ikegami, M., Miwa, K. and Shioji, M A Study of Hydrogen Fuelled Compression Ignition Engines. International Journal of Hydrogen Energy. 7(4): [6] Adnan, R., Masjuki, H. H. and Mahlia, T. M. I Performane and Emission Analysis of Hydrogen Fueled Compression Ignition Engine with Variable Water Injetion Timing. Energy. 43(1): [7] Kuroki, R., Kato, A., Kamiyama, E. and Sawada, D Study of High Effiieny Zero-Emission Argon Cirulated Hydrogen Engine. SAE Tehnial Paper. [8] Mansor, M. R. A., Samiran, N. A., Mahmood, W. M. F. W. and Raja, N. A Effet of Injetor Nozzle Design on Spray Charateristis for Hydrogen Diret Injetion Engine Conditions. Applied Mehanis and Materials. 660: [9] Seneal, P. K., Rihards, K. J., Pomraning, E., Yang, T., Dai, M. Z., Mdavid, R. M., Patterson, M. A A New Parallel Cut-Cell Cartesian CFD Code for Rapid Grid Generation Applied to In-Cylinder Diesel Engine Simulations. SAE Tehnial Paper. 724: [10] Zhiyin, Y Large-eddy simulation: Past, present and the future. Chinese Journal of Aeronautis. 28(1): [11] Toubiana, E., Russeil, S., Bougeard, D. and François, N Large Eddy simulation and Reynolds-averaged Navier Stokes Modelling Of Flow In Staggered Plate Arrays: Comparison At Various Flow Regimes. Applied Thermal Engineering. 89: [12] Killingsworth, N. J., Rapp, V. H., Flowers, D. L., Aeves, S. M., Chen, J. Y. and Dibble, R Inreased Effiieny In SI Engine With Air Replaed By Oxygen In Argon Mixture. Proeedings of the Combustion Institute. 33(2): [13] Yunus A. Cengel, Mihael A.Boles, Thermodynamis, An Engineering Approah 7th edition, M Graw Hill page [14] White, C., Steeper, R. and Lutz, A The Hydrogen- Fueled Internal Combustion Engine: A Tehnial Review. International Journal of Hydrogen Energy. 31(10): [15] Mansor, M. R. A., Nakao, S., Nakagami, K., Shioji, M. and Kato, A Ignition Charateristis of Hydrogen Jets in an argon-oxygen Atmosphere. SAE Tehnial Paper. [16] Tang, C., Man, X., Wei, L., Pan, L. and Huang, Z Further Study On The Ignition Delay Times Of Propane-Hydrogen- Oxygen-Argon Mixtures: Effet Of Equivalene Ratio. Combustion and Flame. 160(11): [17] John B. Heywood Internal Combustion Engine Fundamentals, M Graw Hill Series in Engineering [18] Mohammadi A, Shioji M, Nakai Y, Ishikura W, Tabo E Performane and Combustion Charateristis of a Diret Injetion SI Hydrogen Engine. International Journal of Hydrogen Energy. 32(2):