Numerical Study on the Combustion and Emission Characteristics of Different Biodiesel Fuel Feedstocks and Blends Using OpenFOAM Harun M. Ismail 1, Xinwei Cheng 1, Hoon Kiat Ng 1, Suyin Gan 1 and Tommaso Lucchini 2 1 Department of Mechanical, Materials and Manufacturing Engineering University of Nottingham (Malaysia Campus) 2 Dipartimento of Energia, Politecnico di Milano, Italy
Introduction Motivation Both conventional petroleum industry and bio-fuel industry (Palm Oil) is a multi-million dollar business in Malaysia (biggest Palm Oil exporter in the world ) Limited petroleum resources and increasing emission standards Bio-fuels still at its infancy stage, the development of theories to understand its combustion and emission nature is not widely available Objectives of this work Development and applications fuel thermo-physical and transport properties of Coconut (CME), Palm (PME) and Soy methyl-esters (SME) for in-cylinder (IC) spray combustion CFD modelling Development and applications of generic reduced combustions kinetics suitable for CME, PME and SME for CFD, IC engine applications Analyse the influence and relation between fuel properties and fuel spray structures for different biodiesel/diesel blend levels (B0 B100) and fuel type Investigate combustion and emission characteristics for different blend levels and fuel type (CME, PME and SME) 1
Biodiesel Thermo-physical & Transport Properties Properties calculated using Group contribution method Evaluation of fuel thermo-physical properties & transport properties up to the critical temperature of a respected fuel Type of Fatty-acids Chemical Formula Soy % Palm % Coconut % Saturated C 12 H 26 O 2 - - 48 Saturated C 15 H 30 O 2 - - 17 Saturated C 11 H 22 O 2 - - 9 Saturated C 17 H 34 O 2 18 42 8 Saturated C 19 H 38 O 2 7 5 - Unsaturated C 19 H 36 O 2 10 41 18 Unsaturated C 19 H 34 O 2 60 10 - Unsaturated C 19 H 32 O 2 10 2 - Bio-fuel components Based on information compiled from open literature and lab fuel test Five largest contributing components of ME are identified for properties computation Property CME PME SME Critical Temperature (K) 773.5 789.2 721.2 Critical Pressure (bar) 14.0 13.0 15.3 Critical Volume (ml/mole) 1064.0 1084.0 885.0 2
Biodiesel Thermo-physical & Transport Properties Vapour Thermal Conductivity / [W/m-K] Vapour Diffusivity / log[m^2/s] 0.10 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00 1.0E+00 1.0E-01 1.0E-02 1.0E-03 1.0E-04 1.0E-05 1.0E-06 Diesel CME PME SME 280 380 480 580 680 780 Diesel CME PME SME Temperature (K) 280 380 480 580 680 780 Temperature (K) Some examples of the PME blends (B0 B100) as compared to diesel for illustration Vapour viscosity (Pa-s) 1.6E-04 1.4E-04 1.2E-04 1.0E-04 8.0E-05 6.0E-05 4.0E-05 2.0E-05 0.0E+00 Latent heat of vaporization (kj/mol) 450 400 350 300 250 200 150 100 50 0 B20 B50 B80 Diesel 280 380 480 580 680 780 Temperature (K) B20 B50 B80 Diesel 280 380 480 580 680 780 Temperature (K) Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 6
Biodiesel Thermo-physical & Transport Properties Implementations in OpenFOAM Snippets of properties implemented in OpenFOAM fuel library Interpolate function were utilised to estimate fuel properties at different temperatures 7
Experimental Setup (Nottingham Research Engine) Nottingham Research Engine Specifications Engine Type Light-Duty Diesel Piston Type Bowl-in-piston Displacement Volume per-cylinder 347cm 3 Compression Ratio 19 : 1 Stroke 69 mm Bore 80 mm Connecting Rod Length 114.5 mm Intake Valve Closing (IVC) -140 o ATDC Exhaust Valve Opening (EVO) 140 o ATDC Engine Speed 1500 3500 rev/min Experimental engine consists of 4- hole injector at 90 o apart A 90 o 3D-model of the combustion chamber is generated The mesh set with 1 injector hole at 45 o from X and Y axis. Dynamic mesh with cell size of 1mm 8
Experimental Setup (Chalmers HP/HT Rig) Chalmers High Pressure/High Temperature Rig Type Constant Volume Volume 2.0 l Injection Period 3.5 ms Injection Pressure 1200 bar Fuel Temperature 313.15 K Chamber Pressure (constant) 50 bar Chamber Initial Temperature 830 K Fuel of Interest Diesel(B0), Palm methyl-ester (B100) Cell size increased from 0.25 to 2 mm Chamber size fit to (60 x 60 x 150) mm Chamber air density maintained to match experimental condition at 20.89 kg/m 3 Air Heaters Exhaust Gas Heat Exchanger Gas Exit Increasing cell size Pressurised Air Inlet Regulator Optically Accessed Chamber Raul et al. (SAE 2008-01-1393) Air Flow Controller Fuel Injection Point Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 9
Utilised OpenFOAM ICE-Lib (Polimi) Models Injector Model huhinjector Atomisation HuhGosman Heat Transfer RanzMarshall Collision Model off Injector Type unitinjector Wall Model BaiGosman Break-up Model Reitz-KHRT Wall Heat Transfer Model turbulentheatflux Dispersion Model stochasticdispersionras Turbulence Model RNG-k-epsilon Evaporation standardevaporationmodel Drag Model standarddragmodel Combustion Model PSR model 10
Validations of Fuel Thermo-physical & Transport Properties Using OpenFOAM The simulated spray structure matches with experimental data with good accuracy Fuel spray structure could be simulated using the developed fuel properties 0.14 0.12 0.10 MeasuredDiesel SimulatedDiesel SimulatedPME VpMeasuredDiesel VpSimulDiesel VpSimulPME MeasuredPME SimulatedCME SimulatedSME VpMeasuredPME VpSimulCME VpSimulSME The thermo-physical and transport properties validated against experimental liquid & vapour axial penetration length Axial Penetration Length / [m] 0.08 0.06 0.04 0.02 0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Time / [ms] Difference in liquid penetration length between diesel B(0) and biodiesel (B100) fuels 11
Validations of Fuel Thermo-physical & Transport Properties Using OpenFOAM Illustration of the four fuel species distribution contour plot Biodiesel fuel tends to have SIMILAR spray and evaporation structure, but NOT SAME The dissimilarities can be clearly seen in fuel evaporation simulation without combustion Biodiesels (B100) have longer spray penetration length due to higher viscosity and surface tension Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 12
Effects of Fuel Thermo-physical & Transport Properties (Fuel Type Comparison) Evaporated Mass / [kg] 9.0E-07 8.0E-07 7.0E-07 6.0E-07 5.0E-07 4.0E-07 3.0E-07 2.0E-07 1.0E-07 EvapMassIDEA (B0) EvapMassCME (B100) EvapMassPME (B100) EvapMassSME (B100) CAD range before start of combustion (SOC) -5 o CAD Fuel Composition / [% by vol] 0.0E+00 100 90 80 70 60 50 40 30 20 10 0-11 -9-7 -5-3 -1 1 Crank Angle Saturated Unsaturated CME PME SME Fuel Type % of Unsaturation SME > PME > CME % of individual composition that makes up SME and PME Rate of Evaporation: Diesel (IDEA) > CME > SME > PME Mass of fuel injected equal for all fuel Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 13
Effects of Fuel Thermo-physical & Transport Properties (Fuel Type Comparison) Fuel Composition / [% by vol] 100 90 80 70 60 50 40 30 20 10 0 Saturated Unsaturated CME PME SME Fuel Type % of Unsaturation SME > PME > CME Suggests the importance of % individual composition that makes up SME and PME Rate of Evaporation: Diesel (IDEA) > CME > SME > PME Mass of fuel injected equal for all fuel SME can be evaporated significantly faster Heat of Vaporisation / [kj/kg] 500 450 400 350 300 250 200 150 100 50 0 280 380 480 580 680 780 Temperature (K) Diesel CME PME SME Heat of Vaporisation: From T = 280 650 K SME < PME 14
Effects of Fuel Thermo-physical & Transport Properties (Palm B0-B100) Evaporated Mass / [kg] 9.0E-07 8.0E-07 7.0E-07 6.0E-07 5.0E-07 4.0E-07 3.0E-07 2.0E-07 1.0E-07 EvapMassIDEA (B0) EvapMassPME (B20) EvapMassPME (B50) EvapMassPME (B80) EvapMassPME (B100) -5 o CAD 0.0E+00-11 -9-7 -5-3 -1 1 Crank Angle Suggests the strong influence of IDEA fuel (Diesel) up to 80% biodiesel mixture (B80) Normalised Evaporated Mass / [kg/kg] 1.2 1 0.8 0.6 0.4 0.2 0 B0 B20 B50 NormalisedEvapMass(B0-B100) B80 B100 0 20 40 60 80 100 % Blends (PME) Increasing % of biodiesel 15
Development & Applications of Generic Reduced Biodiesel Fuel Surrogate Combustion Kinetics Detailed LLNL methylester mechanism with 301 species and 1516 reactions Motivation Computational resources (time & cost) Lack of widely available biodiesel mechanism validated for Palm, Coconut and Soy methyl-esters To investigate the combustion and emission characteristics of Palm, Soy and Coconut biodiesel fuels. Adapted from C.K Law et al. [AIAA 2008-969] Reduced mechanism 77 species 212 reactions Reduced and validated for 48 shock-tube conditions (STC) during entire mechanism reduction process with comparison the detailed mechanism Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 16
0-D (PSR) Validations of Generic Reduced Biodiesel Fuel Surrogate Combustion Kinetics 1,000 P= 13.5, Ø= 0.5-1.5 1000 P= 41.0, Ø= 0.5-1.5 100 77 species, 212 reactions 100 77 species, 212 reactions Ignition Time/ [ms] 10 modifiedllnl-eq=0.5 Reduced-EQ=0.5 1 modifiedllnl-eq=1.0 Reduced-EQ=1.0 modifiedllnl-eq=1.5 Reduced-EQ=1.5 0 0.60 0.80 1.00 1.20 1.40 1.60 1000/T Ignition Time/ [ms] 10 1 modifiedllnl-eq=0.5 Reduced-EQ=0.5 modifiedllnl-eq=1.0 0.1 Reduced-EQ=1.0 modifiedllnl-eq=1.5 Reduced-EQ=1.5 0.01 0.60 0.80 1.00 1.20 1.40 1.60 1000/T Gas Phase (PSR) Validations at 48-STC Error in ID less than 13 % during 0-D reductions process as compared to LLNL mechanism Combine with modified skeletal n-heptane mechanism to match energy content and C/H/O ratio ID shown here based on Nottingham Research Engine calibrations 18
3-D CFD Validations of Reduced Biodiesel Fuel Surrogate Combustion Kinetics Using OpenFOAM Diesel Pressure trace comparison for engine conditions at 2000 rpm and power 1.5kW The simulated spray combustion matches with experimental data with good accuracy Future test will be conducted on constant fuel mass and constant ID to study the combustion an emission characteristics for CME, PME and SME PME SME Legend Measured Simulatied 19
Combustion and Emission Characteristics of Palm, Soy and Coconut (B100) Biodiesel Fuels NO plot for engine conditions at 2000rpm and 1.5 kw of power. Start of injection (SOI) and end of injection (EOI) for all cases are the same -4 o CAD ATDC SOI EOI Change in fuel type leads to change in NO distributions Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 20
Combustion and Emission Characteristics of Palm, Soy and Coconut (B100) Biodiesel Fuels SME -4 o CAD ATDC PME NO mass fraction at -4 o CAD ATDC SME > PME Soot mass fraction at -4 o CAD ATDC SME > PME Similar to trend observed in experimental studies Three possibilities for the observed trend : Mass of fuel injected (SME > PME) Ignition Delay (SME > PME) Level of usaturation in the fuel (SME > PME) Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 21
Conclusions Conclusions Main objectives of development and implementations of biodiesel fuel properties and combustion kinetics were achieved As for preliminary validations, good level of agreement was achieved between computed and measured data for both fuel properties and combustion kinetics Effects of fuel properties and chemical kinetics could be isolated using CFD studies to better understand the combustion and emission characteristics of biodiesel fuel in CI engines Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 22
Acknowledgments Authors would like to thank Dr. Tommaso Lucchini for his helpful suggestions and guidance Authors would like to thank Internal Combustion Engine (ICE) Group PoliMi for the collaborative effort given during the research work This research work is currently funded by Ministry of Science oftechnology (MOSTI) of Malaysia. Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011 23
THANK YOU Contact email: harun.ismail@nottingham.edu.my Internal Combustion Engine Group Energy, Fuel and Power Technology Research Division Department of Mechanical, Materials and Manufacturing Engineering University of Nottingham (Malaysia Campus) Meeting on Internal Combustion Engine Simulations Using OpenFOAM, Milan, Italy, 2011