CFD Combustion Models for IC Engines. Rolf D. Reitz

Transcription

1 CFD Combustion Models for IC Engines Rolf D. Reitz Engine Research Center University of Wisconsin-Madison ERC Symposium, June 7, 27

2 Combustion and Emission Models at the ERC Turbulence vs. chemistry time scales 1. Characteristic Time Combustion Model (KIVA-CTC) Auto-ignition model (SHELL) ignition with 8 reaction steps Combustion model (CTC) turbulence mixing timescales 2. Representative Interactive Flamelet Model (KIVA-RIF) Flamelet equations/detailed chemistry coupled with CFD (Peters et al.) Effect of turbulence is modeled with average scalar dissipation rate 3. Direct Integration of CFD with Chemistry (KIVA-CHEMKIN) Each computational cell treated as well stirred reactor ERC reduced mechanism for n-heptane chemistry timescales 4. KIVA-CHEMKIN-G Model Explicit modeling of turbulent flame propagation combined with detailed chemistry Eddy breakup timescales Chemistry timescales

3 Auto-ignition and Detailed Chemistry Models Skeletal Chemistry Mechanism for Multi-Surrogate Fuels Gasoline Iso-octane 857 species 3586 reactions 25 species 51 reactions Diesel Bio-diesel N-heptane Methyl butanoate 57 species 258 reactions 264 species 1219 reactions 3 species 65 reactions 41 species 15 reactions Multi-fuel Mechanism (ERC-PRF-Bio) Aromatic Alcohol Toluene Ehthanol surrogate fuel Multi-component fuel Comprehensive mechanism Skeletal mechanism Dr. Youngchul Ra Jessica Brakora SAE , SAE , SAE

4 Validation - Auto-ignition Model n-heptane-iso-octane/air mixtures, stoichiometric, P=4 Bar) Calculation using ERC-PRF mechanism Ignition Delay Time (ms) % i-octane 9% i-octane / 1% n-heptane 8% i-octane / 2% n-heptane 6% i-octane / 4% n-heptane 1% n-heptane Shock tube measurements (Fieweger et al. 1997) SAE (K)/T

5 Validation Cat HD engine rpm = 1737 eq_ratio =.23 c_ratio = 16.1 inj_timing = -113 ca rpm CAD ATDC Experiment eq_ratio =.27 Shell-CTC c_ratio = 16.1 CHEMKIN inj_timing = ca CAD ATDC CO (g/kgf) Experiment 9 14 Shell-CTC Gasoline HCCI CHEMKIN AHRR (J/deg) AHRR (J/deg) NOx (g/kgf) Experiment Shell-CTC Chemkin.135 Experiment Shell-CTC Equivalence ratio Chemkin Equivalence ratio SAE , SAE

6 Validation - GM Light-duty Diesel Engine KIVA-CHEMKIN SOIC = 39, -32 and 23 SOIC=-32 GM 1.9L 4-cylinder Bore Stroke mm Compression Ratio 16.6: 1 Engine Speed 2 rev/min IMEP 5.5 Bar EGR 67% PCCI SAE

7 Advances in SI Engine Combustion Modeling G-equation Combustion Model with Detailed Chemical Kinetics SAE Discrete particle ignition model Burnt Gas G-Equation Flame propagation End Gas ~ G t + ( v f v vertex ~ ) G = u S T Auto-ignition (detailed kinetics) ~ G D T ~ ~ k G SAE SAE

8 Validation Ford SI Engine EXPT SIMU PFI mode -32 O ATDC 2.5 KIVA-CHEMKIN-G EXPT SIMU PFI mode -44 O ATDC Ford DISI Bore Stroke mm Compression Ratio 12 : 1 Engine Speed 15 rev/min PFI Spark sweep EXPT Spk = -2 o ATDC. 3. SIMU Crank Angle ( o ATDC) EXPT MAP = 75 kpa -.5 SIMU Crank Angle ( o 25 ATDC) Crank Angle ( o ATDC) EXPT Spk = -32 o ATDC. 3. SIMU Heat Release Rate (J/Deg Heat Release Pressure Rate (MPa) (J/Deg Crank Angle ( o ATDC) EXPT MAP = 1 kpa SIMU Crank Angle ( o ATDC) Crank Angle ( o ATDC) Heat Release Rate (J/De Heat Release Rate (J/Deg DI Spark sweep DI MAP sweep SAE

9 Validation - Sandia Cummins Diesel High-T, long-id Condition (Premixed Combustion) O 2 = 21% SOI = -5 AHRR (J/deg) TEST CHEMKIN G_Model Injection Profile CAD ATDC High-T, Long-ID Experiment KIVA-CHEMKIN KIVA-CHEMKIN-G SAE Green: OH Red: Soot

10 Validation - Cummins Dual-Fuel Engine Experimental SAE KIVA-CHEMKIN-G NOx Emissions (grams) Experimental Measurements 18 Case 1 Case 2 Case 3 15 Case 4 Case 5 12 AHRR (J/deg) Measured Predicted CAD ATDC NOx Case1 Case2 Case3 Case4 Case5 C AHRR (J/deg) Model Predictions BMEP 1 bar SOI -87 BTDC PIVC 2.1 bar NG.44 CR 14.5 Case 1 Case 2 Case 3 Case 4 Case CAD ATDC Case TIVC (K) Primary Fuel: Natural Gas Ignition Source: Diesel Pilot (1%)

11 Validation - Cummins Dual-Fuel Engine KIVA-CHEMKIN vs KIVA-CHEMKIN-G AHRR(J/deg) Test KIVA-CHEMKIN Case AHRR(J/deg) Test KIVA-CHEMKIN-G Case CAD ATDC CAD ATDC 2 Without flamepropagation combustion is too slow

12 Summary and Conclusions Current CFD models describe important physical and chemical processes of engine flows, but validation experiments are needed for model development and improvement CFD combustion models capture engine performance and emissions trends and can be combined with optimization tools for design and evaluation of engine design concepts Methodologies are in place for modeling a wide range of combustion regimes: HCCI PCCI Diesel SI flame propagation dual-fuel Multi-component fuel vaporization models can be coupled with advanced chemical kinetics models Surrogate fuel models are under development to represent representative practical fuels (MB, toluene, ethanol..)

13 Multi-Component fuel modeling Dr. Youngchul Ra Gasoline* * Adapted from J. Farrel, Exxon Mobil Corp. Diesel Aromatics % 16 Sulfur ppm 7.3 Paraffins % 42 Napthenes % 42 Olefins %.3 e l o m C/H ratio r o n o i t u b i r t s i d gasoline composition 1.4 Discrete g p (mw i ) iso-octane approximation Single comp approx Continuous f p (I) molecular weight d i s t r i b u t i o n o r m o l

14 PRF5: Emissions HC at EVO [g.kg-f] PRF5 Diesel-sgl CO at EVO [g/kg-f] PRF5 Diesel-sgl 2 HC CO start of injection command [deg ATDC] start of injection command [deg ATDC].25.2 PRF5 Diesel-sgl NOx [g/kg-f].15.1 LTC regime (EGR=65%) Injection sweep.5. NOx start of injection command [deg ATDC]

15 Reduced Biodiesel Mechanism Jessica Brakora LLNL methyl butanoate (MB) * Surrogate to represent the large methyl esters in biodiesel MB chemical formula: C 5 H 1 O species, 1219 reactions Mechanism Reduction Method SENKIN constant volume analysis Identify key species and reaction paths based on ignition timing Peak concentration: remove species that never exceed specified mole fraction thresholds (-112 species) XSENKPLOT reaction flux analysis: identify key reaction pathways, remove less important paths (-111 species) Adjust rate constants for key reactions to optimize ignition timing *Proc. Comb. Inst., Vol 28, 2/pp E+ 1.E-1 1.E-2 1.E-3 1.E-4 Large flux to mb2j (C 5 H 9 O 2 ); test removal of other paths Ignition Delay [s] LLNL detailed (264 species) rm spec < 1^-1 (199 spec) rm spec < 1^-9 (179 spec) rm spec < 1^-8 (162 spec) No noticeable change in ignition; these species can be removed 1.E Initial Temperature [K] XSENKPLOT reaction paths: MB fuel breakup

16 Reduced MB Kinetic Mechanism mb ch3o2h (ch3o, ch2o, ch3o2) mb2ooh mb2oo mb2j These species further break down into lower-level carbon species (ch3oco) mb2o mb2ooh4j mp2d MB reduced coupled with ERC n-heptane: 41 species, 15 reactions me2*o c2h5cho mb2ooh4oo ch2cho 1.E+2 P = 4bar, Phi = 1. LLNL detailed me2j*o (ch2co,ch3oco) (c2h5) mb4ooh2*o mp3j2*o ch2co (ch2o) Ignition Delay [ms] 1.E+1 1.E+ 1.E-1 MB reduced (ch3oco, ch2co) 1.E Initial Temperature [K]

17 Auto-ignition and Detailed Chemistry Models Diesel Gasoline RH + O 2 ROOH First Stage Ignition OH ORO Olefin RH + O 2 HO 2 H 2 O 2 ERC n-heptane mechanism SAE : 34 species H 2 O 2 + (M) OH + OH + (M) Second Stage Ignition ERC PRF Reduced Chemistry SAE , SAE : 47 species

18 Multi-component Fuel DMC/PRF Models PRF5 fuel pressure [MPa] n-heptane iso-octane HRR [J/deg] P, inj= -45 P, inj= -42 P, inj= -39 P, inj= -33 P, inj= -27 P, inj= -21 P, ic8 P, nc7 HRR, inj= -45 HRR, inj= -42 HRR, inj= -39 HRR, inj= CA [deg ATDC] HRR, inj= -27 HRR, inj= -21 HRR, ic8 HRR, nc7