COMPARISON OF VARIABLE VALVE ACTUATION, CYLINDER DEACTIVATION AND INJECTION STRATEGIES FOR LOW-LOAD RCCI OPERATION OF A LIGHT-DUTY ENGINE Anand Nageswaran Bharath, Yangdongfang Yang, Rolf D. Reitz, Christopher J. Rutland University of Wisconsin Madison Engine Research Center North American GT-User Conference 2015
Acknowledgements In addition to our advisors Prof. Rolf Reitz and Prof. Christopher Rutland, and our sponsors DERC, we would also like to thank the following people: Reed Hanson, Shawn Spannbauer & Chris Gross for experimental data Current & Former ERC Colleagues: Dipa, Jae-Hyung Lim, Yifeng Wu, Nitya Kalva, Jian Gong & Xingyuan Su Paramjot Singh & Daniel Schimmel from Gamma Technologies Joshua Leach, ERC Systems Administrator Mike Andrie 2
Motivation High CO & UHC emissions at low-load operation of RCCI (and LTC regimes in general) due to low combustion temperatures At low-load operation, exhaust gas temperatures are insufficient to light off the Diesel Oxidation Catalyst (DOC) Other strategies for DOC light-off: 1. Using electrical heater requires large electrical currents for rapid catalyst warm-up (150 250 A), placing high power demand on vehicle electrical power supply components (Laing, Socha) 2. Varying exhaust gas composition to include hydrogen to lower light-off temperature, but at least one cylinder has to run rich to generate hydrogen, increasing PM emissions (Katare et al.) It can therefore be seen that increasing exhaust temperatures would be a good way to achieve DOC light-off and improve catalyst efficiency. 3
In our previous work, we showed that Early Exhaust Valve Opening (EEVO) with fully flexible VVT raised exhaust gas temperatures sufficient for DOC light-off However, there was a significant deterioration in fuel economy Should a cam phaser be used instead due to cost & complexity? What about other methods, e.g. post-injection, combustion phasing, cylinder deactivation, etc.? Bharath et al., ASME ICEF 2014, Paper No. ICEF2014-5534, Oct. 2014 Compare EEVO, combustion phasing, post-injection & cylinder deactivation in raising exhaust gas temperatures for DOC light-off, and impact of each strategy on fuel economy & emissions at a near-idle load point using coupled GT-Power and KIVA- 3V simulations. 4
5 University of Wisconsin Engine Research Center Simulation Setup Engine & DOC Specifications Bore 82 mm Stroke 90.4 mm No. of Cylinders 4 Displacement 1.9 liters Compression Ratio 16.7 Turbocharger Variable Geometry EGR High Pressure EGR Loop Engine Schematic used in GT-Power Model Catalyst Specifications Volume (L) 0.6 Substrate Dimension D: 90 mm X L: 90 mm Substrate Material Metallic CPSI/Wall Thickness 200/50 micron Loading/Washcoat 90 g ft 3 Platinum Heat Shield Yes Stock Piston Geometry with re-entrant bowl Computational grid Specifications Cells at IVC 9,738 Cells at TDC 3,528 Radial resolution (mm) 4 Azimuthal resolution (deg.) 2.9
6 University of Wisconsin Engine Research Center Simulation Setup Data for Model Calibration Data for Engine System Model Calibration 1 Bar BMEP 4 Bar BMEP ** Intake Manifold Pressure (Bar) 1.006 1.06 Fuel Energy per Cylinder (J) 275.1 563.3 Engine Speed (rev/min) 1,500 Gasoline Quantity (mg/cyl/cyc) 3.525 10.5 Diesel Quantity (mg/cyl/cyc) 2.619 2.1 Gasoline Start of Injection (Deg.) -227.4 Diesel Start of Injection (Deg.) -40-45 Diesel Fuel Rail Pressure (Bar) 400 480 EGR Fraction (%) 49.9 0 Temperature Boundary Conditions for Heat Transfer in KIVA-3V Parameter Cylinder Liner Temp. Cylinder Head Temp. Cylinder Wall Temp. Value 390 K 440 K 440 K **NOTE: 4 Bar BMEP case used for cylinder deactivation study
7 University of Wisconsin Engine Research Center Simulation Setup - Calibration Procedure Experimental HRR Profile IMEP, BMEP, Intake Air Flow Rate, Volumetric Efficiency, etc. P iii, T iii & Gas Component Mole Fractions KIVA-3V
8 University of Wisconsin Engine Research Center Simulation Setup Calibration Results 1 Bar BMEP 4 Bar BMEP Experiment Simulation Experiment Simulation BMEP (Bar) 1.10 1.06 4.00 4.15 Intake Air Flow Rate (kg/h) 41.11 41.86 89.8 93.9 Intake Manifold Gas Temp. (K) 348.1 351.1 312.2 323.0
Simulation Setup Calibration Results Experimental Data for DOC Model Calibration BMEP (Bar) 1 2 3 4 Operating Conditions Exhaust Gas Temp. (K) 410 421 456 484 Intake Air Flow (kg/h) 88 87 87 89 DI (mg/inj) 3.864 4.278 4.264 3.396 PFI (mg/inj) 2.812 4.178 6.7 9.49 Engine Speed (rev/min) 1,500 Emissions Pre-DOC UHC (ppm) 1,192 1,006 1,108 1,091 Post-DOC UHC (ppm) 1,128 1,182 1,019 147 Pre-DOC NO (ppm) 5.5 8.56 4.81 6.26 Pre-DOC CO (ppm) 7,005 5,820 2,458 1,432 Pre-DOC H 2 O (%) 2.3 3.07 4.05 4.69 Pre-DOC CO 2 (%) 2.8 2.54 3.98 4.7 Pre-DOC O 2 (%) 16.96 15.35 14.3 13 DOC Model Calibration Results Platinum Dispersion Factor used: 1.45% 9
Combustion Phasing Study Gasoline/diesel ratios from 10% isooctane/90% n-heptane to 90% isooctane/10% n-heptane, with 5% intervals between ratios Global fuel reactivity denoted by global PRF Octane Number: Ψ PPP = m C8H18 100+m C7H16 0 m C 8H18 +m C7H16 No combustion for Ψ PPP > 75 as fuel reactivity too low Combustion efficiency decreases with increasing Ψ PPP, in turn leading to increased UHC and CO emissions. 10
Combustion Phasing Study Exhaust gas temperatures before DOC were below the DOC light-off temperature of 457 K. Hence UHC and CO conversion efficiencies were poor for all iso-octane/n-heptane ratios In conclusion, varying combustion phasing not effective for DOC light-off. Also, peak pressure rise rates may be a problem at advanced combustion phasing. 11
Post-injection Study Definition of Post-injection for this study: Subsequent injection of fuel into the combustion chamber during the expansion stroke following the main injection Equation for mass of additional fuel required: m e + m f h 457 m eh 442 = m f q LLL,C7 H 16 Assumptions in aforementioned equation: 1. Thermodynamic properties of exhaust gas same as air 2. Complete combustion of post-injected fuel 3. All additional post-injected fuel used to increase exhaust gas temperature to lightoff temperature, and none lost to heat or converted to work. From equation, additional fuel required: 4.11 mm s 1 (0.082 mm ccc 1 ccl 1 or 1.34% of total fueling rate) 2 nd injection at injection timings from 40 degrees ATDC to 90 degrees ATDC, with 10 degree intervals. Additional fuel did not burn so exhaust gas temperatures did not increase. 12
13 University of Wisconsin Engine Research Center VVT System Comparison Study for EEVO Compare engine & DOC performance for fully flexible VVT and cam phaser at 80 degrees ATDC EVO timing: o 80 degrees ATDC gave lowest increase in BSFC using EEVO for fully flexible VVT Fully Flexible VVT Cam Phaser IVC Pressure (Bar) 1.17 1.15 IVC Temperature (K) 390 388 EGR Fraction (%) 45 22* o No external EGR used for cam phaser case, as EGR accomplished internally using zero valve overlap o 22% EGR fraction was initial result obtained from GT-Power to calculate the mole fractions for KIVA.
VVT System Comparison Study for EEVO Engine Performance Comparison Baseline Case Fully Flexible VVT Cam Phaser BMEP (Bar) 0.84 0.77 0.44 PMEP (Bar) 0.15 0.17 0.48 Intake Air Flow Rate (kg/h) 39.15 43.27 64.12 Pre-DOC Exh. Temp. (K) 453.8 463.3 472.45 EGR Fraction (%) 49.9 45 22 DOC Exh. Flow Rate (kg/h) 40.21 44.02 65.22 BSFC (g kkh 1 ) 553 607 1,053 DOC Performance Comparison Fully Flexible VVT Cam Phaser UHC Conversion (%) 94.9 89.7 CO Conversion (%) 99.6 98.1 Larger hot residual gas quantity for cam phaser case causes increased heat loss during compression stroke, giving higher pumping work. Lower EGR for cam phaser leads to higher flow rate through catalyst, thereby lower conversion efficiencies for cam phaser case. 14
Cylinder Deactivation Study Cylinders 2,3 and 4 were motored while cylinder 1 was fired. Actuators added to cylinders 2,3 and 4 to cut off fueling and keep valves closed Heat release rate & fueling for 4 Bar BMEP at 1,500 rev/min operating point used in cylinder 1 15
16 University of Wisconsin Engine Research Center Cylinder Deactivation Study Engine Performance Results All Cylinders Firing Only Cylinder 1 Firing Baseline Case BMEP (Bar) 4.15 0.68 0.84 FMEP (Bar) 1.04 1.04 1.04 PMEP (Bar) 0.24 0.05 0.15 Gross IMEP (Bar) 5.43 1.77 2.03 Intake Air Flow Rate (kg/h) 93.9 22.8 39.15 Pre-DOC Exhaust Temp. (K) 500.15 472.45 453.8 DOC Exhaust Flow Rate (kg/h) 96.18 23.34 40.21 DOC Performance Results All Cylinders Firing Only Cylinder 1 Firing Baseline Case UHC Conversion (%) 79.8 99.6 0 CO Conversion (%) 93.4 99.9 18 Much lower flow rate through DOC and the higher than light-off temperature results in near complete conversion of UHC & CO for cylinder deactivation. Also, significantly reduced pumping work due to closure of intake and exhaust valves in motoring cylinders. (BSFC: 351 g kkh 1, best of all 4 strategies)
Conclusions 1. Varying combustion phasing via changing gasoline/diesel ratio did not raise exhaust gas temperatures sufficiently to reach DOC light-off temperature, so catalyst efficiency was poor. 2. 1.34% of fueling rate required for post-injection to raise exhaust gas temperatures to 457 K for DOC light-off. (Lower limit, assuming no heat loss and no conversion of additional fuel to heat.) However, additional fuel did not burn. 3. EEVO using either cam phaser or fully flexible VVT raises exhaust gas temperatures high enough to light-off DOC, but significant deterioration in fuel economy observed (607 g kkh 1 for fully flexible case & 1,053 g kkh 1 for cam phaser case.) 4. Cylinder deactivation the best strategy for near-idle operation for LTC, because of superior fuel economy of 351 g kkh 1, and near complete conversion of UHC and CO by DOC due to lower exhaust flow rate and high exhaust temperatures. 17
Future Work Planning of experiments to verify and validate the cam phaser and fully flexible VVT strategies. Simulation & Experimental studies of cylinder deactivation strategies: 2-cylinder operation (fire either cylinders 1 & 4, or cylinders 2 & 3) Transient 1-cylinder operation that takes into account firing order Noise, Vibration & Harshness (NVH) issues to determine stresses on engine 18
19 University of Wisconsin Engine Research Center