Mixed-mode Diesel HCCI with External Mixture Formation:
Preliminary ResultsShawn Midlam-Mohler, Yann Guezennec, Giorgio Rizzoni
The Ohio State UniversityCenter for Automotive ResearchMichael Bargende, Simon Haas
FKFS, Universität [email protected]
http://car.eng.ohio-state.edu
DEER 2003Newport, August 26th, 2003
Outline
1. Mixed-mode HCCI/DI concept2. Experimental results in single-cylinder
engine3. Future Work
1. Mixed-mode HCCI technologies
• Single fuel system: Diesel• Two injection systems
– Port/manifold injection – low pressure atomizer system– Direct injection - high-pressure injection system
• Atomizer system delivers fuel as a pre-mixed lean homogeneous mixture in the cylinder.
• Homogeneous charge ignites due to compression, or alternatively due to triggering from direct injection pulse– Allows more control over HCCI SOC
Basic concept
Characteristics• Small droplet size (< 1µm mean diameter) allows
rapid evaporation during compression stroke, removing the need for intake air heating (leading to higher CR).
• EGR and valve actuation control (based on models of combustion delay and reaction rates) permits SOC control.
• Direct injection supplements the reaction with additional fuel as a function of load for high-torque output.
Engine Operation
• Low load– Main torque from
homogeneous charge (HC) fuel
– Direct Injection (DI) mainly for ignition
• Mid load– Increasing HC fuel– Increasing DI fuel
• High load– Max HC fuel– Increasing DI to full load
SPEED
LOA
D
Increasing DI fuel
Increasing HC fuel
Basic Injection Scheme
Diesel HCCI Methods
• External Mixture Formation
– Port-Injection/Fumigation: fuel is injected in the intake air as it enters the cylinder
• Internal Mixture Preparation
– Early In-Cylinder Injection: Fuel is injected in the cylinder well in advance of TDC
– Late In-Cylinder Injection: Fuel is injected in the cylinder near TDC
HCCI combustionwith External Mixture Preparation
• Advantages:
– Utilizes turbulence at intake port to promote mixing; very homogenous charge
– For mixed mode combustion, the conventional direct injection fuel system can be optimized for just direct injection operation; it does not have to be a compromise between the different requirements of HCCI injection and DI injection.
• Disadvantages:
– Diesel fuel difficult to atomize with conventional atomization techniques
– Wall wetting can lead to high HC and smoke emissions and oil dilution
Generation 1 Atomizer
Maximum Power Consumption 320 W
Maximum Flow Rate 20 mL/min
Environmental Temperature Limits:
Lower Tested to -5 deg. C with no loss of performance
Upper Not evaluated; no problems are expected
Physical Dimensions 1” x 1” x 4”
Fuel Pressure Requirement < 40 psi
System “warm-up time” 5 seconds
Power source 12 VDC
Atomization Quality
T = 0 min.
T = 2 min.
T = 8 min.
2 mL of diesel in a 2 L flask at room temperature
Droplet Size Distribution
-10
0
10
20
30
40
50
1 10 100
Vol
ume
Perc
ent
particle size [microns]
Sample Combustion Quality
The fine atomization allows for spark ignition of diesel fuel
10 kW premixed, diesel flame, AFR = 16 burning at atmospheric conditions
Generation 2 Atomizer
Status of the Technology• Patent pending• Gen. 3 prototypes available for testing in September
2003– Technology is still under development– Current system is suitable for laboratory testing
• Improvements to be made in:– Reducing power consumption– Improving flow rate control– Reducing size
• Cost to mass produce has not been evaluated– Estimated to be between the cost of a gasoline port fuel
injector and a direct injection diesel fuel injector
• Low momentum particles follow air streamlines– Minimal wall impingement = low wall wetting
• Lower HC, smoke, and reduced oil dilution
– Mixing properties are very good in turbulent flows
• Can effectively atomize diesel fuel at low temperatures
• Suitable for injection of water- in- diesel emulsion
The atomizer and HCCI mixture preparation
2. Single-cylinder engine experiments
Experimental study
• Conducted in collaboration with Forschungsinstitut für Kraftfahrwesen und Verbrennungsmotoren, FKFS, UniversitätStuttgart (Prof. M. Bargende).
• Single cylinder engine in controlled environment
• First round of tests completed in Spring 2003.
Cylinders 1
Valves 4
Displacement 537,7 cm^3
Bore 88,0 mm
Stroke 88,4 mm
Connecting Rod 149 mm
Geometric CR 18 : 1
Nozzle Type: 6 Holes
Injection System: Common Rail System 1350 bar
Single-Cylinder Engine OM 611
Single-cylinder engine
Single-Cylinder Test Cell at FKFS
EGRcontrol valve
exhaust back-pressure
control valve
exhaust gas analyser
smokemeter
atomizer
damping volume
filter
supercharger
damping volumes
heatexchanger
coolant conditioning system
intake port deactivation
oil conditioning system
combustion air temperature control unit
mass flow meter
Effect of uncooled EGRR
ate
of H
eat R
elea
se [J
/°CA
]
5
15
25
35
45
Crank angle [°CA]140 150 160 170 TDC 190 200
without EGR 35 % EGR 58 % EGR
HC = 1220 ppmNOx = 7 ppmFSN = 0.53sigma pmi = 0.16
HC = 1230 ppmNOx = 4 ppmFSN = 0.41sigma pmi = 0.14
HC = 1350 ppmNOx = 3 ppmFSN = 0.50sigma pmi = 0.17
σ
dQ
___
dθ
KJ/
deg
Effect of Boost PressureR
ate
of H
eat R
elea
se [J
/°CA
]
0
5
10
15
20
25
30
35
40
45
Crank angle [°CA]140 150 160 170 TDC 190 200
p2 = 1090 mbar p2 = 1200 mbar p2 = 1400 mbar
HC = 1220 ppmNOx = 7 ppmFSN = 0.53
HC = 1170 ppmNOx = 3 ppmFSN = 0.18
HC = 840 ppmNOx = 6 ppmFSN = 0.29
dQ
___
dθ
KJ/
deg
Effect of AFR Increase on IMEP
RateofHeatRelease[J/°CA]
10
30
50
70
Crank angle [°CA]140 160 TDC 200 220 240
linerpressure[ba]
405060708090100
IMEP = 1.6 barIMEP = 3.3 barIMEP = 4.0 bar
No EGR
pressure oscillation: light knock
dQ
___
dθ
KJ/
deg
Indicated
Pressure
Extension of Load RangeR
ate
of H
eat R
elea
se [J
/°CA
]
0
10
20
30
40
50
60
70
80
90
100
Crank angle [°CA]140 150 160 170 TDC 190 200
IMEP = 4.0 bar, without EGR, light knock IMEP = 4.7 bar, 60 % EGR, low noise
HC = 1250 ppmNOx = 35 ppmFSN = 1.64
HC = 2450 ppmNOx = 3 ppmFSN = 0.83
dQ
___
dθ
KJ/
deg
Effect of Intake Air Temperature
Rat
e of
Hea
t Rel
ease
[J/°C
A]
0
5
10
15
20
25
30
35
40
Crank angle [°CA]140 150 160 170 TDC 190 200
T2 = 15 °C T2 = 30 °C T2 = 60 °C
HC = 1030 ppmNOx = 4 ppmFSN = 0.23
HC = 1022 ppmNOx = 5 ppmFSN = 0.27
HC = 1050 ppmNOx = 7 ppmFSN = 0.30
dQ
___
dθ
KJ/
deg
Effect of Swirl (Intake-Port Deactivation)
Rat
e of
Hea
t Rel
ease
[J/°C
A]
0
5
10
15
20
25
30
35
Crank angle [°CA]140 150 160 170 TDC 190 200
low swirl, basis high swirl, boost pressure = const. high swirl, air mass = const.
HC = 940 ppmNOx = 13 ppmFSN = 0.98
HC = 1070 ppmNOx = 12.7 ppmFSN = 0.44
HC = 1090 ppmNOx = 11.5 ppmFSN = 0.45
dQ
___
dθ
KJ/
deg
Effect of Engine Speed (Crank Angle Domain)
Rat
e of
Hea
t Rel
ease
[J/°C
A]
0
5
10
15
20
25
30
35
40
45
Crank angle [°CA]140 150 160 170 TDC 190 200
1600 rpm 2000 rpm 3000 rpm
HC = 860 ppmNOx = 5 ppmFSN = 0.68
HC = 1100 ppmNOx = 8 ppmFSN = 0.64
HC = 1120 ppmNOx = 9 ppmFSN = 0.54
dQ
___
dθ
KJ/
deg
Effect of Engine Speed (Time Domain)
-5
0
5
10
15
20
25
30
35
40
45
50
-4 -3 -2 -1 0 1 2 3 4
Time [ms]
Rat
e of
Hea
t Rel
ease
[J/°C
A]
1600 rpm2000 rpm3000 rpm
TDC
Cyclic variability (100 cycles)R
ate
of H
eat R
elea
se [J
/°CA
]
0
10
20
30
40
50
60
70
80
90
Crank angle [°CA]140 150 160 170 TDC 190 200 210 220
n = 2000 rpmIMEP = 4.7 barEGR Rate: 60%σ (IMEP) = 0.06
dQ
___
dθ
KJ/
deg
Goals for atomizer development• Generation Three Atomizer with:
– Flow rate sufficient to fuel a single cylinder HCCI engine at at all engine speeds
– Closed loop fuel control system with high bandwidth control
– 50% reduced power consumption
– Appropriate packaging for port injection
• Development of a detailed thermo-fluid dynamic model of the atomizer system to support the above efforts
3. Future work
Atomizer characterization• Proposed work in collaboration with Pacific
Northwest National Laboratory (Dan Imre, Alla Zelenyuk-Imre):– Characterize the physical and chemical properties
of the spray formed by the Ohio State novel atomizer.
• Radial and axial variation in droplet size distributions and number concentrations will be measured as a function of an array of controllable atomization parameters: flow rate, electrical current, fuel type, fuel viscosity, ambient pressures and temperatures, etc.
• These data will provide information on the kinetics of spray propagation, coagulation, and gas to particle partitioning to yield predictive relationships between atomizer operation and spray properties.
• Similar collaborations with Sandia, ORNL?
Additional single-cylinder engine tests
Multi-cylinder engine tests• Implementation on a multi-
cylinder, direct-injection engine in engine test cell.
• Investigation of steady state engine characteristics with mixed mode HCCI.– Torque production, emissions,
fuel consumption
Fiat JTD 2.4 liter I-5 engine
VM-Motori/DDC 2.5 liter I-4 engine
Vehicle tests
FutureTruck 2004, Ford Motor Co. Michigan Proving Ground, Romeo, MI
