GASOLINE COMPRESSION IGNITION – A PROMISING TECHNOLOGY TO MEET FUTURE ENGINE EFFICIENCY AND EMISSIONS TARGETS drhgfdjhngngfmhgmghmghjmghfmf STEPHEN CIATTI Principal Mechanical Engineer Argonne National Laboratory 2 nd CRC Advanced Fuels and Engine Efficiency Workshop Wednesday, November 2, 2016 Work funded by DOE Office of Vehicle Technologies – Gurpreet Singh and Leo Breton This presentation does not contain any proprietary, confidential or otherwise restricted information KHANH CUNG Postdoctoral Fellow Argonne National Laboratory
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
GASOLINE COMPRESSION IGNITION – A PROMISING TECHNOLOGY TO MEET FUTURE ENGINE EFFICIENCY AND EMISSIONS TARGETS
drhgfdjhngngfmhgmghmghjmghfmf
STEPHEN CIATTIPrincipal Mechanical EngineerArgonne National Laboratory
2nd CRC Advanced Fuels and Engine Efficiency WorkshopWednesday, November 2, 2016Work funded by DOE Office of Vehicle Technologies – Gurpreet Singh and Leo Breton
This presentation does not contain any proprietary, confidential or otherwise restricted information
KHANH CUNGPostdoctoral FellowArgonne National Laboratory
or GCI utilizing LTC process to reduce soot formation, and avoid high NOx (more premixing and dilution) GCI has shown to have advantages over
traditional diesel compression ignition due to: low soot and emission, high thermal efficiency Initial hardware tends to influence desired
strategies and tradeoffs Many dependent variables (Pin, Tintk, mixture
stratification) on the ignition, combustion and the emissionRecently explored: Effect of multiple parameters (injection pressure, mixture composition, boost,
injection timing, and fuel reactivity) on engine outputs (ignition, combustion phasing, emission, etc.) using design of experiment approach
Distinguish the transition from “quasi” HCCI to GCI as injection timing is swept with constant air/fuel mixture for two different fuels
Kitamura, T., et al. International Journal ofEngine Research 3.4 (2002): 223-248.
ENGINE SPECIFICATION4-cylinder light duty diesel engine (Euro IV
General Motors 1.9 L)Diesel fuel injection system (Bosch CRIP2)Eaton supercharger equipped for study with
Numbers of hole 7Nozzle diameter 0.141 mmSpray angle 120 deg.
Injection pressure 400 to 600 bar
Graphics courtesy ORNL (Curran & Dempsey)
OBJECTIVES OF WORK - MULTI-CYLINDER, HIGH EFFICIENCY GASOLINE COMPRESSION IGNITION
Current Specific Objectives:1. Evaluate effect of Low Pressure EGR upon auto-ignition
and engine performance characteristics 2. Quantitatively study effect of injection strategy upon auto-
ignition to develop approach for transient operation and reduced fuel sensitivity
3. Perform factorial experiments to quantify the effect of important input parameters upon engine performance, noise and emissions
4
Long-Term ObjectiveUnderstand the physical and chemistry characteristics ofGasoline Compression Ignition (GCI) in a multi-cylinder engine toaid industry in developing a practical high efficiency, lowemission combustion system
Mixing Limited GCIHCCI PFS
Majority Premixed GCI
Majority Stratified GCI
INJECTION STRATEGY: E10 MINIMUM FUELING – FULL SOI SWEEP (-141 ATDC TO TDC)
5
SOI [deg. aTDC]
-140 -120 -100 -80 -60 -40 -20 0
Lam
bda
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
Lambda
Minimum Fueling approach: least fuel requirement for stable combustion (COVIMEP < 3%)
Combustion mode (HCCI vs. GCI) characterized by SOIs
Discontinuity due to fuel rate changed
A-period Location of ignition (CA10), and combustion phasing (CA50) seems stay constant
B-period More fuel (smaller lambda) is needed to have stable combustion, but CA10/CA50 also seems constant
C-period IMEP shows a drop near -60 deg. aTDC due to possible fuel entering squish region, as indicated by simulation
A B C
Fuel in squish region
Both SOI and Lambda show to have effect on CA10/CA50, but lambda is more effective
There seems to be a condition with constant lambda to fix CA10/CA50
GM 1.9 L 17.8:1 (CR)Engine speed 1000 rpm
Injection pressure 400 bar
Injector-Bosch
7 hole 120 deg.
cone angleFuel E10
Constant lambda at some SOI range
SOI [deg. aTDC]
-140 -120 -100 -80 -60 -40 -20 0C
A [d
eg. a
TDC
]-10
-5
0
5
10
15
20 CA10CA50CA90
E10 CONSTANT LAMBDA – SOI SWEEP
Early to near TDC SOI effect
Minimum fuel provides fuel requirement (least fuel) for combustion stability, but does not give same mixture to study SOI effect explicitly on ignition and combustion constant lambda approach
Fuel rate was adjusted to keep same lambda through all SOIs
λ calculated from emission bench
“Sweet spot” (early ignition)
High noise level near
TDC
quasi HCCI Transition GCI
Less fuel in squish region
Almost similar ignition location in “quasi HCCI” (also similar CA50/CA90)
Existing region where earliest ignition occurs (near -30 deg. aTDC) – reduction in fuel in squish region
Near TDC, short residence time for ignition
IMEP increases slightly near TDC (less fuel in squish)
It was harder to control noise level with late injection
very lean combustion (λ = 4.4)
6
COVIMEP<3%, max = 5%)
LOW SPEED/LOAD E10 CONSTANT LAMBDA – SOI SWEEP: EMISSION
Early injection: Low level of NOx in highly
homogeneous mixture (quasi HCCI region)
Incomplete combustion (high HC)
Low combustion temperature high CO
Smoke number (FSN) was very low (<0.1) due to lean condition at all SOIs
Late injection (GCI mode): Insufficient time for air/fuel mixing High NOx due ~ high combustion
temperature; richer zones for ignition
Less HC and CO (aggressive reaction leads to more complete combustion)
Injector Current and HRR of 5 bar and 8 bar at 2000 RPM with ~30% EGR (NOx<0.45 g/kWhr, CO<1 g/kWhr, HC<3g/kWhr, FSN < 0.03)
20160404_000
20160404_001
20160404_002
20160405_002
20160405_003
HRR_0404_000
HRR_0404_001
HRR_0404_002
HRR_0405_002
HRR_0405_003
SOI OF MULTIPLE INJECTION AND HRR: 5 BAR AND 8 BAR BMEP
11
SO
I 1st
= -1
00 a
TDC
SO
I 2nd
= -7
0 aT
DC
SO
I 3rd
= -2
5 aT
DC
Pinj = 600 bar5 bar
(48 C, 1.37 bar)
8 bar(48 C, 1.63
bar)
CAD
0 10 20 30 40 50
HR
R [J
/CA
D]
-20
0
20
40
60
80
100800 bar SOI (47/29/16)600 bar SOI (100/70/25)
INJECTION PRESSURE & SOI EFFECTIVELY CONTROL COMBUSTION PHASING
12
CA10/50/90 delayed by 4
deg
Retarded SOI further delay ignition and combustion phasing: this results in slight reduction of noise, but most of all relax the constraint of EGR
Higher peak pressure near TDC for late SOI case implies higher heat loss. This resulted in lower performance of BSFC (increase by 22 g/kW-hr). Adjustment in 1st
and 2nd injection and/or increase in EGR would possibly help
CHALLENGE IN FUNDAMENTAL STUDY OF SOI EFFECT
13
Fueling rate varied when changing SOI Injector per cylinder needs to adjusted SOI duration to
match load, combustion phasing
Slight change in 3rd
injection timing (0.2 deg) could make a difference in noise level (~4 dB level)
CAD
-5 0 5 10 15 20 25 30 35N
orm
aliz
ed M
FB
0.00.10.20.30.40.50.60.70.80.91.01.1
Cyl 1Cyl 2Cyl 3Cyl 4
LP EGR SETUP & TESTING CONDITIONLP EGR Adjustment by means
ExhaustValve
Throttle valve on overall exhaust discharge
Inlet ValveThrottle on fresh intake air upstream of turbo, to drive LP-EGR
LP EGR Valve
Throttle valve between post DPF exhaust and turbocharger intake
Test condition: EGR% sweep at constant load
(BMEP ~ 3 bar at 2000 RPM) EGR% adjusted by separate valves
(most effective exhaust valve) Triple injection (SOI of each: 100-70-
25 deg. bTDC) Supercharger (ON) for P intake = 0.6
bar• Allows for precise control of
intake pressure!
Inj 1 Inj 2 Inj 3
Pinj = 400 bar
14
Smoke meter
DPF
AMA2000
Lambda sensor
RECENT RESULTS WITH LP-EGR: APPROACHING AN OPTIMAL INJECTION STRATEGY FOR LOW NOISE, BSFC AT FULL LOAD RANGE
15
S/C OFF (new test) Since start of injection is very in
controlling the combustion phasing; noise is also reduced
EGR was even reduced from 30% to 21% even for late injection
Higher EGR may be required to maintain location of combustion phasing
Advantage of using higher injection pressure for fuel atomization and promote mixing to achieve lean combustion (lambda ~ 1.6-1.9); more stable combustion
Study at UCB (Vuillemier) shows that LTHR has significant effect on gasoline HCCI ignition
His conclusions indicate:
A fuel’s Octane Index is a good indicator of its GCI Low-Load Performance
LTHR Onset Pressure in an HCCI engine correlates very well with GCI Low-Load Performance.
Increased intake pressure increases low-temperature heat release, enabling lower loads in a GCI engine
COLLABORATION - UCB WORK SHOWS SIGNIFICANT LTHR DEPENDENCE FOR GCI
Courtesy of David Vuilleumier (August 2015 AEC meeting)
SUMMARY New approach to study the ignition of E-10 in GCI engine using “constant lambda” to
investigate ignition behavior Combustion modes were seen distinctively from CA10/50/90 vs SOI plot: “quasi”
HCCI, transition, and GCI modes Design of experiment (DOE) provide overall view of different parametric study using
minimal number of test runs: constant lambda, constant boost– Boost, lambda, and SOI affect strongly on ignition, combustion phasing, and emission– Injection pressure has minimal effect at low speed; but contribute significantly at high speed
in controlling NOx/HC/CO, and combustion stability
Installation of the new LP-EGR has been evaluated with benefit of using smoke-filtered and cooled EGR to effectively control the combustion phasing, therefore noise and emission– Higher EGR decrease combustion noise but fuel consumption is still high due to incomplete
combustion– Reduced EGR required for 5 bar BMEP was achieved by retarding SOIs. This was
performed with higher injection pressure with the same emission output (low emission, and ultra-low smoke)
– Combustion phasing was further delayed to facilitate with combustion noise– The SOI of the 3rd injection (determine load and combustion phasing the most) is very
sensitive to both noise and BSFC. Further investigation is needed to identify the flexibility when changing higher load/speed
18
FUTURE WORK Continue to explore/understand effect of injection strategy
upon GCI operation E10 is sensitive to both boost and EGR
Explore more conditions with LP-EGR Provide more boost at low speeds/loads with EGR
Examine influence of CR upon combustion noise Alter IVC relative to exhaust cam Use lower CR piston crowns (we have 16, 15 and 14:1)
Continue to develop strategy for transient operation with injection, boost and EGR
Continue to track and account for USCAR guidelines combustion noise – later injections at higher pressure show significant promise! Target <90 dB for high load, <85 dB for low load
Continue to characterize GCI particulate emissions TEM sampling and analysis19