U.S. patent application number 11/380019 was filed with the patent office on 2007-10-25 for combustion diagnostic for active engine feedback control.
Invention is credited to Charles Stuart Daw, Johney Boyd JR. Green, Robert Milton Wagner.
Application Number | 20070250251 11/380019 |
Document ID | / |
Family ID | 38535880 |
Filed Date | 2007-10-25 |
United States Patent
Application |
20070250251 |
Kind Code |
A1 |
Green; Johney Boyd JR. ; et
al. |
October 25, 2007 |
COMBUSTION DIAGNOSTIC FOR ACTIVE ENGINE FEEDBACK CONTROL
Abstract
This invention detects the crank angle location where combustion
switches from premixed to diffusion, referred to as the transition
index, and uses that location to define integration limits that
measure the portions of heat released during the combustion process
that occur during the premixed and diffusion phases. Those
integrated premixed and diffusion values are used to develop a
metric referred to as the combustion index. The combustion index is
defined as the integrated diffusion contribution divided by the
integrated premixed contribution. As the EGR rate is increased
enough to enter the low temperature combustion regime, PM emissions
decrease because more of the combustion process is occurring over
the premixed portion of the heat release rate profile and the
diffusion portion has been significantly reduced. This information
is used to detect when the engine is or is not operating in a low
temperature combustion mode and provides that feedback to an engine
control algorithm.
Inventors: |
Green; Johney Boyd JR.;
(Knoxville, TN) ; Wagner; Robert Milton;
(Knoxville, TN) ; Daw; Charles Stuart; (Knoxville,
TN) |
Correspondence
Address: |
UT-Battelle, LLC;Office of Intellectual Property
One Bethal Valley Road
4500N, MS-6258
Oak Ridge
TN
37831
US
|
Family ID: |
38535880 |
Appl. No.: |
11/380019 |
Filed: |
April 25, 2006 |
Current U.S.
Class: |
701/105 ;
123/435; 123/698; 701/108 |
Current CPC
Class: |
F02D 35/023 20130101;
Y02T 10/44 20130101; F02D 2041/0017 20130101; F02B 77/085 20130101;
Y02T 10/40 20130101; F02D 35/025 20130101; F02D 35/028 20130101;
F02D 41/3035 20130101; F02D 41/0057 20130101; F02D 41/401
20130101 |
Class at
Publication: |
701/105 ;
701/108; 123/435; 123/698 |
International
Class: |
G06F 17/00 20060101
G06F017/00; F02M 7/00 20060101 F02M007/00; F02D 41/00 20060101
F02D041/00 |
Goverment Interests
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH
[0001] This invention was made with United States Government
support under Contract No. DE-AC05-00OR22725 between the United
States Department of Energy and U.T. Battelle, LLC. The United
States Government has certain rights in this invention.
Claims
1. A combustion control apparatus for a diesel engine, comprising;
at least one fuel injector projecting into a combustion chamber of
at least one cylinder of an engine, timing means for adjusting
injection timing of said at least one fuel injector, exhaust gas
recirculation means for adjusting an amount of recirculated exhaust
gas to said combustion chamber, means for measuring dynamic
combustion parameters in said engine, means for determining a
combustion index comprising a ratio between a premixed portion and
a diffusion portion of heat release data, means for controlling
said timing means and said exhaust gas recirculation means in
response to a predetermined range of said combustion index.
2. A combustion control apparatus of claim 1 wherein said dynamic
combustion parameters is at least one parameter selected from the
group consisting of in-cylinder pressure, carbon dioxide, carbon
monoxide, hydrocarbons, nitrogen oxides, oxygen, net heat release
over an entire power stroke, and indicated mean effective pressure
over said entire power stroke.
3. A combustion control apparatus of claim 1 wherein said
combustion index further comprises the sequential steps of: a.
acquiring in-cylinder pressure and dynamic combustion signals, b.
determining dynamic combustion state, c. determining a degree of
transition between said premixed portion and said diffusion
portion, d. calculating said combustion index, e. comparing said
combustion index with predetermined limits, and f. modulating
engine parameters in response to said combustion index.
4. A combustion control apparatus of claim 3 wherein said degree of
transition between said premixed portion and said diffusion portion
further comprises the sequential steps of: a. reading crank-angle
resolved data for heat release over an entire cycle, b. processing
cycle data to find first derivative and second derivative of heat
release function, c. finding a minimum first derivative of heat
release function for said cycle, d. finding a peak heat release
rate, e. finding a second minimum first derivative of heat release
function within 30 degrees after said peak heat release rate, f.
finding next peak first derivative after said second minimum first
derivative, g. finding maximum second derivative between said peak
heat release rate and said peak first derivative, and h. performing
burn mode ratio calculation using integrated premixed portion and
integrated diffusion portion.
5. A combustion control apparatus of claim 1 wherein said means for
controlling further comprises at least one action selected from the
group consisting of modifying injection timing, modifying fuel rail
pressure, and modifying exhaust gas recirculation.
6. A method for calculating a combustion index comprising the
sequential steps of: a. acquiring in-cylinder pressure and dynamic
combustion signals, b. determining dynamic combustion state, c.
determining a degree of transition between premixed and diffusion
dominated combustion, d. calculating a combustion index, e.
comparing said combustion index with predetermined limits, and f.
modulating engine parameters in response to said combustion
index.
7. A method of claim 6 wherein said degree of transition further
comprises the sequential steps of: a. reading crank-angle resolved
data for heat release over an entire cycle, b. processing cycle
data to find first derivative and second derivative of heat release
function, c. finding a minimum first derivative of heat release
function for said cycle, d. finding a peak heat release rate, e.
finding a second minimum first derivative of heat release function
within 30 degrees after said peak heat release rate, f. finding
next peak first derivative after said second minimum first
derivative, g. finding maximum second derivative between said peak
heat release rate and said peak first derivative, and h. performing
burn mode ratio calculation using integrated premixed portion and
integrated diffusion portion.
8. A method of claim 6 wherein said engine parameters further
comprises at least one selected from the group consisting of
injection timing, fuel rail pressure, and exhaust gas
recirculation.
Description
NAMES OF PARTIES TO JOINT RESEARCH AGREEMENT
[0002] This invention was made under a joint research agreement
between UT-Battelle, LLC and Ford Motor Company executed on Jun. 1,
2002. The field of the claimed invention is combustion diagnostics
for diesel engines.
FIELD OF THE INVENTION
[0003] This invention is a combustion control method and apparatus
for improving exhaust gas recirculation (EGR) utilization in
compression ignition direct injection (CIDI) engines that lowers
performance requirements for post-combustion emissions
controls.
BACKGROUND OF THE INVENTION
[0004] Exhaust gas recirculation (EGR) has been used in recent
years to reduce NOx emissions in light duty diesel engines. EGR
involves diverting a fraction of the exhaust gas into the intake
manifold where the recirculated exhaust gas mixes with the incoming
air before being inducted into the combustion chamber. EGR reduces
NOx because it dilutes the intake charge and lowers the combustion
temperature. A practical problem in fully exploiting EGR is that,
at very high levels, EGR suppresses flame speed sufficiently that
combustion becomes incomplete and unacceptable levels of
particulate matter (PM) and hydrocarbons (HC) are released in the
exhaust. This transition to incomplete combustion is
characteristically very abrupt due to the highly nonlinear effect
of EGR on flame speed. In a transient operating environment, it is
particularly difficult to reliably approach this instability limit
without occasionally producing undesirable bursts of HC and PM
emissions. The result is that diesel engines must be typically
operated significantly below their maximum EGR potential, thus
penalizing NOx performance. The effect of EGR on development of
combustion instability and particulate formation was discovered and
options for maximizing the practical EGR limit were identified. The
dynamic details of the combustion transition with EGR and how the
transition might be altered by appropriate high-speed adjustments
to the engine is taught herein. Using the combustion diagnostic of
this invention, one can alter the effective EGR limit (and thus NOx
performance) by using advanced engine control strategies. All
experiments described here were performed on a modern
turbo-charged, direct-injection automotive diesel engine. This
engine was selected on the basis that it is likely to reflect the
EGR response of more advanced diesel engines proposed for
automotive use. The results of this study are applicable to
stationary CIDI engines, especially those experiencing transient
load and/or speed demands.
[0005] In general, a direct-injection diesel engine injects a fuel
into a combustion chamber with high pressure and high temperature
at around the top dead center of the compression stroke of
cylinder, so that the fuel is burned by its self-ignition. Here,
the fuel injected into the combustion chamber proceeds being
divided (atomized) into minute liquid drops by a collision with air
having high density, and forms substantially a cone-shape fuel
spray. The fuel evaporates from surfaces of the fuel drops and
forms a fuel mixture by involving air surrounding mainly a front
end and a periphery of the fuel spray. Then, the fuel mixture is
self-ignited when it becomes a certain condition with its
appropriate concentration and temperature necessary for an
ignition, and begins to burn (premixed combustion). Then, it is
considered that the portion beginning to burn becomes a core and
diffusion combustion is performed involving surrounding fuel vapor
and air.
[0006] In such a normal combustion of the diesel engine
(hereinafter, referred to as diesel combustion), the initial
premixed combustion may be followed by the diffusion combustion
that burns most part of fuel. Here, nitrogen oxides (NOx) is
produced at a portion in which an air excessive ratio .lamda. is
nearly 1 in the fuel spray (fuel mixture) having un-homogeneous
concentration due to a rapid generation of heat. Also smoke is
produced at a portion in which a fuel concentration is too high due
to a lack of air. Conventionally, some measures to reduce NOx or
smoke are taken, such as recirculating part of an exhaust gas into
an intake air (Exhaust Gas Recirculation, hereinafter, referred to
as EGR) and increasing injection pressure of fuel.
[0007] Recirculating the inert exhaust gas into the intake air
system by EGR may suppress production of NOx by decreasing
combustion temperature but, on the other hand, promote production
of smoke with a large amount of EGR decreasing oxygen in the intake
air. Further, increasing injection pressure of fuel may promote
minute fuel spray and improve air utilization rate by increasing
penetration of the fuel spray, resulting in suppression of smoke,
but, on the other hand, it may make a condition where NOx is
produced easily. In other words, the conventional combustion of
diesel engine provided a trade-off relationship on NOx reduction
and smoke reduction, so that it was difficult to reduce both NOx
and smoke coincidentally.
[0008] In contrast, new combustion modes have been recently
proposed that provide a combustion state consisting of premixed
combustion mainly by advancing the timing of fuel injection and
thereby reducing NOx and smoke coincidentally and greatly. These
are generally known as the name of diesel premixed combustion or
premixed compression ignition combustion. This is, for example, a
new combustion mode, in which a large amount of exhaust gas is
recirculated by EGR and a fuel is injected at relatively early
timing of the compression stroke of cylinder to mix with air
sufficiently, so that the premixed mixture is self-ignited at the
end of the compression stroke of cylinder and burns (for example,
as shown in Japanese Patent Laid-Open Publication No.
2000-110669).
[0009] It is preferable that the rate of recirculated exhaust gas
into intake air by EGR (EGR ratio) at such combustion state is set
at a much higher level than that at the above-described diesel
combustion. That is, a large amount of exhaust gas having larger
thermal capacity than that of air is mixed and thereby density of
fuel and oxygen in the premixed mixture is reduced, and as a
result, the timing of self-ignition of the premixed mixture may be
delayed until near the top dead center of compression of cylinder
(TDC) by extending its delay time of ignition. Further, the inert
exhaust gas disperses evenly around fuel and oxygen in the premixed
mixture and absorbs the heat by combustion, and thereby the
production of NOx may be suppressed greatly.
[0010] However, because increasing the rate of recirculated exhaust
gas in the intake air by EGR means decreasing the amount of air in
return, it may be difficult to perform the above-described
combustion at an engine operating area where the engine load is
relatively high. Thus, conventionally, when the engine operation is
at relatively low load, an early fuel injection like the above is
performed and EGR ratio is controlled higher than a first
predetermined value that is relatively high, resulting in premixed
compression ignition combustion. Whereas, when the engine operation
is at relatively high load, fuel is injected at around the top dead
center by changing fuel injection mode, resulting in diesel
combustion.
[0011] In the meantime, in a case where the engine combustion mode
is changeable between the premixed compression ignition combustion
and the diesel combustion, problems exist such as a transient
deterioration of exhaust gas condition at its changing and an
occurrence of large noise. That is, when changed from the premixed
compression ignition combustion to the diesel combustion, the EGR
ratio is changed from a state where it is higher than the first
predetermined value to another state where it is lower than the
second predetermined value, by reducing the amount of recirculated
exhaust gas by EGR. Here, if only fuel injection mode is changed at
once to its injection at around TDC for the diesel combustion, the
combustion consisting of the diffusion combustion mainly is
performed along with an excessive EGR ratio because controlling the
amount of exhaust gas recirculation needs a certain time. As a
result, smoke is produced.
[0012] On the other hand, in changing from the diesel combustion to
the premixed compression ignition combustion, if only fuel
injection mode is changed to its early injection when the EGR ratio
is not sufficiently high, the fuel may be self-ignited at the
too-early timing because adjusting the amount of exhaust gas
recirculation needs a certain time as well. As a result,
considerably large noise of combustion is produced and an increase
of NOx is produced rapidly as well. In addition, a large amount of
smoke is produced by combustion of fuel having an insufficient
mixture with intake air.
[0013] In view of the above-described problems, the present
invention has been devised to diagnose the combustion condition of
diesel by applying an effective algorithm to better sense and
control combustion of a diesel engine in which its combustion is
changeable between a first combustion state where its combustion
consists of premixed combustion mainly (for example, premixed
compression ignition combustion described above) and a second
combustion state where its combustion consists of diffusion
combustion mainly (for example, conventional diesel
combustion).
BRIEF DESCRIPTION OF THE INVENTION
[0014] During high EGR combustion, heat release rate data show that
in the low temperature combustion regime where there are low NOx
and PM emissions, the portion of premixed combustion increases
significantly and the diffusion portion decreases significantly.
This invention detects the crank angle location where combustion
switches from premixed to diffusion, referred to as the transition
index, and uses that location to define integration limits that
measure the portions of heat released during the combustion process
that occur during the premixed and diffusion phases. Those
integrated premixed and diffusion values are used to develop a
metric referred to as the combustion index. The combustion index is
defined as the integrated diffusion contribution divided by the
integrated premixed contribution. As the EGR rate is increased
enough to enter the low temperature combustion regime, PM emissions
decrease because more of the combustion process is occurring over
the premixed portion of the heat release rate profile and the
diffusion portion has been significantly reduced. This information
is used to detect when the engine is or is not operating in a low
temperature combustion mode and provides that feedback to an engine
control algorithm.
[0015] The combustion control apparatus is for a diesel engine
having at least one fuel injector projecting into the combustion
chamber of at least one cylinder of an engine, timing means for
adjusting the injection timing of the fuel injectors, exhaust gas
recirculation means for adjusting the amount of recirculated
exhaust gas, means for measuring dynamic combustion parameters in
the engine, means for determining a combustion index comprising the
ratio between the premixed portion and the diffusion portion of
heat release data, and a means for controlling the timing means and
the exhaust gas recirculation means in response to a predetermined
range of the combustion index.
BRIEF DESCRIPTION OF THE DRAWINGS
[0016] FIG. 1 is graphs showing lower engine-out NOx and PM
observed at higher EGR rates with throttling; Approach 1; 1500 rpm,
2.6 bar BMEP; Production Calibration is 24% EGR.
[0017] FIG. 2 is graphs showing simultaneous low NOx and low PM
observed at lean air-fuel ratios; Approach 1.
[0018] FIG. 3 is a graph showing the Not-so-"classic" PM and NOx
tradeoff at 1500 rpm, 2.6 bar BMEP; Approach 1, including the low
temperature combustion region.
[0019] FIG. 4 is graphs showing significant shift in heat release
observed at higher EGR levels; Approach 1, the transition index,
the premixed region, and the diffusion region.
[0020] FIG. 5 is a graph showing the transition index and PM as a
function of EGR rate.
[0021] FIG. 6 is a graph showing the combustion index and PM as a
function of EGR rate.
[0022] FIG. 7 is a flow diagram for the combustion diagnostic.
[0023] FIG. 8 is a table showing data from attempts to recover load
at high EGR rates revealing significant NOx and PM reduction with
some fuel penalty; COV of IMEP comparable for all cases.
[0024] FIG. 9 is graphs from a Wiebe analysis that reveals
correlation between late-stage combustion and emissions.
[0025] FIG. 10 is a graph showing that late-stage combustion
contributes strongly to emissions at high EGR levels.
DETAILED DESCRIPTION OF THE INVENTION
[0026] First experiments were performed on a 1.9 liter,
four-cylinder Volkswagen turbo-charged direct injection engine
under steady state, low load conditions. Engine speed was
maintained constant at 1200 rpm using an absorbing dynamometer and
fuel flow was set to obtain 30% full load at the 0% EGR condition.
A system was devised to vary EGR by manually deflecting the EGR
diverter valve. The precise EGR level was monitored (on a volume
basis) by comparing NOx concentrations in the exhaust and intake.
NOx concentrations were used because of the high accuracy of the
analyzers at low concentrations found in the intake over a wide
range of EGR levels. In typical experiments to date, fuel flow rate
and injection timing were maintained constant while EGR was
increased. This operating strategy introduces a complication in the
analysis because the engine air-to-fuel ratio is decreased with
increasing EGR due to the displacement of intake air by
recirculated exhaust gases. The effect of this decrease in
air-to-fuel ratio on our observations is discussed below. In future
experiments, fueling adjustments will also be made with EGR changes
to keep the total air-fuel ratio constant. Numerous steady state
and crank angle resolved measurements were made for each EGR level,
including in-cylinder pressure and exhaust gas constituents and PM.
Measurement details are described below.
[0027] High- and low-speed data acquisition systems (DAS) were used
for recording engine measurements. At steady state conditions, the
low-speed DAS was used to record various engine temperatures,
pressures, and engine out emissions. A high-speed DAS developed by
Real Time Engineering (RTE; Dearborn, Mich.) was used to record
crank angle resolved in-cylinder pressure from three cylinders,
exhaust HC concentration, and exhaust particle "density".
[0028] In-cylinder pressure data were recorded using Kistler
piezoelectric pressure transducers mounted with glow plug adapters
in three cylinders. Three thousand cycles of pressure data were
recorded on a crank angle resolved basis for each EGR level using
the RTE system. Integrated combustion parameters (e.g., work and
heat release) were calculated by integrating the in-cylinder
pressure data.
[0029] Steady state measurements were made of CO, CO.sub.2, HC,
NOx, and O.sub.2 concentrations in the raw engine-out exhaust using
Rosemount and California Analytical analyzers. Crank angle resolved
measurements were also made of HC concentration in the exhaust
using a Fast Flame Ionization Detector (Cambustion HFR400Fast FID;
Cambridge, England). The HC sampling probe was located in the
exhaust manifold and the data were recorded using the RTE
system.
[0030] Conventional measurement of automotive exhaust particulates
requires that the exhaust be diluted. Dilution serves two purposes:
it cools the exhaust and it lowers the dew point. In these
experiments, the exhaust was diluted with clean, filtered air to
approximately 35.+-.5:1 by volume. Several instruments were used
for characterizing particulates in the exhaust stream on a steady
state (dilute) and crank angle resolved (raw) basis.
[0031] Dilution of Exhaust: A slipstream of exhaust is fed through
a heat-traced 25 mm stainless steel line to a 100 mm dilution
tunnel. A 250 mm blower is used to pull HEPA-filtered air through
the tunnel, and samples are taken 1 m downstream of the exhaust
inlet to the tunnel. This ensures uniform mixing of the filtered
air and exhaust gases before being sampled by the PM
instrumentation.
[0032] Particle Mass Concentration: A Tapered Element Oscillating
Microbalance (TEOM model 1105; R&P Co., New York, N.Y.) was
used to measure particulate mass concentration and total mass
accumulation as a function of time. A sample of diluted exhaust is
pulled through a 12 mm filter to the end of a tapered quartz
element. The frequency of the element changes with mass
accumulation. The instrument has approximately 3 sec resolution on
mass concentration.
[0033] Particle Size Distribution: A Scanning Mobility Particle
Sizer (SMPS; TSI, Inc., St. Paul, Minn.) was used to measure the
steady state size distribution of the particulates in the exhaust
stream. The SMPS is the scanning version of an Electrical Mobility
Analyzer which is used extensively in aerosol work. The particles
are neutralized and then sorted based on their electrical mobility
diameter. The range of the SMPS was set at 11 nm-505 nm.
[0034] Rapid Particulate Mass Emissions: A Diesel Particle
Scatterometer (DPS) was used to obtain rapid scattering
measurements of a raw exhaust sample taken directly from the
exhaust manifold. The DPS was designed and built by Lawrence
Berkeley National Laboratory. Under normal operation, it measures
size distribution with a 1 sec response time. For this experiment,
we used it to rapidly measure gross quantity of particulates by
monitoring individual signals from its photomultiplier tubes so
that it functioned as a fast "smoke" or particle "density" meter.
The signals were acquired at each crank angle by the RTE
system.
[0035] The previously discussed instrumentation was used to
investigate the effect of EGR level on time-averaged and
time-resolved combustion, emissions, and particulate behavior. When
interpreting the discussion and data, recall that air-to-fuel ratio
decreases toward stoichiometric with increasing EGR level. Also
note that measured torque decreased approximately 15% from the
lowest to the highest EGR level. All gaseous emissions and
particulate data were measured in the exhaust stream before the
catalyst.
[0036] Combustion Characterization with Cylinder Pressure: Net heat
release (HR) and indicated mean effective pressure (IMEP) values
were calculated for each cycle using cylinder pressure measurements
according to well-established definitions. These specific
combustion parameters were selected for the initial analysis
because they are widely used for engine combustion
characterization. However, it is important to note that because HR
and IMEP are evaluated over the entire power stroke, they represent
an integrated assessment of the entire combustion event in the
selected cylinder. Details of the combustion sequence such as
ignition delay or the relative contributions of premixed and
diffusion combustion are typically not clearly revealed by HR and
IMEP. Mean HR and IMEP showed no sudden changes as EGR was
increased, but instead they decreased in a manner consistent with
the overall engine efficiency. Similar trends were also exhibited
by the coefficients of variation (COV) for these quantities, except
that the change was positive with increasing EGR. This slight
increase in COV may be indicative of the onset of combustion
instability. However, the small magnitude of the COV change is in
sharp contrast to the behavior observed for HC, NOx, and PM, which
changed dramatically. Apparently, the flame changes that produce
these emissions are relatively subtle, and such subtle details are
obscured by the pressure integration process. The COV in IMEP and
heat release ranged from 1.0 to 1.8% for all data sets, which is
well within accepted driveability limits. Thus, one would not
expect a driver to "feel" the onset of combustion events that are
bad enough to greatly impact emissions. In addition to
insensitivity due to integration, the HR and IMEP response may have
been reduced because of the constant fueling rate in these
experiments. One would expect that the corresponding decrease in
air-to-fuel ratio would tend to somewhat counter the effect of
dilution and help to restabilize the combustion. Specific pressure
features can be identified that relate to the sudden shift in
emissions described below. When such relationships are identified,
evaluating possible combustion mechanisms with detailed combustion
models such as KIVA (A Hydrodynamics Model for Chemically Reacting
Flow with Spray) which can simulate engine pressure profiles based
on fluid mechanics and chemistry will be possible.
[0037] Combustion Characterization with HC and NOx Emissions:
Time-averaged HC and NOx concentrations in the raw engine-out
exhaust were measured versus EGR level. This data shows NOx
concentration decreasing and HC increasing with increasing EGR as
would be expected. A sudden increase in HC and leveling-off in NOx
occurs at approximately 45% EGR, where there appears to be a
significant shift in combustion chemistry. This major transition is
in sharp contrast to the slight changes observed in the integrated
pressure parameters, HR and IMEP. Because of the suddenness of the
emissions change at 45% EGR, it is clear that dynamic engine
behavior at or above this operating point will be highly nonlinear.
Thus it is imperative that any control strategies being considered
should be able accommodate such behavior. It was not possible to
obtain accurate HC concentration measurements on a crank angle
resolved basis for EGR levels greater than 45% because the Fast FID
sampling probe was fouled by particulates at the higher levels. A
preliminary analysis of the Fast FID data for EGR levels less than
45% indicates no major changes in cycle-resolved HC up to that
point.
[0038] Combustion Characterization with PM: Measurements identified
significant changes in PM emissions with EGR level, as was
expected. Similar to the gaseous emissions (e.g., HC and NOx),
there was a sharp increase in PM at a critical EGR level. This
critical level corresponding to a sharp increase in PM was observed
in mass concentration, particle size, and particle density.
[0039] Mass Concentration: Particle mass concentration and total
mass accumulation were measured on dilute exhaust using the TEOM.
The dilution ratio was maintained at 35.+-.5:1. Mass accumulation
rates were calculated based on over 100 mass data points as a
function of EGR level. Mass accumulation rates begin to increase
significantly at 30% EGR and continue to increase rapidly until the
maximum EGR level. The intersection of the particulate mass and NOx
curves represents a region where the engine out particulate mass
and NOx concentration are minimized for this engine condition.
[0040] Particle Size: Particle sizing was performed on dilute
exhaust using the SMPS. The dilution ratio was maintained at
35.+-.5:1. Number concentration vs. particle diameter was measured
for several EGR levels. Two aspects of the data stand out. The
first is the increasing number concentration with level of EGR. The
second is the increasing particle size. The particle size at the
peak concentration increases by a factor of approximately two
between 30% and 53% EGR.
[0041] The likely mechanism for particle growth is the
reintroduction of particle nuclei into the cylinder during EGR. The
recirculating exhaust particles serve as sites for further
condensation and accumulation leading to larger particles. A
significant fraction of the measured size distribution appears
larger than the 500 nm upper bound of the SMPS for the highest EGR
rates. This is significant because these particles contain much of
the exhaust particulate mass. While exhaust dilution tunnels often
are the source of artifact in the measurement of ultrafine
particles, the effects are greatest for low dilution (<15:1) and
the smallest particles (<20 nm).
[0042] A frequency plot was used to illustrate the disappearance of
small particles and the growth of much larger particles. The
divergence between the curves for particles >100 nm and
particles 60-100 nm increases significantly at 30% EGR and
continues to increase. The data appear to show that the smallest
particles are contributing to the growth of the largest ones. The
increase in larger particles is less steep than the increase in
particle mass. The particle mass, however, increases as a function
of the cube of particle diameter, and thus can be expected to
increase more rapidly.
[0043] Rapid Particle "Density": Fast particle "density"
measurements were made in the raw exhaust using the DPS. Recall
that the photo-multiplier tube (PMT) voltage is effectively a
measure of "smoke" or particle "density". A preliminary analysis of
the data indicates no significant cycle-to-cycle variations in mean
particle "density". While the data are preliminary, this instrument
in conjunction with fast emissions and in-cylinder pressure
measurements is expected to be very useful for improving our
understanding of PM formation.
[0044] The results of this investigation give insight into the
effect of EGR level on the development of gaseous emissions as well
as mechanisms responsible for increased particle density and size
in the exhaust. The time-averaged gaseous emissions results were
similar to those seen in other studies. Particle sizing data showed
some of the most interesting results. The results indicate that it
is possible to directly measure particle growth of the exhaust
particulate mass during high rates of EGR. This growth is likely
caused by the recirculating exhaust particulates serving as
nucleation sites for further particle growth. The DPS fast particle
"density" instrument demonstrated in this study will be used
extensively in future studies with fast emissions and in-cylinder
pressure measurements to improve our understanding of PM formation
and growth. The observation of only slight changes in the
conventional engine combustion parameters of HR and IMEP at high
EGR indicates that these integrated quantities are not adequate for
monitoring the combustion processes responsible for the increased
emissions. Quantities related to specific crank-angle-resolved
details of the cylinder pressure trace are used for the diagnostic
herein.
[0045] Improved exhaust gas recirculation (EGR) utilization in
compression ignition, direct injection (CIDI) engines can lower the
performance requirements for post-combustion emissions controls.
The combustion diagnostic can potentially have >50% NOx
reduction with minimal PM penalty via aggressive use of EGR in CIDI
engines. Development of diagnostic tools and/or feedback control
strategies to allow closed-loop control of CIDI engines in low NOx
and low PM regimes will improve utilization of the diagnostic.
[0046] Actual EGR utilization is typically less than optimal
because of PM and HC emissions. Contributing factors are
cycle-to-cycle and cylinder-to-cylinder variations in combustion,
mixing, and EGR. Extensive experiments were performed under high
EGR conditions with full-pass engine control. There is potential
for recovery of BSFC/BMEP penalty using the combustion
diagnostic.
[0047] The engine platform used for these subsequent experiments
was a Mercedes 1.7 L common rail diesel engine with cooled EGR. The
engine had full-pass control (modified SwRI RPECS system) of
electronic throttle, EGR valve, and fueling parameters.
Instrumentation for each individual cylinder included in-cylinder
pressure, regulated steady-state gaseous emissions, fast HC
emissions, PM size and mass (TEOM, SMPS, Diesel Particle
Scatterometer).
[0048] Two approaches for reducing engine-out NOx and PM at high
EGR levels were discovered. Approach 1 utilized EGR valve control,
sweeping the EGR rate until the EGR valve was fully open and
employing intake throttling to further increase EGR rate with EGR
valve fully open. Approach 2 utilizing EGR valve control, sweeping
the EGR rate until the EGR valve fully opened, and retarding
injection timing with EGR valve fully open. Lower engine-out NOx
and PM were observed at higher EGR rates with throttling
[0049] As stated previously, Approach 1 utilizes a throttle to
increase EGR rate beyond the maximum rate possible with only the
EGR valve. EGR rate is swept with the EGR valve until the maximum
rate achievable with the valve for the particular engine condition.
While maintaining the EGR valve fully open, an intake throttle is
utilized to further increase EGR rate. All other engine conditions
including the injector parameters were held constant. Our
combustion diagnostic detects shift in combustion corresponding to
a drop in PM and NOx emissions observed during low temperature
combustion.
[0050] The production calibration of the engine utilizes an EGR
rate of 24% EGR. The maximum rate of EGR using only the EGR valve
is approximately 44% in this engine under the specified base
conditions. A steady decrease in NOx emissions was observed with
increasing EGR rate. PM emissions increased with EGR rate until
approximately 44%. At this point, a further increase in EGR rate
resulted in a significant decrease in PM emissions. The EGR rate
where the PM maximum was achieved ranged from 42% to 44% and
appeared to be somewhat dependent on the humidity of the combustion
air. Combustion air quality was not controlled and varied
significantly between the days on which these experiments were
repeated. Achieving simultaneously low NOx and low PM resulted in a
penalty (or increase) in brake specific fuel consumption (BSFC).
Attempts at the recovery of this penalty were successful. The low
NOx and low PM combustion regime was also accompanied by an
increase in unburned HC emissions. It is important to note that all
of the data shown for Approach 1 were for conditions lean of
stoichiometry. Intake temperatures increased from 19.degree. C. to
61.degree. C. and exhaust temperatures increased from 221.degree.
C. to 245.degree. C. over the EGR range of 24% to 51%,
respectively, under the 1500 rpm and 2.6 bar BMEP base condition.
The increase in exhaust temperatures and HC emissions observed with
increasing EGR rate may be beneficial to the conversion efficiency
and regeneration of NOx adsorber catalysts.
[0051] Approach 2 uses a combination of EGR and injection timing to
explore the potential for achieving a simultaneous reduction in NOx
and PM emissions. EGR rate is set at the maximum rate possible by
maintaining the EGR valve at its fully open position. No intake
throttle is used with Approach 2. With the EGR valve fully open,
injection timing was retarded from the production calibration
timing. All of the results discussed here correspond to 1500 rpm
and 2.6 bar BMEP with EGR rate held fixed at 44%. The time duration
between the pilot and main was held constant during the injection
timing sweep. A preliminary investigation was performed on the
effect of the relative timing of the pilot and main and will be
discussed in an upcoming report. Simultaneous low NOx and low PM
were observed at lean air-fuel ratios using Approach 2. There was
also a significant shift in heat release observed at retarded
injection timings. Trends in combustion and emissions were similar
to those seen for Approach 1. Attempts to recover load at high EGR
rates using this approach revealed significant NOx and PM reduction
with some fuel penalty.
[0052] The combustion diagnostic of this invention performs the
following logic:
[0053] During high EGR combustion, heat release rate data show that
in the low temperature combustion regime where there are low NOx
and PM emissions, the portion of premixed combustion increases
significantly and the diffusion portion decreases significantly.
This invention detects the crank angle location where combustion
switches from premixed to diffusion, referred to as the transition
index, and uses that location to define integration limits that
measure the portions of heat released during the combustion process
that occur during the premixed and diffusion phases. Those
integrated premixed and diffusion values are used to develop a
metric referred to as the combustion index. The combustion index is
defined as the integrated diffusion contribution divided by the
integrated premixed contribution. As the EGR rate is increased
enough to enter the low temperature combustion regime, PM emissions
decrease because more of the combustion process is occurring over
the premixed portion of the heat release rate profile and the
diffusion portion has been significantly reduced. This information
is used to detect when the engine is or is not operating in a low
temperature combustion mode and provides that feedback to an engine
control algorithm.
[0054] While there has been shown and described what are at present
considered the preferred embodiments of the invention, it will be
obvious to those skilled in the art that various changes and
modifications can be made therein without departing from the
scope.
* * * * *