The 13

advertisement
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
Characterization of Methanol and Ethanol Sprays Using Mie-scattering and
Laser Induced Fluorescence under Engine Cold-start Conditions
Ming Zhang1, Min Xu1, Wei Zeng1, Gaoming Zhang1, Yuyin Zhang1, David J Cleary2
1
Institute of Automotive Engineering, Shanghai Jiao Tong University, Shanghai200240, E-mail: mxu@sjtu.edu.cn
2
Powertrain Systems Research, GM China Research & Development, No 99 Fucheng Rd., Shanghai 200120, China
Abstract
An instantaneous LaVision laser sheet imaging system with Mie scattering and Laser Induced Fluorescence (LIF) were
used to investigate the spray characteristics of gasoline, methanol and ethanol fuels. The sprays from a high-pressure
swirl DI injector were observed in a constant volume pressure chamber. A combination of the second harmonic (532nm)
and the fourth harmonic (266nm) was generated simultaneously by an Nd:YAG laser system to illuminate the spray. By
adding fluorescence dopants (TEA) in the methanol and ethanol and using a band-pass filter with a centerline
wavelength of 378nm for fluorescence, the vapor phase signal can be captured, while the scattered light from the liquid
phase is transmitted through a Mie filter with a centerline wavelength of 532 nm. By using the image doubler, both the
Mie and the LIF signals from the same spray are simultaneously recorded by a CCD camera. As a result, the methanol
and ethanol sprays were characterized and compared with the gasoline spray at different injection timing (ASOI).
Moreover, the Mie and LIF images of the same spray were used to analyze the effects of ambient pressure and fuel
temperature on the distribution of liquid and vapor phase qualitatively.
Keywords: Mie-scattering, Laser Induced Fluorescence (LIF), spray, evaporation, methanol, ethanol
1. Introduction
Alcohol fuels, especially ethanol and methanol, have
higher octane numbers, broader flammability limits,
higher flame speeds and higher latent heats of
vaporization compared to gasoline [1]. These properties
allow for a higher compression ratio, shorter burn time
and cleaner combustion, which lead to thermal
efficiency advantages over gasoline in the internal
combustion engines. However, there are also many
challenges which affect the performance and practical
implementation of alcohol fuel engines due to the
differences between alcohol fuels and gasoline. In
particular, difficult cold start remains as one of the
major concerns.
The poor performance and high engine-out HC
emissions during the engine cold start are due to the
poor spray atomization and evaporation. Such a
phenomenon is more severe for methanol and ethanol
fuels. Compared to gasoline, the atomization process of
methanol and ethanol fuels was a little slower and the
droplet size was slightly larger because of the higher
viscosity [2, 3]. The evaporation of methanol and
ethanol fuels is much slower than that of gasoline
during cold-start because the lower vapor pressures and
the higher latent heats of vaporization [4]. Therefore,
the optimization strategies for alcohol fuel spray
atomization and evaporation should be investigated.
To improve the engine cold-start performance, a direct
injection spark ignition engine is anticipated to have
advantages over the port-fuel injected engines [5, 6].
The DISI engine exhibits a rapid rise in the IMEP
following the first or second injection event, whereas
the PFI engine requires about more than 10 cycles for
the engine to attain stable combustion [7]. This is
attributed to the more accurate fuel injection mass, less
cycle-to-cycle variation, higher compression ratio and
higher injection pressure of DISI engines compared to
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
PFI engines. Therefore, direct injection is expected to
following the excitation by a laser beam. The
mitigate some of the vaporization challenges when
fluorescent agents in the multi-composition fuels such
using methanol or ethanol fuels.
as gasoline, or a particular fluorescent dopant dissolved
However, there are still some issues for the cold-start of
in a non-fluorescent fuel such as methanol, can be used
DISI alcohol fuel engines. First, the fuel should be
to visualize liquid fuel and fuel vapor simultaneously.
injected into the cylinder at a high pressure to improve
This is because the absorption and fluorescence spectra
the atomization and evaporation process, but during
of organic molecules dissolved in nonpolar solvents are
cold-start, the mechanical fuel pump which is supposed
virtually identical to the spectra of the same molecules
to provide the high fuel pressure is unable to
in the vapor phase [9]. Therefore, the LIF signal is
immediately deliver the appropriate fuel pressure for the
applicable for detecting both liquid and vapor phases. In
injector [8]. The fuel injector operates at pressures lower
this study, the Mie-scattering technique was used to
than the designed. Such phenomenon results in a poor
characterize the liquid phase of sprays. The LIF
spray atomization process. And it is also difficult to mix
technique was used to detect both the liquid and vapor
the air and the fuel very well under such fuel injection
phases of sprays. Both Mie and LIF signals were
pressure. On the other hand, the lower vapor pressures
captured simultaneously using the same experimental
and the higher latent heats of vaporization are still the
setup to generate the images of the same spray. The Mie
challenges for the evaporation of alcohol fuels. These
and LIF images of the spray were analyzed to derive the
challenges are more severe in the DISI engines because
distribution of the liquid and vapor qualitatively.
the necessary time of vaporizing the fuel is not enough.
With the laser at 266nm, the gasoline can be excited
Therefore, the injection scheme should be investigated
because of the natural fluorescent components, such as
to optimize the spray atomization formation process.
aromatic compounds and benzene. For alcohol fuels,
The temperature of the fuel or surrounding air should be
TEA (triethylamine) was used as the dopant to visualize
increased to accelerate the evaporation process.
the liquid fuel and vapor [10]. Because of the different
In this study, the sprays of gasoline, methanol and
fluorescence species and concentration, the difference in
ethanol fuels were investigated experimentally to
absorption between gasoline and alcohol fuels is
understand the spray characteristics of these three fuels
significant. The absorption of gasoline is much stronger
under different operating conditions. The study was
than those of alcohol fuels. Therefore, the Mie scattered
focused on the effects of ambient pressure and fuel
light of 266nm of gasoline is weaker compared to
temperature on the sprays structure. The fuel of elevated
alcohol fuels because the Mie scattered ray of refraction
temperatures injected into the vacuum was investigated
is attenuated by the absorption [11]. In this study, the
as
laser at 532nm was used as the incident light for
a
possible
solution
to
improve
the
spray
characteristics under engine cold-start conditions.
Mie-scattering so that the scattered light was not
affected by the absorption. The laser at 266nm was used
as the exciting light for LIF. A coaxial laser sheet
2. Experimental Techniques and Apparatus
consisting of the second harmonic (532nm) and the
Laser Diagnostic Techniques
fourth harmonic (266nm) was generated by a laser sheet
The elastic scattering light from spray particles with the
optics in addition to an Nd:YAG laser system to
wavelength of the incident light, so called Mie
illuminate the spray.
scattering,
can
be
used
for
spray
geometry
determination. However, the Mie signal is only
Apparatus
applicable for measuring the liquid phase, and its
Figure 1 shows the schematic of the experimental
intensity scattered from a droplet is proportional to the
apparatus consisting of a constant volume pressure
total surface area of the droplet. Laser Induced
chamber, a fuel supply system, a fluid temperature
Fluorescence (LIF) is the emission of light emitting
control system, a vacuum pump system and a laser
from an atom of the absorbing species molecule
diagnostic system.
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
Figure 1. Experimental apparatus
A high-pressure swirl DI injector was installed in the
The injected spray was illuminated by a thin laser sheet
constant volume chamber. The constant volume pressure
generated by an Nd:YAG laser (Litron, Pulse Width: 4ns,
chamber for spray visualization has an inner diameter of
Max. Power: 220mj at 532nm, 22mj at 266nm). Images
203mm and a height of 692mm. Four quartz windows
of the illuminated sprays were captured by a LaVision
are installed in the chamber to provide fully optical
instantaneous imaging system. A Programmable Timing
access to the spray. A vacuum system was used to
Unit (PTU) was used to synchronize the laser, camera,
control the chamber ambient pressure. A fluid
and injector systems. The Mie scattering light is
temperature controller and heat exchange system were
separated from the fluorescence light using optical
installed to regulate the fuel temperature from -20°C to
filters (Filter for Mie: BP at 532nm, Filter for LIF: BP at
99°C.
378nm) that are mounted in front of a UV lens (Nikon
Three hydraulic piston accumulators were used to
Rayfact PF10545MF-UV, focal length: 105, f/#=4.5)
pressurize methanol, ethanol and gasoline fuels at a
and a CCD camera (12 bit LaVision Imager Intense,
desired injection pressure and to maintain the pressure
1376×1040 resolution, 15fps recording rate). Both Mie
constant during testing. A flexible fuel supply system
and LIF images were captured simultaneously using an
was designed to allow switching among the three types
image doubler. A LaVision Intensified Relay Optics
of the fuels. With the hydraulic piston accumulators and
(IRO) was used to intensify the signals. After the spray
a high pressure nitrogen bottle, different injection
image is captured by the LaVision high resolution
pressures up to 20.7MPa can be obtained.
camera, the image is post-processed using the software
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
Davis and Matlab.
Experimental Conditions and Testing Fuels
The experimental conditions are shown in Table 1. An
injection pressure of 5MPa was selected as the typical
cold-start fuel injection pressure. The test ambient
pressures were 40KPa and 100KPa, corresponding to
the typical DISI engine in-cylinder pressures at the early
stage of the intake stroke and the late stage of the intake
stoke. The ambient temperature was around 25°C. The
fuel temperature was varied from 25°C to 90°C to
investigate the effect of fuel temperature on the spray
characteristics under engine cold-start conditions.
The test fuels were gasoline (octane number of 97), pure
Figure 2. Temperature dependence of the evaporation
ethanol fuel, and pure methanol fuel. The TEA
properties of the test fuels
(triethylamine) was used as the fluorescent dopant for
methanol and ethanol fuels. The physical properties of
Experimental Procedure
the three fuels and TEA are summarized in Table 2 and
A number of experiments were designed to investigate
Fig. 2.
the spray characteristics in the constant volume pressure
chamber. The ambient pressure was maintained by
Table 1. Experiment conditions
97# gasoline;
Test fuel
controlling
the
vacuum
system.
While
a
fuel
Ethanol (10% TEA vol. %);
temperature controller was used to control the fuel
Methanol (10% TEA vol. %)
temperature. The test fuel was injected into the chamber
through an electronically controlled DI swirl injector.
Injection pressure
5MPa
Back pressure
40KPa, 100KPa
For methanol and ethanol, 10 vol. % TEA (triethylamine)
was used as the dopant,while gasoline contains many
natural fluorescent components and needs no dopant. To
Ambient temperature
25±1℃
balance the simultaneous LIF and Mie signal intensity,
Fuel temperature
25℃,55℃,90℃
an ND filter (OD=0.6) was used in front of the LIF filter
for gasoline test.
Table 2. Physical properties of the test fuels
97#
gasoline
At each test condition, spray images were recorded and
the background was also recorded as a reference which
methanol
ethanol
then was subtracted from the spray images during post
processing. For the purpose of determining the spray
Surface tension
(mN/m, 20 °C)
20-25
22.5
22.39
penetration and angle, a threshold value was used to
distinguish between background noise and fuel spray.
Pixels with intensity below the threshold value were set
Viscosity
0.42/
0.541/
1.052/
(mPa·s,
-/
0.365/
0.6476/
25/55/90℃)
-
0.251
0.4069
Density
0.740/
0.784/
0.782/
(g/mL,
-/
0.7485/
0.746/
25/55/90℃)
-
0.7037
0.702
to zero. The threshold value is selected according to the
SAE standard J2715.
3. Results and Discussion
The spray characteristics for methanol, ethanol, and
gasoline fuels injected from the high-pressure swirl DI
fuel injector were investigated to understand the effect
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
Gasoline
Methanol
Ethanol
100kPa
40kPa
100kPa
40kPa
100kPa
40kPa
100kPa
40kPa
25℃
25℃
90℃
90℃
25℃
25℃
90℃
90℃
Figure 3. Mie (left) and LIF (right) images of sprays at various ambient pressures and fuel temperatures
of fuel temperatures and the ambient pressures on spray
The penetration in the 40KPa ambient is larger than that
structure. Both Mie and LIF images were captured
in atmospheric pressure for all fuels due to the reduced
simultaneously. The difference in the liquid phase
resistance to the spray in the vacuum.
among these three fuels was investigated by Mie images.
And the LIF images were used to analyze the effect of
chamber ambient pressure and fuel temperature on the
both liquid and vapor phases. All images were taken at
0.2ms interval after the start of injection. The laser has
traversed the spray from the right to left side of the
spray.
Effect of Ambient Pressure on Spray Structure
Figure 3 shows the spray Mie and LIF images under
various test conditions. When the chamber ambient
pressure decreased from atmospheric to a vacuum of
40KPa at the same fuel temperature of 25℃, a greater
Figure 4. Spray penetrations at ambient pressures of 40 KPa
penetration and a wider spray angle for all the test fuels
and 100KPa
were observed.
Figure 4 shows the spray penetration comparison of
No significant difference of spray angle among the test
gasoline, methanol and ethanol fuels in such conditions.
fuels was found for the ambient pressure tested in the
At the chamber ambient pressure of 40KPa, the ambient
experiment. Ethanol was observed to have the smallest
density was lower which led to a longer penetration and
spray angle at different ambient pressure at the fuel
a wider spray distribution for all three fuels. The spray
temperature of 25℃. As shown in Fig. 5, there is a
penetration of gasoline was larger than those of alcohol
1-2°spray angle increment for all the test fuels when the
fuels. These results were consistent with the theoretical
ambient pressure varied from 100KPa to 40KPa.
analysis, where the fluid with lower viscosity and
Furthermore,
density were anticipated to have a larger penetration.
comparison of Mie and LIF images in Fig. 6 can be used
take
ethanol
as
an
example,
the
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
to analyze the difference in spray distribution of liquid
differences of all the test fuels at 55℃ and 90℃ fuel
and vapor phase. The difference of the Mie and LIF
temperatures relative to those at 25℃ are shown in
images was insignificant, indicating negligible vapor
Figure 8. When the fuels were heated up to 55℃, the
concentration of the test fuel for the ambient pressure of
penetration rates increase relative to those of 25℃.
40KPa. Fig. 3 indicates that for all three fuels, the spray
However, when further heating the fuel to 90℃, the
atomization and evaporation at the vacuum of 40KPa
penetration rates decrease. The spray Mie and LIF
are much worse compared to the ambient pressure of
images shown in Fig.3 demonstrate totally changed
100KPa at the same fuel temperature of 25℃. Among
spray structure, and enhanced spray atomization and
them, gasoline has the best atomization, liquid phase
evaporation at 90℃, which are considered to cause the
distribution and vaporization, ethanol is next, but
rapid spray deceleration. The opposite effects of the
methanol shows the worst scenario.
55℃ and 90℃ fuel temperatures on penetration indicate
the combined effects of fuel temperature on viscosity
and on spray atomization/evaporation. In the case of
55℃ fuel temperature, the reduced viscosity results in
the increased penetration, while at 90℃, spray
atomization and evaporation effect become dominant
and cause the reduced penetration.
Figure 5. Ambient pressure effect on spray angle
Mie
Figure 7. Effect of fuel temperature on the spray penetration
LIF
Figure 6. Mie and LIF images of ethanol (40KPa, 25℃)
Effect of Fuel Temperature on Spray Characteristics
The fuel temperature has significant impact on spray
characteristics. For all test fuels higher temperature
resulted in a bigger penetration initially (at 0.2ms ASOI),
as shown in Fig. 7. The increased fuel temperature tends
to reduce viscosity, thus the fuel exits the injector
quicker. However, after 0.2ms ASOI, the spray
decelerates quickly with time as compared to 25℃ fuel
temperature
condition.
The
difference
in
spray
penetration for the three fuels was also observed for the
high fuel temperature condition. Spray penetration
Figure 8. Temperature effect on spray penetration for test fuels
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
Spray Characteristics in the Vacuum with Elevated
Fuel Temperature for Cold-start Application
When a fuel with a temperature larger than the boiling
point was injected into a chamber where the ambient
pressure is lower than the saturation pressure, the
flash-boiling happens. Such phenomenon results in a
better spray distribution and a faster evaporation. The
better spray distribution can be due to the low chamber
ambient pressure, decreased fuel density, viscosity and
surface tension. To investigate such a potential way of
obtaining better atomization and evaporation spray
under engine cold-start condition, fuels are tested at the
fuel temperature of 90℃ and the chamber ambient
Figure 9. Spray angle at different fuel temperature
pressure of 40KPa.
The spray angle sensitivity to the fuel temperature was
also investigated. Fig. 9 shows that the fuel temperature
of 55℃ does not affect the spray angle very much.
While at the fuel temperature of 90℃ there is a
8-10°spray angle decrease for gasoline and methanol.
But for ethanol, only slight change of spray angle was
observed.
Mie
LIF
Figure 11. Mie and LIF images of ethanol (40KPa, 90℃)
At the fuel temperature of 90℃ and ambient pressure of
40KPa, a better fuel distribution can be observed in Fig.
11. Much stronger LIF signal than Mie signal is
observed comparing to other test conditions, which
Ethanol
indicates increased vapor concentration. The LIF image
Figure 10. Mie images of the test fuels (100KPa, 90℃)
cone is larger than that of Mie image. The most difficult
Methanol
Gasoline
to evaporate fuel, ethanol, can evaporate very well
Further examination of the Mie images at 90℃ fuel
under this particular condition.
temperature is shown in Fig. 10. The spray structure
becomes more solid cone, compared to the hollow cone
at 25℃, especially for gasoline and methanol. Although
spray penetration difference among the three fuels is
narrowed, the difference in spray distribution becomes
more
obvious.
Gasoline
spray
cone
collapses
completely into a uniform solid cone spray, and sac
spray no longer exists. Ethanol spray still remains some
structure similar to the hollow cone with significant
presence of fuel droplets inside the cone, and the sac
spray still exists as attaching to the main spray. The
methanol spray is between the two and shows a
bell-shape structure.
Figure 12. Axial spray penetration difference between the
conditions of 100KPa, 90℃ and 40KPa, 90℃
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
A greater penetration is obtained compared to test
temperature (90℃) injecting into a vacuum of 40KPa is
condition at ambient pressure of 100KPa and fuel
proven to be able to improve the spray atomization and
temperature of 90℃, as shown in Fig. 12. Ethanol gains
evaporation.
the least penetration increase which may help to
Through the analysis of the experimental data, the
mitigate wall wetting.
proposal of injecting fuels of elevated temperatures into
4. Conclusions
In this study, the sprays of gasoline, methanol and
ethanol were investigated experimentally to understand
the vacuum was expected to improve the engine
cold-start performance.
5. Acknowledgement
the effects of ambient pressure and fuel temperature on
The research was carried out at National Engineering
the sprays structure. And Spray characteristics at
Laboratory
for
Automotive
elevated fuel temperatures injected into the vacuum
Technology,
and
sponsored
were investigated as a potential solution to improve the
Company.
spray characteristics under engine cold-start conditions.
The main conclusions are as follows.
Electronic
by
General
Control
Motors
6. References
1. By comparing the Mie scattering images at various
[1] Heather L, MacLean, Lester B. Lave. Evaluating
ambient pressures and the room temperature, ethanol
automobile
was expected to have the smallest penetration due to its
Progress
highest viscosity. The spray structure of ethanol also has
2003.29:1-69.
a different response to the ambient pressure and fuel
[2] C Stan and R Tröger. Internal Mixture Formation
temperature compared to that of gasoline and methanol.
and Combustion-From Gasoline to Ethanol. SAE Paper
2. The greater spray penetration and wider spray angle
2001, 2001-01-1207.
were obtained when the fuel was injected into a vacuum
[3] Gong Yanfeng, Liu Shenghua, Li Yu. Investigation
of 40KPa due to reduced air resistance. However, less
on Methanol Spray Characteristics. Energy & Fuels
difference between the Mie and LIF images was
2007; 21: 2991-2997.
observed, indicating that the fuel injected into the
[4] Brusstar, M.J., Mark Stuhldreher, M., David Swain,
vacuum is more difficult to evaporate.
D., William M. Pidgeon, W.M., High Efficiency and
3. High fuel temperature results in collapsed sprays
Low Emissions From a Port-Injected Engine With Neat
which have smaller spray penetration and angle, but a
Alcohol Fuels. SAE paper number 2002-01- 2743.
better spray atomization and distribution were observed.
[5] Turner, J. W. G., Pearson, R. J., Holland, B., Peck,
Difference in spray penetration among these three fuels
R., Alcohol-Based Fuels in High Performance Engines.
is small. However, there is still very limited difference
SAE paper number 2007-01-0056.
between LIF and Mie signal, indicating very poor
[6] Takashi Tsunooka, Yohei Hosokawa, Shintaro
evaporation. Moreover a high fuel temperature leads to
Utsumi, Takashi Kawai, and Yukihiro Sonoda, High
the spray deformation for gasoline and methanol. But
Concentration Ethanol Effect on SI Engine Cold
ethanol spray still keeps the main spray shape nearly
Startability, SAE 2007-01-2036
unchanged.
[7] F. Zhao, M.-C. Lai, D.L. Harrington. Automotive
4. Fuels are tested at fuel temperature of 90℃ and the
spark-ignited direct-injection gasoline engines. Progress
chamber ambient pressure of 40KPa as a potential way
in Energy and Combustion Science. 1999.
of obtaining better atomization and evaporation spray,
[8] Anand H. Gandhi, Corey E. Weaver and Eric W.
Both a wider spray distribution and better spray
Curtis, Spray Characterization in a DISI Engine during
uniformity were observed. Moreover, outlines of LIF
Cold
signal for all the test fuels seems much more extensive
2006-01-1004.
compared to that of Mie signal, indicating strong vapor
[9] Hua
phase signals captured. Thus a fuel with a high
Combustion Instrumentation and Diagnositcs, SAE
fuel/propulsion
in
Start:
Energy
(1)
Zhao,
and
Imaging
Nicos
system
technologies.
Combustion
Investigation,
Ladommatos,
Science
SAE
Engine
The 13th Annual Conference on
Liquid Atomization and Spray Systems- Asia
October 15-17, 2009
Wuxi, P.R.China
books, 2001.
Conference
[10] Gashi Sara, Yan YouyouLockett, Russel D,
December 16, 2002 - December 20, 2002.
Arcoumanis
induced
[11] MounaIm-Rousselle Christine and Pajot Olivier,
fluorescence measurements of fuel concentration in a
Droplet sizing by interferometric Mie scattering in
gasoline direct injection optical engine, 1st International
engine environment, SPIE proceedings series, 1997.
Constantine,
Planar
laser
on
Optical
and
Laser
Diagnostics,
Download