-
gnb
vin
f Tec
Flame kernel development
Laser beam prole
Plasma formation
ced
ed
ion
ce
for
ar
as
el d
different relative airfuel ratios (l1.21.7) and the images were
interrogated for temporal propaga-tion of ame front. Pressure-time
history inside the combustion chamber was recorded and
analyzed.
This data is useful in characterizing the laser ignition of
natural gasair mixture and can be used in
developing an appropriate laser ignition system for commercial
use in SI engines.
withdegradblems haveimprothe n(CNG)
engine operation, higher pressure of combustible charge can
be
mita-lean
lasertched
exceeds certain threshold intensity level, breakdown of
medium
Contents lists available at ScienceDirect
lse
Optics and Lasers
Optics and Lasers in Engineering ] (]]]]) ]]]]]]one is the spark
creation due to the local deposition of energy.E-mail address:
[email protected] (A.K. Agarwal).used. This further increases
in-cylinder pressure towards the end occurs leading to the
formation of a plasma spark, whose sizedepends on the numerical
aperture (NA) of the focused laser beam.If the energy content of
the spark is high enough, the mixtureignites. Laser ignition can be
divided into two main parts. The rst
0143-8166/$ - see front matter & 2011 Elsevier Ltd. All
rights reserved.
doi:10.1016/j.optlaseng.2011.04.015
n Corresponding author. Tel.: 91 512 259 7982; fax: 91 512 259
7408.Pleasvoluthis leads to considerable reduction in power density
of theengine. To compensate for lower power density due to
leaner
laser are focused by a lens system inside the chamber
containingcombustible fuel-air mixture. If the peak intensity in
the focal regionengines can be potentially operated at relatively
higher compres-sion ratios, thus leading to higher thermal
efciencies. The mostimportant pollutant of concern from CNG fueled
engine is oxidesof nitrogen (NOx). Emissions from a CNG fueled
engine can befurther improved by igniting leaner fuelair mixtures,
however
system is a desirable option, which will overcome these litions
to achieve higher engine efciency by igniting an ultramixture in
reciprocating engines.
Laser is an alternative ignition source for engines. Shortpulses
of few nanoseconds pulse duration delivered by a Q-swifor some of
these problems. CNG is regarded as one of the mostpromising
alternative fuels and is probably the cleanest commer-cial fuel.
This fact has resulted in an increased interest in usingCNG as a
fuel for transport engines all over the world. CNG hashigher octane
number compared to gasoline, thus CNG fueled
these electrodes. An increase in chamber pressure with the
sameelectrode distance means an increase in the required
secondarycoil voltage applied to the spark plug. Therefore in order
to realizecleaner combustion of leaner CNGair mixture at higher
chamberlling pressures, a durable high-energy electrode-less
ignition1. Introduction
The world is presently confrontedfuel depletion and
environmentalconcern towards environmental prosions, stringent
emission regulationthe world. Thus, an alternative andwill be
helpful in coping up withregulations. Compressed natural gase cite
this article as: Srivastava DK,me combustion chamber. Opt
Laser& 2011 Elsevier Ltd. All rights reserved.
the twin crises of fossilation. With increasingdue to vehicular
emis-been imposed all overved engine technologyew requirements
andis one of the solutions
of compression stroke, i.e. at the time of combustion.
Howeverleaner airfuel mixtures combined with higher pressures at
thetime of ignition require relatively much higher voltages,
whenconventional spark plug technology is used. Providing the
neces-sary spark energy at these relatively higher voltages to
operatethese engines signicantly reduces the lifetime of spark
plugs [1].The amount of the energy released at the spark
electrodesdepends mainly on the pressure of the combustion
chambertowards the end of compression stroke and the distances
betweenFlame kernel characterization of laser iin a constant volume
combustion cham
Dhananjay Kumar Srivastava, Kewal Dharamshi, A
Engine Research Laboratory, Department of Mechanical
Engineering, Indian Institute o
a r t i c l e i n f o
Article history:
Received 13 January 2011
Received in revised form
27 April 2011
Accepted 27 April 2011
Keywords:
Laser ignition
Constant volume combustion chamber
a b s t r a c t
In this paper, laser-indu
Experiments were perform
compression stroke condit
conditions except turbulen
locations, which are used
were conducted at 10 b
combustion phenomena w
CMOS camera. Flame kern
journal homepage: www.eet al. Flame kernel characteEng (2011),
doi:10.1016/j.oition of natural gasair mixtureer
ash Kumar Agarwal n
hnology Kanpur, Kanpur 208016, India
ignition was investigated for compressed natural gasair
mixtures.
in a constant volume combustion chamber, which simulate end of
the
s of a SI engine. This chamber simulates the engine combustion
chamber
of airfuel mixture. It has four optical windows at diametrically
opposite
laser ignition and optical diagnostics simultaneously. All
experiments
chamber pressure and 373 K chamber temperature. Initial stage
of
visualized by employing Shadowgraphy technique using a high
speed
evelopment of the combustible fuelair mixture was investigated
under
vier.com/locate/optlaseng
in Engineeringrization of laser ignition of natural gasair
mixture in a constantptlaseng.2011.04.015
-
kernel growth in methaneair mixtures at 10 bar and different
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]]2This can be achieved in any gas. Breakdown is
associated withplasma formation and shock wave generation. The
second part ofthe laser ignition is the ignition itself based on a
positive balancebetween the deposited energy and the losses. In
this case, a amekernel can develop.
There are four mechanisms by which laser radiations interactwith
medium/combustible fuelair mixtures: thermal ignition[24],
photochemical ignition [57], resonant ignition [89] andnon-resonant
breakdown [10]. Non-resonant breakdown of gas ismore favorable
because it does not require a close match betweenthe laser
wavelength and the target molecules [11]. This processgenerally
begins with multi-photon ionization of a few gasmolecules, which
release electrons that can then readily absorbmore photons,
increasing their kinetic energy. The electronsliberated by this
means collide with other molecules and ionizethem, leading to an
electron avalanche and breakdown of the gas.It is important to note
that this process requires initial seedelectrons. These electrons
are produced from impurities presentin the combustible gas mixtures
[12], which absorb the laserradiations and lead to very high local
temperatures and as aconsequence, free electrons start the
avalanche process. Multi-photon processes are usually essential for
the initial stages ofbreakdown because the available laser photon
energy similar towhat is employed in this work is approximately 1
eV, whereas theionization potential for most molecules is more than
10 eV [13].Initial ame growth from laser ignition resembles in some
waysthe process of electric spark ignition; however, the initial
stages ofenergy deposition differ considerably. In laser ignition,
most ofthe energy transfers to plasma within the pulse duration of
laser[14] is of the order of nano-seconds whereas in electric
sparkignition, energy transfer lasts in microseconds range.
There are several potential benets of laser ignition over
theconventional spark plug. Detailed advantages of laser
ignitionwere reviewed by Paul [15]. The choice of location of
plasmainside the combustion chamber is one of the several
importantadvantages of laser ignition. Location of ignition
initiator sparkcould be placed at any optimum location inside the
combustionchamber using a suitable focal length of lens, which is
notpractically feasible in any conventional spark plug engine.
Thisway, ame propagation distance could be reduced and combus-tion
duration could also be decreased. This may also potentiallyhelp in
ignition of relatively leaner airfuel mixtures, where theslower
combustion is the main issue. Since the laser ignition doesnot
employ any spark electrode, there is no erosion effect asobserved
in case of spark plug engines therefore the life span oflaser
ignited engine system is expected to be signicantly longerthan that
of spark plug [16]. A diode-pumped laser ignitionsystem has
potential lifetime up to 10,000 h compared to sparkplug lifetimes
of the order of 20004000 h. McMillian et al. [17],McIntyre et al.
[18], Tauer et al. [19] have developed a miniaturelaser that can be
mounted directly on the cylinder head. Multi-point ignition in the
combustion chamber is also possible withlaser ignition. Phuoc [20]
found that multi-point ignition ofcombustible gas mixture increases
the combustion chamberpressure and shortens the combustion
duration. This furtherenhances the possibility of using laser
ignition system for ignitinglean airfuel mixture and enhances the
combustion speed.
In the present investigation, early stages of
laser-inducedignition of a CNGair mixture at 10 bar lling pressure
wereexperimentally investigated for potential application in
anengines simulated environment. To improve the understandingof
in-cylinder combustion, it is important to understand the
amepropagation. The burning speed or rate of ame propagation is
afundamental parameter, which inuences the engine perfor-mance and
emissions. In the early stages of ignition, the relative
importance of the shape and development of ame kernel is
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.oair/fuel equivalence ratios. The data points of this
study exhibiteda signicant spread because only one picture was
taken duringeach test and temporal growth of ame kernel was
composed of asequence of ignition tests. However, it was concluded
that amegrew faster at stoichiometric air/fuel ratio than the lean
air/fuelmixtures. Tewari and Wilson [24] investigated the effect of
highfrequency electric eld on the ame propagation generated
bylaser-induced spark. They conducted experiments at one
atmo-spheric pressure in mixtures of methaneair, methaneoxygenargon
and hydrogenair. They found that the ame propagationrate increases
to almost double in the methaneair mixture inpresence of high
frequency electric eld however the identicaleld condition has an
insignicant effect on the ame propaga-tion rate in the hydrogenair
mixture. Beduneau and Ikeda [25]investigated the laser-induced
spark kernel in a premixed laminarmethaneair burner. The ame kernel
is in asymmetric toroidalshape, which is caused by the expansion
mode of the shock wave.The asymmetric behavior was attributed in
part to the plasmacharacteristics. In the initial stages of ame
kernel growth, ameexpansion velocity strongly correlated to the
spark energy. In thelater stage of expansion, the velocity was
found to depend mainlyon the relative airfuel ratios. There are
several others researchers[2629] who performed optical
investigations to visualize theame evolution. The objective of this
study is to investigate thelaser ignition behavior of CNGair
mixture in a constant volumecombustion chamber. CNG is regarded as
one the most promisingalternative fuel for the engines; thus
performing laser ignition inCNGair mixture would be one step closer
towards the develop-ment of laser red natural gas engine. To
understand thecombustion, ame kernel shape and propagation in the
earlystages of combustion was investigated for different relative
airfuel ratios. In addition to providing time resolved images of
amekernel growth and its speed, pressure-time history inside
thecombustion chamber was also investigated.
2. Experimental setup
2.1. Experimental apparatus
The schematic of experimental setup is depicted in Fig. 1.
Laserignition of CNGair mixture was performed in a
speciallydesigned constant volume combustion chamber to gain
thefundamental information like minimum laser ignition
energyrequired for plasma generation, ame kernel development
andgrowth, ame speed and pressure-time history inside thecombustion
chamber, whose conditions typically represent the endof the
compression stroke conditions of a typical internal combus-tion
engine. The internal diameter and length of combustionchamber is 72
and 220 mm, respectively. Constant volume com-particularly high. At
this time, the expansion speeds for theplasma growth are low. In
the earlier work, Srivastava et al. [21]measured the size and
propagation behavior of laser plasma inatmospheric air. It was
found that the plasma propagates towardsthe incoming laser. This
backward moving plasma (towards thefocusing lens) grows much faster
than the forward movingplasma (along the direction of laser). Phuoc
and White [22]reported the laser spark size in methaneair mixture
at 1 atm. Itwas observed that laser plasma elongated in the
direction of laserbeam. The shape of laser spark was oval and lean
for richmethaneair mixture whereas it becomes cylindrical in
shapefor stoichiometric and near stoichiometric methaneair
mixture.The spark length and radius were about 0.8 and 0.3 mm,
respec-tively. Lackner et al. [23] also investigated the
laser-induced amebustion chamber was able to simulate the real
engine combustion
rization of laser ignition of natural gasair mixture in a
constantptlaseng.2011.04.015
-
ratios and this was kept xed for all the experiments in
thisinvestigation.
For all the experiments, moisture-free compressed air
andcommercially available CNG were used for performing combus-tible
fuelair mixtures. Commercially available CNG contains95.6% methane,
1.2% ethane, 1.4% carbon dioxide and 1.7%nitrogen [30]. Properties
of the CNG are given in Table 1.
For achieving the required relative airfuel ratio (l) ofthe
gaseous mixtures (1.21.7 in this study), it was necessaryto measure
the partial pressure of air and CNG using a high-resolution digital
manometer. Gas lling arrangement is shown inFig. 2. Cylinders lled
with air and CNG are tted with pressureregulators. Outlet pressure
of the cylinder is kept slightly higherthan what is required for a
particular airfuel ratio. Gases owfrom the cylinder via the high
pressure regulator through the highpressure pipe to the combustion
chamber. To achieve therequired airfuel ratio inside the combustion
chamber, partialpressure of gas was measured by digital manometer.
Requiredairfuel ratio was controlled according to Daltons law of
partialpressures. CNG was lled rst because required partial
pressureof CNG was low for all intended airfuel ratios. Then, air
was lledwith high partial pressures. High turbulence generated by
airlling inside the combustion chamber helps in the formation of
ahomogeneous mixture. Additionally, the mixture of air and fuelare
left for 1 min to doubly ensure the thermal stabilization
andhomogeneity of the combustible mixture. After combustion
using
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]] 3chamber conditions except turbulence. It was
provided with fouroptical windows at diametrically opposite
locations so that theycould be used for laser ignition and optical
diagnostics/Shadow-graphy simultaneously. Combustion chamber was
designed to beable to withstand 300 bars static pressure and could
be heated upto 300 1C.
A Q-switched Nd: YAG laser (NanoL 200-30, Litron UK) wasused for
the ignition of CNGair mixture, which is capable ofdelivering
maximum pulse energy up to 200 mJ and a pulseduration of 69 ns at
full width half maximum (FWHM) atfundamental wavelength. The beam
diameter was 5 mm (1/e2).An aperture of 2.5 mm was inserted between
the safety shutterand an output coupler of the laser to enhance the
beam qualityand improve the M2 value. This resulted in maximum
pulseenergy being limited to 38 mJ/ pulse. The beam prole and
M2
values were measured using a laser beam prolometer (Win-CamD,
DataRay Inc. USA). The laser energy could be attenuatedcontinuously
using an external wave plate/ polarizer setup with-out affecting
any laser parameters such as pulse duration orspatial beam prole.
The energy of each pulse was measuredusing a pyro-electric detector
and laser energy meter (FieldMax,Coherent UK). A piezoelectric
pressure sensor (6052C, KistlerSwitzerland) was installed in the
chamber for measuring pres-sure-time history inside the combustion
chamber. A high speedcamera (SA 1.1, Photron UK) was used to
visualize the amekernel growth. Minimum frame rate for this camera
is 5400 fps atmaximum resolution and 6,75,000 fps at minimum
resolution. Awhite light source (OSL1-EC, Thorlabs USA) was used to
illumi-nate the ame kernel.
2.2. Experimental procedure
Fig. 1. Schematic diagram of laser ignition shadowgraphy of
CNGair mixture.Before conducting experiments, it was necessary to
character-ize the laser beam. Laser beam prole and beam quality
wasmeasured by a laser beam proler. Laser beam prole and
beamquality are important parameters in laser ignition. These
affectthe minimum energy required for formation of plasma
andinitiation of combustion. Laser beam was expanded from 2.5 to15
mm using a set of lenses. A 100 mm focal length of plano-convex
lens was used to focus the laser beam. The location offocal point
was kept inside the combustion chamber in such away that the plasma
is formed in front of the orthogonal windowsso that laser plasma
and ame kernel growth can be visualized asshown in Fig. 1. It was
necessary to measure minimum laser pulseenergy required for the
formation of plasma in atmosphericconditions as well as at 10 bar
pressure. Plasma formation prob-ability was calculated at different
pulse energies and chamberpressures. Laser pulse energy for
combustion experiments waschosen such that the laser ignition was
successful for all airfuel
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.olaser ignition, exhaust valve was opened and the
contents of thecombustion chamber are evacuated by a vacuum pump so
that noresidual gases are left out in the CVCC. It was also ushed
outusing fresh air to remove the traces of exhaust. The CVCC
thusbecomes ready for next combustible mixture preparation for
thenext combustion event.
The pressure-time history inside the combustion chamber
wasrecorded using a piezoelectric pressure transducer. The
signalsfrom the pressure transducer were amplied using a
chargeamplier and were recorded in a digital storage
oscilloscope.Then pressure-time history signals were sent to
computer forfurther data analysis.
Table 1Properties of CNG.
Sl. no. Properties Values
1. Relative density 0.64
2. Auto ignition temperature (1C) 5403. Flammability range (%
v/v) 515
4. Octane no. 127
Fig. 2. Gas lling arrangement for the constant volume combustion
chamber withprecise control of relative airfuel ratio. 1High
pressure pipe, 2Pressureregulating valve, 3Digital manometers and
4Vacuum Pump.
rization of laser ignition of natural gasair mixture in a
constantptlaseng.2011.04.015
-
The early development of ame kernel was observed byShadowgraphy
technique [31]. Collimated white light beam wasdirected into the
combustion chamber and ame kernel wascaptured on the other side
with the help of the high speed cameraoperating at 54,000 fps. The
camera was also used to trigger thelaser system. Computer, camera,
laser and oscilloscope are syn-chronized for combustion and
visualization of ame kernel.Camera is triggered from the computer
software. A 5 V outputsignal is generated by camera. This signal is
divided into two sothat laser and oscilloscope are triggered
simultaneously. Laser istriggered in the external mode so that as
soon as laser receivessignal from the camera it will re the laser
pulse. If the energydeposited in the plasma is above the critical
value for combustion,then combustion event takes place and is
recorded by the camera.After receiving the trigger signal from
camera, oscilloscope isactivated in the measurement mode. During
combustion, varia-tion of pressure inside the combustion chamber is
recorded bypiezoelectric transducer and stored in the computer
connectedto oscilloscope.
investigated rst. Shape of plasma is shown qualitatively inFig.
4a. Laser enters from right to left. A prior study by Srivastavaet
al. [21] measured the shape and propagation behavior ofplasma
generated in atmospheric air. The maximum diameterand length of the
laser plasma 30 ns after the ring of the laserwere 0.01 and 0.27
mm, respectively, as depicted in Fig. 4b. It wasobserved that
plasma growth took place towards the incominglaser beam much faster
compared to along the laser beam. Thereason for this reverse
propagation of plasma is that the layers ofgas outside the plasma,
although transparent to the laser beam,get heated by the plasma
radiation. This outside gas close to theplasma will in turn get
ionized to such an extent that it willstrongly absorb the laser
beam [32]. As a result, these gas layerswill then get further
heated rapidly and their temperaturesincrease. By this time,
another layer of plasma near to the laserwould become strongly
absorbing, and hence the boundary of theplasma will move toward the
focusing lens.
Once the plasma is generated in the atmospheric air, itbecomes
pertinent to study the probability of plasma formationupon ring a
laser pulse of certain energy. Fig. 5 shows the plasmaformation
probability in atmospheric air and 10 bar of air pres-sure (in the
chamber lled with only air) with different laserpulse energies.
Plasma formation probability is dened as numberof successful
breakdown events divided by number of attempts.For each pulse
energies investigated, laser was red 500 times.The breakdown
threshold was dened as the laser energy atwhich the air would
breakdown for more than 50% of theattempts.
It can be noticed very clearly from Fig. 5 that the air
break-down pulse energy thresholds at atmospheric condition and
Laser Pulse Energy (mJ/Pulse)Fig. 5. Plasma formation
probability in air at atmospheric pressure and 10 barpressure.
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]]4Fig. 3. Laser beam prole at 1 m distance from laser
head (a) 2-D prole, (b) 3-Dprole, (c) Line prole of beam along
X-axis and (d) Line prole of beam along3. Results and
discussion
Laser beam prole is a critical parameter for laser
ignitionexperiments. It is usually necessary to measure it in order
toensure the adequate beam quality prole. Before starting
theexperiments in combustion chamber, ofine test was carried outto
characterize the laser beam. Laser beam prolometer was usedto
measure the beam quality with a cavity aperture of 2.5 mm at1 m
distance from the laser head. Fig. 3a and b show 2-D and 3-Dbeam
proles, respectively. Fig. 3c and d show the line prolealong the X
and Y axis, respectively.
It can be seen from Fig. 3 that the laser beam prole
deviatesslightly from TEM00 mode. Beam quality factor (M
2 value) wasmeasured for the cavity aperture of 2.5 mm. M2 denes
thefocussability of laser beam. It is directly related to the
diameterof beam at focal point. M2 value of perfectly Gaussian
laser beamis 1.0. In the present case, M2 value of laser beam was
observed tobe 4.6. For laser ignition, the M2 value closer to the
Gaussian laserbeam is desired however it also leads to severe
reduction inmaximum laser pulse energy therefore a balance between
the twois required to be found out.
For systematic investigation of laser ignition of fuelair
mix-tures, plasma generated in atmospheric condition needs to
beY-axis.
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.oFig. 4. (a) Plasma formation in air (b) emission
photography of the laser plasma30 ns after the laser pulse
[21].
0102030405060708090
100
0 3 6 9 12 15 18 21
Plas
ma
Form
atio
n Pr
obab
ility
(%) 1 bar
10 bar10 bar pressure were 17 and 7 mJ, respectively. In
summary, the
rization of laser ignition of natural gasair mixture in a
constantptlaseng.2011.04.015
-
breakdown threshold energy decreases with the increasing
cham-ber pressure. The reason is that at higher pressure
conditions, thenumber of gas molecules in the focal region
increases and laserenergy can be absorbed more efciently. In an
engine operatingwith lean airfuel mixture, in-cylinder pressure at
the time ofignition should increase in order to compensate for the
resultingpower density loss. From Fig. 5, it could be seen that
this willcreate a favorable condition in case of laser ignition,
where laserpulse energy required to breakdown decreases with
increasingchamber pressure. This trend is completely opposite to
the oneobserved in conventional spark plug system. It is reported
thatthe spark energy required for igniting the mixture increases
withincreasing in-cylinder pressures [1]. Since at 9 mJ laser
pulseenergy, the plasma formation probability is 100% at 10
barchamber pressure, a notch higher level of pulse energy i.e.12 mJ
pulse energy is chosen for carrying out the combustionexperiments
in the present study in order to be able to cover thewhole spectrum
of experimental conditions.
Once the minimum laser pulse energy for this set of experi-ments
was determined, ignition experiments and ame kernelvisualization of
different airfuel ratios were carried out in theconstant volume
combustion chamber. The images showing thedevelopment of the early
ame kernel stages and its growth withtime were recorded by
employing Shadowgraphy technique. Highspeed camera was used to
capture the image at 54,000 fps. Theconsecutive images were
captured at an interval of 18.5 ms for anygiven single combustion
event. These images provided useful
At the early stages of ame development (to92.5 ms, notshown in
these pictures), a toroidal shape of the kernel wasobserved.
Toroidal shape of ame kernel shape is similar to theone observed in
conventional spark electrode ignition system.Maly and Vogel [33],
conducted experiments in methaneairmixture using conventional spark
plug. They changed the spark
T = 1.87ms T =2.70ms T = 3.53ms T = 4.37ms
Fig. 8. Shadowgraph image of ame kernel development (l1.4).
T = 37s T= 370s T = 703s T = 1036s
T = 2.37ms T = 3.70ms T = 5.03ms T = 6.36ms
Fig. 9. Shadowgraph image of ame kernel development (l1.5).
T = 2.70ms T = 4.37ms T = 6.03ms T = 7.70ms
T = 37s T = 370s T = 703s T = 1036s
Fig. 10. Shadowgraph image of ame kernel development (l1.6).
T = 6.03ms T = 11.03ms T = 16.02ms T = 21.02ms
T = 37s T= 370s T = 703s T = 1036s
Fig. 11. Shadowgraph image of ame kernel development (l1.7).
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]] 5information about the ame kernel development, and
the shapeof the ame kernel as a function of time for different
airfuelratios (l1.21.7). Figs. 611 show the images of ame
kerneldevelopment process at different time scales. Time scale of
rstfour images was kept constant in all gures however inFigs. 911,
the time scale for the last four images was observedto be different
because of longer combustion duration for theleaner mixtures. In
all these gures, laser beam is entering fromright to the left.
Vertical and horizontal dimensions of all imageswere kept constant
at 1.15 and 1.45 cm, respectively.
T = 37s T= 370s T = 703s T = 1036s
T = 1.87ms T = 2.70ms T = 3.53ms T = 4.37ms
Fig. 6. Shadowgraph image of ame kernel development (l1.2).
T = 37s T= 370s T = 703s T = 1036s
T = 1.87ms T = 2.70ms T = 3.53ms T = 4.37msFig. 7. Shadowgraph
image of ame kernel development (l1.3).
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.oT = 37s T= 370s T = 703s T = 1036senergy to see its
effect on ame kernel size and expansion
rization of laser ignition of natural gasair mixture in a
constantptlaseng.2011.04.015
-
velocity. A smooth ame front appeared for high-energy
arc.Initially, the shape of kernel was cylindrical for very short
timethen spherical changing to toroidal for intermediate time
andspherical again for longer time durations. Other
researchers[3436] also visualized the ame kernel formation in
combustionchamber for different airfuel mixture using conventional
sparkplug. The evaluation behavior of ame kernel was found
bespherical. In the present experiment with laser ignition, the
toroidshape continues to develop radially away from the spark
centertill 92.5 ms. After approximately t92.5 ms, a front lobe is
formedand propagates towards the incoming laser beam. This is
apeculiar feature of laser-induced ignition. Similar
expansionbehavior of laser ignited ame kernel is observed by
otherresearchers as well [14,23,37]. The shape of the ame kernel
isobserved to be structurally identical for all airfuel ratios. It
isobserved from Figs. 10 and 11 that the front lobe of the
amekernel disappeared after approximately 1 ms for relative
airfuelratios of 1.6 and 1.7 (image 5 onwards). Based on these
observa-
for every airfuel ratio, the images were analyzed and
averagevalues are presented in these graphs to reduce effect of
experi-mental errors. Centroid of the rst kernel is taken as origin
forcalculating distance in the three directions.
It can be observed from Fig. 12 that during the early
stages(t666 ms), ame propagation distance increases rapidly,
sug-gesting higher ame velocity initially. For leaner mixtures
(l1.6and 1.7), ame propagation distance in X-direction decreases
withtime in the later stages of combustion (t1.16 ms) because
theleaner mixture is unable to sustain the combustion in front
lobebeyond a limit, resulting in reduction in ame velocity.
However,for relatively richer mixtures (l1.2, 1.3 and 1.4), ame
propaga-tion distance increases up to 2.8 ms and then it becomes
almostconstant. The maximum distance of ame propagation observedfor
lambda 1.2 is 0.53 cm.
Flame kernel propagation in the direction of laser beam, i.e.X
direction is shown in Fig. 13. It can be observed from thisgure
that the propagation of ame kernels were varying almostlinearly for
relative airfuel ratio (l1.2, 1.3 and 1.4) with timeand this
suggest almost constant ame velocity. The maximumpropagation
distance for lambda 1.2 is 0.3 cm. It can be concludedfrom Figs. 12
and 13 that forward propagation distance of amekernel i.e. in the
direction of laser beam, is less than thepropagation distance of
ame kernel in backward direction, i.e.opposite to laser beam
direction, for all the relative airfuel ratio.
Propagation of ame kernel in Y and Y direction wasalmost
identical and therefore only one direction is reported.Fig. 14
shows the temporal variation of ame kernel in Ydirection. It can be
seen from the gure that the maximum ame
Fig. 14. Temporal variation of ame kernel development in the
orthogonal to thelaser beam propagation (Y-direction).
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]]6tions, shape of the laser-induced ame kernel can be
thought tobe having two stages of development. In early stage
(to92.5 ms)of ame kernel development, the kernel develops radially
to forma toroidal shape. In the latter stage (t492.5 ms) of
development, afront lobe is formed, which propagates backwards
towards theincoming laser beam. This expansion phenomenon is
alsoreported earlier [14] however the reasons for this have not
beenexplained fully and convincingly. Spiglanin et al. [37]
proposedpossible explanation for the formation of front lobe and
suggestedthat this phenomenon may be related to the initial ow
eldcreated by the propagation of a radiation transport wave up
tothe laser beam, arising from the high rate of energy transferred
atthe leading edge of the plasma. An additional factor may be
thepreheating of the gases by the focused laser beam that ignites
themixture. This preheating gas readily ignites in a ame front
thatpropagates much faster than it would through cold
combustiblemixture.
After understanding the ame kernel development, it is logicalto
analyze the temporal variation of the ame kernel develop-ment in
various directions with time inside the combustionchamber for
varying relative airfuel ratios. This informationcan be attained
from the analysis of different photographicimages captured by the
high speed camera. Temporal develop-ment of ame kernel was analyzed
using MATLAB. Lasers direc-tion of propagation is taken as X and
direction opposite to thelaser propagation is taken as X , which is
also the direction ofpropagation of front lobe. The temporal
variation of ame kerneldevelopment in X , X and Y direction are
given in Figs. 1214.Five combustion events are carried out under
identical conditions
Fig. 12. Temporal variation of ame kernel development in the
direction of
opposite to laser beam propagation (X-direction).
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.oFig. 13. Temporal variation of ame kernel
development in the direction of laserbeam propagation (X
direction).kernel propagation distance was 0.38 cm for relative
fuelair
rization of laser ignition of natural gasair mixture in a
constantptlaseng.2011.04.015
-
earlier observations, which indicate that the rich
fuelairmixtures give higher ame velocities and faster propagation
ofame kernel. It can also be noticed that the peak cylinder
pressuredecreases with leaner mixtures, as expected. The
experimentswere also carried out with very lean mixture (l1.7)
however thepressure-time history showed unacceptable variations
reectinguncertain combustion behavior. Further leaner mixtures
couldnot be ignited by lasers and this suggested that for a
practicalapplication of laser ignition system applied to the engine
will notbe able to deal with CNGair mixtures leaner than l1.6.
Thisstatement is however subject to the given type of optics.
Ifimproved optics and better quality laser beam is used,
possiblylean combustion limit in an engine can be pushed
further.
To study the event-to-event variation in combustion
pressure,peak pressure variations were analyzed for different
airfuelratios. One of the important advantages of laser ignition is
the
dt
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]] 7Fig. 15. Flame propagation speed in the
X-direction.
05
101520253035404550
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Pre
ssur
e (b
ar)
Time (ms)
Lambda = 1.2Lambda = 1.3
Lambda = 1.4 Lambda = 1.5Lambda = 1.6ratios (l1.2 and 1.3). The
ame kernel propagation distance wasfound to consistently decrease
with the leaner mixtures and it isfound to be lowest for l1.7.
These ndings are almost similar tothe ones for X directions
suggesting that the volumetric growth ofthe ame kernel reduces
substantially with reduction of airfuelmixture strength.
Flame speed was derived from propagation distance. Fig. 15shows
the propagation speed of ame in direction opposite tolaser beam
direction, i.e. X-direction for different relative airfuelratio. It
is observed from gure that ame propagation speeddecreases with
increasing relative airfuel ratio and alsodecreases with time.
Initially propagation speed for higher rela-tive airfuel ratios,
i.e. leaner mixture strength is observed to behigher than that of
relatively richer mixtures however thispropagation speed decreases
very fast. At 0.33 ms, ame propaga-tion speed for lambda 1.2 and
1.6 are 5.54 and 4.07 m/s,respectively, while at 6.67 ms, it
becomes 1.62 and 3.21 m/s,respectively. Propagation speed becomes
negative for lambda 1.5,1.6 and 1.7 at some time because of
reduction of ame propaga-tion distance and suggests shrinking of
ame kernel, particularlyfor leaner mixture (l1.6 and 1.7).
It is also important to experimentally evaluate the
pressure-time history in the combustion chamber for varying
fuelairmixture as this will provide vital information about the
kind ofpressure rise, which can be expected in an engine system
ignitedusing laser.
Fig. 16 shows a pressure-time history of the combustionchamber
for different relative airfuel (l1.21.6) at initialchamber lling
pressure and temperature of 10 bar and 373 K,respectively, ignited
by laser. It can be clearly seen from thisgure that there is a
clear trend towards longer combustionduration with leaner CNGair
mixtures. This is also supported by
Fig. 16. Pressuretime history of the combustion chamber for
different l.
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.oI. The cylinder charge was considered to behave as
an ideal gas.II. Distribution of thermodynamic properties inside
the combus-
tion chamber was considered to be uniform.III. Dissociation of
combustion products was neglected.IV. Heat transfer from the
combustion wall is neglected in
this model.
Since experiments were done in constant volume
combustionchamber, the rate of volume change parameter in heat
releaseequation will be zero. Rate of heat release for constant
volumecombustion chamber will therefore be:
dQ
dt 1g1
VdP
dt
0
5
10
15
20
25
30
35
40
45
50
1.2 1.3 1.4 1.5 1.6
Pea
k P
ress
ure
(bar
)
Relative Air-Fuel Ratio (-)TFi
rizaptlahe following assumptions were made in this
calculation.g1 V dt g1 P dtdQ 1
dP g
dVRexact regulation of deposited energy in the focal volume. So
it isexpected that the variation in combustion is lower in case of
laserignition. Fig. 17 shows the variation of peak combustion
pressureinside the combustion chamber for different airfuel ratios.
Errorbars in Fig. 17 show that there is very little variation in
peakchamber pressure. There is a good repeatability of
experimentsand the observed data.
Rate of heat release (ROHR) was calculated from the
acquiredpressure-time history data of the CVCC using zero
dimensionalheat release analysis model [38].
ate of heat release was calculated asg. 17. Peak pressure
variations for laser ignition of different airfuel ratios.
tion of laser ignition of natural gasair mixture in a
constantseng.2011.04.015
-
for providing funding for carrying out this project. The
Engine
ignition in 02/03 mixtures. Applied Physics B 1985;37:18995.[3]
Hill RA. Ignition-delay times in laser initiated combustion.
Applied Optics
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]]84. Summary
In this study, laser-induced ignition of CNGair mixtures
wasexperimentally investigated. Experiments were conducted in
aconstant volume combustion chamber at 10 bar initial llingpressure
and 373 K temperature. A Q-switched Nd: YAG laserwas used for the
ignition of CNGair mixture at the fundamentalwavelength (1064 nm).
The beam quality was measured using anoptical prolometer. Plasma
was generated in the atmosphericcondition as well as 10 bar chamber
pressure, and the minimumenergy required for plasma generation with
50% probability werefound to be 17 and 7 mJ, respectively. It was
observed thatbreakdown threshold energy decreases with increase in
chamberpressure. This trend is completely opposite to the one
observed inconventional spark plug system. Since at 9 mJ laser
pulse energy,the plasma formation probability is 100% at 10 bar
chamberand the net heat release is
Q Z t0
dQ
dt
dt
where Q is net heat release, t is the time, V is the volume
ofcombustion chamber, P is combustion pressure and g is the ratioof
specic heat.
Fig. 18 shows the net heat release for different relative
airfuelratio. Net heat release decreases with increase in relative
airfuelmixture. Signicant differences in slope of heat release
indicatesthe variations in combustion duration. Combustion
durationincreases with increasing relative airfuel i.e. for leaner
mixtures.This observation is also supported by the decrease in ame
speedwith increasing relative airfuel. Heat release for lambda 1.2
is12.4 kJ whereas for leaner mixture, (lambda1.6) it is 8.8 kJ.
Fig. 18. Net heat release versus time for different relative
airfuel ratios.pressure, 12 mJ pulse energy is chosen for carrying
out thecombustion experiments in the present study in order to be
ableto cover the whole spectrum of experimental conditions.
Ignition experiments and ame kernel visualization of differ-ent
airfuel ratios were carried out in the constant volumecombustion
chamber. The images showing the development ofthe early ame kernel
stages and its growth with time wererecorded by employing
Shadowgraphy technique. At the earlystages of ame development
(to92.5 ms), a toroidal shape of thekernel was observed. The toroid
continues to develop radiallyaway from the spark center till 92.5
ms. After t92.5 ms, a frontlobe is formed and propagates towards
the incoming laser beam.This is a peculiar feature of laser-induced
ignition. The shape ofthe ame kernel is observed to be structurally
identical for allairfuel ratios of the combustible airfuel
mixtures. Based onthese observations, shape of the laser-induced
ame kernel can bethought to be having two stages of development. In
early stage
Please cite this article as: Srivastava DK, et al. Flame kernel
charactevolume combustion chamber. Opt Laser Eng (2011),
doi:10.1016/j.o1981;20(13):223956.[4] Hill RA, Laguna GA. Laser
initiated combustion of CH4O2 mixtures. Optics
Communications 1980;32(3):4359.[5] Lavid M, Stevens JG.
Photochemical ignition of premixed hydrogen/oxidizer
mixtures with excimer lasers. Combustion and Flame
1985;60:195202.[6] Chou Mau-Song, Zukowski TJ. Ignition of
H2/02/NH3, H2/Air/NH3, and
CH4/O2/NH3 mixtures by excimer-laser photolysis of NH3.
Combustion andFlame 1991;87:191202.
[7] Lucas D, Dunn-Rankin D, Hom K, Brown NJ. Ignition by excimer
laserphotolysis of ozone. Combustion and Flame 1987;69:17184.
[8] Forch BE, Miziolek AW. Laser-based ignition of H2/02 and
D2/02 premixedgases through resonant multiphoton excitation of H
and D atoms near243 nm. Combustion and Flame 1991;85:25462.
[9] Forch BE, Miziolek AW. Oxygen-atom two-photon resonance
effects inmultiphoton photochemical ignition of premixed H2/02 ows.
Optics Letters1986;11(3):12931.
[10] Morgan CG. Laser-induced breakdown of gases. Reports on
Progress inPhysics 1975;38:62165.
[11] Kopecek H, Maier H, Reider G, Winter F, Wintner E. Laser
ignition ofmethaneair mixtures at high pressures. Experimental
Thermal and FluidScience 2003;27:499503.
[12] Yablonovich E. Self phase modulation and short pulse
generation from laserbreakdown plasmas. Physical Review A
1975;10:188895.
[13] Radziemski LJ, Cremers DA. Laser-induced plasmas and
applications. New York,Basel: Marcel Dekker; 1989.
[14] Bradley D, Sheppard CGW, Suardjaja IM, Woolley R.
Fundamental of high-energy ignition with lasers. Combustion and
Flame 2004;138:5577.
[15] Paul DR. Laser versus conventional ignition of ames.
Optical Engineering1994;33(2):51021.
[16] Weinrotter M, Kopecek H, Wintner E, Lackner M, Winter F.
Application oflaser ignition to hydrogenair mixtures at high
pressures. InternationalResearch Laboratory staff members Mr.
Roshan Lal and Mr. RaviSingh are also acknowledged for their help
during the experimentalsetup development and execution of
experiments.
References
[1] Wienrotter M, Srivastava DK, Iskra K, Graf J, Kopecek H,
Klausner J, et al. Laserignition of engines: a realistic option!
Proceedings of SPIE 2006;6053:605316.
[2] Raffel B, Warnatz J, Wolfrum J. Experimental study of
laser-induced thermal(to92.5 ms) of ame kernel development, the
kernel developsradially to form toroidal shape. In the latter stage
(t492.5 ms) ofdevelopment, a front lobe is formed, which propagates
backwardstowards the incoming laser beam. Front lobe of ame
kerneldisappeared after approximately 1 ms for leaner mixtures
(l1.6and 1.7). The temporal variation of the ame kernel
developmentin various directions with time inside the combustion
chamberfor different airfuel ratios is analyzed by interrogation
ofdifferent images. The ame kernel propagation distance wasfound to
be consistently decreasing with leaner mixtures and itfound to be
the lowest for (l1.7). These ndings are almostsimilar to the ones
for X directions suggesting that the volumetricgrowth of the ame
kernel reduces substantially with reductionof airfuel mixture
strength. The pressure-time history of thecombustion chamber for
different airfuel mixture was investi-gated. There was a clear
trend towards longer combustionduration with leaner CNGair
mixtures. This is also supportedby earlier observations, which
indicate that the richer fuelairmixtures give higher ame velocities
and faster propagation ofame kernel. It can also be noticed that
the peak cylinder pressuredecreases for leaner mixtures, as
expected. A practical applicationof laser ignition system applied
to the engine will not be able todeal with CNGair mixtures leaner
than (l1.6).
Acknowledgments
SERC (Mechanical and Civil Engineering), Department of
Scienceand Technology, Government of India, is gratefully
acknowledgedJournal of Hydrogen Energy 2005;30:31926.
rization of laser ignition of natural gasair mixture in a
constantptlaseng.2011.04.015
-
[17] McMillian MH, Woodruff SD, Richardson SW, McIntyre DL.
Laser spark ignition:laser development and engine testing,
ICEF2004-917, Fall technical conferenceof the ASME ICE division,
Long Beach, California, USA, October 2427, 2004.
[18] McIntyre DL, Woodruff SD, Ontko JS. Lean burn stationary
natural gas engineoperation with a prototype laser spark plug.
Journal of Engineering for GasTurbines and Power
2010;132(072804):16.
[19] Tauer J, Koer H, Wintner E. Laser initiated ignition. Laser
and PhotonicReviews 2009:124.
[20] Phuoc TX. Single-point versus multi-point laser ignition:
experimentalmeasurements of combustion times and pressures.
Combustion and Flame2000;122:50810.
[21] Srivastava DK, Weinrotter M, Iskra K, Agarwal AK, Wintner
E. Characteriza-tion of laser ignition in hydrogenair mixtures in a
combustion bomb.International Journal of Hydrogen Energy
2009;34(5):247582.
[22] Phuoc TX, White FP. Laser-induced spark ignition of CH4/air
mixtures.Combustion and Flame 1999;119:20316.
[23] Lackner M, Charareh S, Winter F, Iskra K, Rudisser D, Neger
T, et al. Investigationof the early stages in laser-induced
ignition by Schlieren photography and laser-induced uorescence
spectroscopy. Optics Express 2004;12(19):454657.
[24] Tewari GP, Wilson JR. An experimental study of the effects
of high frequencyelectric eld on laser-induced ame propagation.
Combustion and Flame1975;24:15967.
[25] Beduneau JL, Ikeda Y. Application of laser ignition on
laminar ame frontinvestigation. Experiments in Fluids
2004;36:10813.
[26] Morsy MH, Chung SH. Laser-induced multi-point ignition with
a single-shotlaser using two conical cavities for hydrogen/air
mixture. ExperimentalThermal and Fluid Science 2003;27:49197.
[27] Lee JH, Lee SI, Kwon OC. Effect of ammonia substitution on
hydrogen/airame propagation and emissions. International journal of
hydrogen Energy2010;35:1133241.
[28] Mohammadi A, Shioji M, Matusi Y, Kajiwara R. Spark-ignition
and combustioncharacteristics of high-pressure hydrogen and
natural-gas intermittent jets.Journal of Engineering for Gas
Turbine and Power 2008;130:062801-17.
[29] Parsinejad F, Keck jC, Metghalchi H. On the location of ame
edge inshadowgraph pictures of spherical ames: a theoretical and
experimentalstudy. Experimental Fluids 2007;43:88794.
[30] Mishra DP. Experimental studies of ame stability limits of
CNGair pre-mixed ame. Energy Conversion and Management
2007;48:120811.
[31] Settles GS. Schlieren and Shadowgraph techniques:
visualizing phenomena intransparent media. Berlin Heidelberg New
York: Springer-Verlag; 2001.
[32] Lamoureux N, Djebaili-Chaumeix N, Paillard CE. Laminar ame
velocitydetermination for H2airHeCO2 mixtures using the spherical
bombmethod. Experimental Thermal and Fluid Science
2003;27:38593.
[33] Maly R, Vogel M. Initiation and propagation of ame fronts
in lean CH4-airmixtures by the three modes of the ignition spark.
Symposium (Interna-tional) on Combustion 1979;17:82131.
[34] Rozenchan G, Zhu DL, Law CK, Tse SD. Outward propagation,
burningvelocities and chemical effects of methane ames up to 60
atm. Proceedingof the Combustion Institute 2002;29:14619.
[35] Xiong Y, Roberts WL, Drake MC, fansler TD. Investigation of
pre-mixed amekernel/vortes interactions via high-speed imaging.
Combustion and Flame2001;126:182744.
[36] Lee K, Ryu J. An experimental study of the ame propagation
and combustioncharacteristics of LPG fuel. Fuel
2005;84:1116127.
[37] Spiglanin TA, Mcilroy A, Fournier RB, Cohen RB. Time
resolved imaging ofame kernels: laser spark ignition of H2/O2/Ar
mixtures. Combustion andFlame 1995;102:31028.
[38] Hi-techniques revelation operator reference manual.
Madison, USA: Hi-Techniques, Inc.; 2004.
D.K. Srivastava et al. / Optics and Lasers in Engineering ]
(]]]]) ]]]]]] 9Please cite this article as: Srivastava DK, et al.
Flame kernel charactevolume combustion chamber. Opt Laser Eng
(2011), doi:10.1016/j.orization of laser ignition of natural gasair
mixture in a constantptlaseng.2011.04.015
Flame kernel characterization of laser ignition of natural
gas-air mixture in a constant volume combustion
chamberIntroductionExperimental setupExperimental
apparatusExperimental procedure
Results and discussionSummaryAcknowledgmentsReferences