Optics & Laser Technology 64 (2014) 343–351

Contents lists available at ScienceDirect

Optics & Laser Technology

journal homepage: www.elsevier.com/locate/optlastec

A comparative study of laser ignition and spark ignition

with gasoline–air mixtures
Cangsu Xu n, Donghua Fang, Qiyuan Luo, Jian Ma, Yang Xie
Department of Energy Engineering, Yuquan Campus, Zhejiang University, Hangzhou 310027, People's Republic of China

art ic l e i nf o a b s t r a c t

Article history: The ignition probability and minimum ignition energy (MIE) of premixed gasoline–air mixture for
Received 16 February 2014 different equivalence ratio was experimentally studied using a nanosecond pulse at 532 nm and
Received in revised form 1064 nm from a Q-switched Nd:YAG laser in a constant-volume combustion chamber (CVCC) The result
3 May 2014
was compared with the spark ignition. The initial pressure and temperature of the mixture was 0.1 MP
Accepted 14 May 2014
Available online 26 June 2014
and 363 K, respectively. The research indicates that within the flammable range, the probability
increases when the ignition energy increases and the distribution of MIE with the equivalence ratios
Keywords: is U-shape for both laser and spark ignition. For laser ignition with 532 nm, when the incident energy is
Laser ignition higher than 110 mJ or the absorbed energy is high than 31 mJ, 100% of ignition could be obtained within
Spark ignition
equivalence ratios of 0.8–1.6. For 1064 nm it is 235 mJ and 30 mJ. To get the same ignition probability of
Minimum ignition energy
mixture with identical equivalence ratio, the incident energy of 1064 nm is twice more than the incident
energy of 532 nm, while the absorbed energy values are virtually the same. It indicates that significant
wavelength dependence is expected for the initial free electrons but irrelevant for the process of
absorbing energy. The initial free electrons are produced from impurities in gasoline–air mixture
because the intensity in the focus (1012 W/cm2) is too low to ionize gas molecules via the multi-photon
ionization process, which requires higher irradiance ( Z1014 W/cm2). The MIE obtained with a laser-
spark ignition is greater than that measured by electrical sparks. The MIE for laser ignition was obtained
at equivalence ratio of 1.0 both of 532 nm and 1064 nm, and it was 13.5 mJ and 9.5 mJ, respectively. But
for spark ignition, the MIE is 3.76 mJ with equivalence ratio of 1.6. What's more, laser ignition extends
the lean flammability limit from 0.8 to 0.6.
& 2014 Elsevier Ltd. All rights reserved.

1. Introduction
With the development of industry and economy, energy short-
age and environmental problems have been becoming more and
more serious. The internal combustion engine and automotive
industry are confronted with severe and realistic challenges. Lean
burn is one of the effective methods to solve the above-mentioned
problems. However, current engines cannot be operated suffi-
ciently lean due to ignition related problems such as the sluggish
flame initiation and propagation along with potential misfiring [1].
It is anticipated that the spark ignition engine of the future will
operate with much higher compression ratios, faster compression
rates, and much leaner fuel-to-air ratios, which will aggravate the
electrode degradation and erosion [2].
In order to improve the ignition stability and reduce the cycle-
by-cycle variations, researchers are exploring new techniques,
including plasma jet igniters [3], laser induced spark ignition [4]
n
Corresponding author. Mobile: þ 86 135 0581 4628.
E-mail address: xucangsu@zju.edu.cn (C. Xu).
and rail-plugs ignition [5]. Most of the problems listed above could
potentially be solved by the use of laser induced spark ignition
because it has many potential benefits. The major benefits are
greater control over the timing and locations of ignition. Moreover,
it is accomplished without electrodes, which allows the lifetime of
a laser-ignition system to be significantly longer. What is more,
laser induced spark ignition would also allow ignition in multiple
locations inside the chamber to shorten the combustion time of
lean mixtures.
In recent years, laser ignition has become an active research
topic because of its many potential benefits over the conventional
electric spark ignition. Laser ignition of reactive mixtures can be
divided into four categories: laser thermal ignition, laser induced
photochemical ignition, laser-induced resonant breakdown igni-
tion and laser induced spark ignition [6]. Laser induced spark
ignition begins with the initial seed electrons produced from
impurities in the gas mixture (e. g dust, aerosol or soot particles).
It is very unlikely that the initial electrons are produced by
multiphoton ionization because the intensities in the focus
(1012 W/cm2) are too low to ionize gas molecules via this process,
which requires intensities of more than 1014 W/cm2 [7,8]. Kopecek

http://dx.doi.org/10.1016/j.optlastec.2014.05.009
0030-3992/& 2014 Elsevier Ltd. All rights reserved.
344 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351
et al. [7] maintained that no wavelength dependence was expected
for this initiation effect while Dhananjay et al. [8] hold the
opposite idea that significant wavelength dependence was
expected for this initiation effect based on Drude model. Initial
electrons readily absorb more photons via the inverse bremsstrah-
lung process to increase their kinetic energy. If the electrons gain
sufficient energy, they can collide with other molecules and ionize
them, leading to an electron avalanche, and breakdown the gas.
This process is repeated until the spark plasma of high tempera-
ture and high pressure is created. This extreme condition relative
to the ambient gas leads to the development of a rapidly expand-
ing shock wave that is of sufficient strength to ignite flammable
mixtures [6].
The laser induced spark ignition has previously been found to
be associated with the laser pulse width, laser energy, the size of
focusing spot, the composition of mixture and its initial conditions.
Studies have mainly focused on mixtures containing hydrogen [9],
methane [7,10] or propane [11] although some studies have also
been performed on hydrocarbon fuels such as dodecane, isooctane
or Jet-A [11–13]. In these studies, the ignition characteristics are
usually expressed according to the energy delivered by laser.
Dhananjay et al. [8] studied characterisation of laser ignition in
hydrogen–air mixtures in a combustion bomb at initial pressure of
3 MPa and temperature 323 K and the results are compared with
the laser ignition ones. They found that the rate of pressure rise
inside the combustion chamber was higher when the mixture was
ignited by laser plasma compared with spark plug ignition.
Weinrotter et al. [9] investigated laser ignition to hydrogen–air
mixtures at high pressures and their results showed that with
increasing initial pressures the minimum pulse energy was
decreasing. Measurements and model calculations of ignition by
electrical sparks and nonresonant laser sparks show that the
minimum ignition energy (MIE) for laser sparks is higher than
for electrical sparks [10,11,14]. Donald [14] attributed it to the
higher energy cost of laser spark formation (both breakdown and
heating) and more efficient removal of the energy absorbed in
laser sparks to regions outside the nominal ignition kernel by the
shock. Ternel et al. [12] found that a too small spark size or spark
Fig. 1. Schematic of laser-induced ignition experimental apparatus. (a) laser beam propagation in the laser-induced ignition experiment, (b) top view of the constant-volume
combustion chamber, (c) photograph of the experimental apparatus for laser-induced ignition.
energy might result in ignition failure if the spark size is not large
enough to support the flame growth or the deposited energy is
less than the minimum ignition energy. Lawes [13] found that the
minimum ignition energy depended on the likelihood of a droplet
existing at the focus of the laser beam.
As far as we know, few papers were devoted to liquid fuel with
laser induced spark ignition and there is controversy about
whether the wavelength plays a role in laser ignition. This paper
focuses on the impact of the equivalence ratio (Φ) and wavelength
(λ) on the ignition energy [7,8]. We use a parametric study to
examine the effects of the equivalence ratio and wavelength on
the ignition energy values of gasoline–air mixture and results are
compared with the spark ignition ones.
2. Experimental setup and experiments
Fig. 1 shows a schematic of the experimental apparatus for
laser ignition, which can be separated into five main parts: ignition
laser, combustion chamber, temperature and pressure measuring
system, heating system and intake and exhaust system.
The optical scheme of the igniting beam is depicted in Fig. 1a, A
Q-switched Nd:YAG (Dawa-100) was used as the ignition source
operating at 532 nm and 1064 nm with pulse duration of 6.6 ns
and 7.9 ns, respectively. The initial diameter of laser beam was
6.0 mm. After being broadened and collimated, the laser beam was
focused on the center of the combustion chamber by a 500-mm
focal length lens. The operating frequency range of laser is 1–20 Hz
and has the single-shot mode.
The energy was controlled by the operating voltage. Two
energy meters were used to measure the beam energy before
and after the chamber as shown in Fig. 1a. The energy meter Ea
(COHERENT J-25-MB-LE) obtained energy before the chamber by
using a beam-splitter that reflected a fraction of the incident
energy (2.4%@532 nm &2.0%@1064 nm) while transmitting the
rest measured by the energy meter Eb(COHERENT J-50-MB-YAG).
Laser energy power meter (EMP2000) could record the incident
energy and transmitting energy simultaneously.
 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351
The combustion chamber equipped with a heating system is
cubical (200 mm  200 mm  200 mm) chamber with a volume of
1.69 L. Before experiment, the chamber was heated up to give an
initial temperature to ensure a homogenous mixture reliably. It
was capable of withstanding high pressures and temperatures up
to 2 MPa and 600 K. The chamber comprised three pairs of
orthogonal windows used to allow the passage of the laser beam
and view the entire volume of the chamber. The windows were
equipped with quartz glass of diameter 120 mm and thickness
23 mm. Two valves were installed on the two corners located on
the top-side of the CVCC for liquid injection and gas inlet and
outlet shown in Fig. 1b. Temperature and pressure sensors were
also installed on the other corners. The pressure transducer
(Kistler, Model 6115B) combined with a charge amplifier (Kistler,
Type 5018A) was used to measure the dynamic pressure during
the combustion.
1-Nd:YAG-Laser (Dawa-100), 2-beam attenuator (wave plate/
polariser), 3-beam-splitter(2.4%@532 nm &2.0%@1064 nm), 4-
concave lens, 5-convex lens, 6-mirror(reflectivity  99.5%), 7-
spherically corrected convex lens f ¼500 mm, 8-constant-volume
combustion chamber (with quartz glass window diameter
120 mm,thickness 23 mm), 9-laser energy sensor Eb(COHERENT
J-50-MB-YAG), 10-laser energy sensor Ea(COHERENT J-25-MB-LE),
11-laser energy power meter (EMP2000) 1-temperature meter, 2-
thermocouple sensor 3-piezoelectric pressure sensor (Kistler,
Model 6115B), 4-charge amplifier (Kistler, Type 5018A), 5-digital
storage oscilloscope (Tektronix DPO2024B).
The investigation of spark ignition characters could be con-
ducted by using the spark plug installed on the metal disc (Fig. 2a)
instead of the quartz glass. Fig. 2b shows a schematic of the
experimental apparatus for spark ignition. The central electrode
was made out of the spark plug. The central electrodes used in the
present investigation were equipped on the opposite sides of the
combustion chamber for spark ignition. The spark was produced
by high voltage in the mixture, which was characterized by a flash
of light and a sharp acoustic sound. The spark released the energy
between the two electrodes. If the voltage varying with time on
both ends of an igniting bridge thread can be measured and
recorded and so can the current, the accurate ignition energy of
spark ignition systems can be obtained, which is given by the
Fig. 2. Spark ignition systems. (a) The metal disc, (b) photograph of the experi-
mental apparatus for spark ignition.
345
following equation (1)
Z t
EIn ¼ uðtÞiðtÞdt ð1Þ
0
where EIn is an accurate value of energy produced by high voltage,
u(t) is the voltage between both ends of the electrode measured by
voltage probe(Tektronix,P6015A), i(t) is the current through the
electrode measured by current probe(Tektronix,TCP2020), and t is
the duration of ignition process. The oscilloscope (Tektronix
DPO2024B) used in this paper has operation and integration
function.
As a general rule, the spark plug has built-in resistance to
reduce wave interference. The energy consumption of the resis-
tance is considered as well, which is given by the following
equation
Z t
EΔ ¼ i2 ðtÞRdt ð2Þ
0
where R is the built-in resistance of 5.1 kΩ.
Therefore, the spark ignition energy could be calculated by the
following equation
ESI ¼ EIn  EΔ ð3Þ
EIn and EΔ could be calculated by the oscilloscope.
Different air–fuel ratio of the gasoline–air mixtures at ambient
temperature 298 K and air humidity 56% were investigated in this
study. Before each experiment, the chamber was initially evacu-
ated by vacuum pump to remove residual gases and moisture. The
initial pressure in the chamber was 0.1 MPa and the chamber was
heated up to an initial temperature 363 K for each combustion
attempt to guarantee a homogeneous mixture. The liquid fuel used
in the experiment is No. 93 gasoline of Sinopec. With certain
volume and temperature, the volume of the fuel was calculated
using the expression:
pV c M r
mair ¼ ð4Þ
T R  103
mair
Vf ¼ ϕ ð5Þ
αρf
Here the mair is the air mass, Vf the volume of the gasoline, Vc
the volume of the chamber (equal to 1.67 L), p the initial pressure
(0.1 MPa), T air temperature (equal to 298 K), Mr the relative
molecular mass of air (equal to 28.8 g/mol), R molar gas constant
(equal to 8.314 J/(mol K)), α theoretical air–fuel ratio of gasoline
(equal to 14.7), ρf the density of gasoline (equal to 0.725 g/mL) and
φ the equivalence ratio.
3. Results and discussion
The incident energy (Ea) has to be discerned from the absorbed
energy (Eabs); the absorbed energy is exactly the energy necessary
to yield ignition inside the combustion chamber, incident energy
determined by working voltage is the total pulse energy needed to
generate ignition. The laser energy was deposited in the plasma
after breakdown which leaded to a shock wave generation. If the
plasma is intense enough, then the resultant ignition kernel is
sufficiently strong to permit transition into full scale combustion.
The probability of ignition was defined as the ratio of effective
trials and the number of shots. The ignition phenomenon involved
the formation of a flame front, a sharply increase of pressure inside
the chamber (Fig. 3) and the combustion of the mixture, as
described in Ref. [15].
346 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351
Fig. 3. Typical time-resolved pressure chart of an ignition event.
3.1. Energy loss
The incident energy and absorbed energy were measured by
the energy meter Ea and Eb, respectively. The main absorbed
energy was deposited in the plasma while the rest was lost with
the development of the plasma. Ma et al. [16] insisted that losses
due to diffraction were negligible and the attenuation could be
considered to be due solely to the absorption by the plasma. In
order to obtain the accurate absorbed energy, we also took into
account the energy loss in the optical devices, which was mea-
sured by placing the energy meter Eb in different locations of the
light path. The experiment results showed that a large part of the
energy loss was in the quartz glass. Fig. 4 shows that the energy
absorbed by the quartz glass depends on the laser wavelength and
increases linearly with the incident energy. The relationship
between the energy loss in the quartz glass and the incident
energy could be deduced according to the experiment results by
the least square method, as shown in Eq. (6) for 532 nm and Eq.
(7) for 1064 nm, respectively:
ΔE532 ¼ 0:053Ea þ 0:216
ΔE1064 ¼ 0:067Ea  0:226
Here ΔE is energy loss and Ea is the incident energy. The root
mean squared error (RMSE) is 0.299 and 0.332, respectively.
3.2. Incident energy and ignition probability
The relationship between incident energy and ignition prob-
ability was discussed in detail. Fig. 5a and b show the probability
of ignition in gasoline–air mixture for 532 nm and 1064 nm. The
probability changes depending on incident energy and mixture
composition. The ignition of gasoline–air mixtures in different
equivalence ratio was investigated. For equivalence ratios between
0.8 and 1.6, the ignition probability could reach 100%. For equiva-
lence ratios equal to 0.6 and 1.8, we failed to obtain a reliable
ignition during our experiments. The maximum probability for
532 nm was 65% and 12% for equivalence ratios of 0.6 and 1.8,
respectively; it was 70% and 35% for 1064 nm. As shown in Fig. 5,
the ignition probability increased with incident energy, which was
consistent with the studies by Bradley [6]. However, Virginie et al.
[17] suggested that the probability of ignition do not increase
necessarily with increasing incident energy. This is due to the
variability of absorbed energy by plasma for the same incident
energies. Only by absorbing enough energy can the mixture be
ignited.
Fig. 4. The energy loss in the quartz glass.
Comparing Fig. 5a and b, we found that getting the same
probability the incident energy for 532 nm differed from that for
1064 nm. For equivalence ratios of 0.8, 1 and 1.2, 100% of ignition
was indeed obtained when the incident energy was greater than
ð6Þ
110, 105 and 108 mJ, respectively, when it was 240, 220 and
ð7Þ 230 mJ for 1064. To get the same ignition probability of mixture
with identical equivalence ratio, the incident energy of 1064 nm is
twice more than the incident energy of 532 nm. This is due to the
difference of pulse width and divergence angle. The divergence
angle is obtained by measuring the spot radius in the focal plane
with CCD; the smaller the spot radius, the smaller the divergence
angle is. The divergence angle was 0.409 and 0.417 mrad for 532
and 1064 nm, respectively. Fig. 6 shows the far field intensity
distribution of 532 and 1064 nm.
If we assume that the beam is cylindrical [10] with a Gaussian
beam profile, its radius r and length l at the focal point are:
  
2λ f
r¼ ð8Þ
π D0
pffiffiffi θ 2
l ¼ ð 2  1Þ f ð9Þ
D0
V ¼ πr 2 l ð10Þ
Here f is the focal length of the focusing lens (mm), D0 the laser
beam diameter before the focusing lens (mm), θ the beam divergence
(mrad) and V the volume of focal spot. In this paper, f¼ 500 mm,
D0 ¼24 mm (the expanded laser beam diameter), θ532 ¼0.409 mrad
and θ1064 ¼0.417 mrad. The calculating results are as following:
r532 ¼ 0.007 mm, l532 ¼1.764 mm, V532 ¼2.76  10  13 m3; r1064 ¼
0.014 mm, l1064 ¼ 1.798 mm, V1064 ¼11.26  10  13 m3. The focal spot
 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351
Fig. 5. Incident energy and probability. (a) 532 nm of incident laser beam wave length, (b) 1064 nm of incident laser beam wave length.
radius of 1064 nm is greater than that of 532 nm. This would mean
that more incident energy is required for ignition of 1064 nm.
Based upon this result, the focal intensity can be immediately
calculated:
ðEa  ΔE1 Þ
If ¼
t pulse πr 2
Here ΔE1 is the absorption of the first window of the chamber,
tpulse is the pulse width(6.6 ns@532 ns and 7.9 ns@1064 nm). For
532 nm, the focal intensity for 100% of ignition is greater than
9.682  1012 W/cm2; for 1064 nm, it is greater than 4.348 
1012 W/cm2. The results confirm that the initial free electrons are
produced from impurities in gas mixture because the intensity in
the focus(1012 W/cm2) is too low to ionize gas molecules via the
multiphoton ionization process, which requires higher irradiance
(Z 1014 W/cm2) [7,8]. The focal intensity changes depending on
wavelength, which means that significant wavelength dependence
is expected for this initiation effect.
3.3. Absorbed energy and ignition probability
The absorbed energy was deposited in the plasma to yield
ignition inside the combustion chamber. A part of incident energy
was absorbed by plasma and optical devices, the rest transmitted
347
through the chamber measured by the energy meter Eb. The
absorbed energy Eabs can be calculated by the expression:
Eabs ¼ Ea  Eb  ΔE ð12Þ
Fig. 7 shows the relationship between the absorbed energy and
ð11Þ incident energy at stoichiometric ratio. For the same incident
energy, the absorbed energy by plasma is not a constant value and
only by absorbing enough energy can the mixture be ignited,
which is consistent with the studies by Virginie [17].
In most of the previous studies of laser-induced ignition, the
probabilities of ignition are expressed according to the incident
energy [9,12]. Given the variability of deposited energy for same
incident energy, it seemed more appropriate to express the
probability of ignition versus energy absorbed by the plasma
during breakdown. Fig. 8a and b shows the ignition probability
of all equivalence ratios for 532 and 1064 nm, respectively. As
mentioned above, getting the same ignition probability of mixture
with identical equivalence ratio, the incident energy of 1064 nm is
twice more than the incident energy of 532 nm, while the
absorbed energy values for 532 and 1064 nm are approximately
equal, particularly for equivalence ratios between 0.8 and 1.2. It
indicated that significant wavelength dependence is expected for
the initial free electrons but irrelevant for the process of absorbing
energy. For equivalence ratios between 0.8 and 1.2, little energy is
needed to ignite the mixture. The absorbed energy for 100% of
348 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351
Fig. 6. The far field intensity distribution. (a) 532 nm of incident laser beam wave
length, (b) 1064 nm of incident laser beam wave length.
ignition is greater than 19 mJ. For equivalence ratios equal to
1.4 and 1.6, 100% of ignition is obtained with absorbed energies
were greater than 25 and 30 mJ, respectively. For equivalence
ratios equal to 0.6 and 1.8, we failed to obtain 100% of ignition
during our experiments. The maximum probability was 59 and
48% for equivalence ratios of 0.6 and 1.8, respectively. What is
more, it indicates that the equivalence ratio also plays a role on the
shape of the curve. For equivalence ratios between 0.8 and 1.2 with
1064 nm, the energy gap between an ignition probability of 0 and
100% is thin. As described in [17] the gap becomes increasingly
large when equivalence ratios deviate from these values. It is
therefore more interesting to use a mixture with an equivalence
ratio to ensure the ignition and the reproducibility of the
phenomenon.
3.4. The minimum ignition energy
The minimum ignition energy is one of the important para-
meters for safety production, design of combustion device and
system reliability. The ignition energy is defined as the amount of
laser energy that is absorbed by the plasma [2]. The minimum
ignition energy in this paper is defined as the minimum absorbed
Fig. 7. Absorbed energy versus incident energy at stoichiometric ratio. (a) 532 nm
of incident laser beam wave length, (b) 1064 nm of incident laser beam wave
length.
energy at 100% of ignition. Fig. 9 presents the results. The
distribution of minimum ignition energy with the equivalence
ratios is U-shape. With leaner mixtures the ignition energy has to
be higher to produce more radicals in the focal region to create a
self-sustaining flame front. With richer mixtures the number of
molecules in the focal region is increased and more energy is
needed to ignite the mixture. The least values of minimum ignition
energy, both of 532 and 1064 nm, were obtained at stoichiometric
ratio and were 13.5 and 9.5 mJ, respectively. When equivalence
ratios deviate from stoichiometric ratio, the minimum ignition
energy becomes increasingly large. The minimum ignition ener-
gies remain at lower values for mixtures of equivalence ratio
between 0.8 and 1.4. Increases in the minimum ignition energies
toward both the lean side and the rich side of the equivalence ratio
are similar to the reports of other researchers [11,15,18,19]. For
equivalence ratios of 0.6 and 1.8, the minimum ignition energies of
532 nm is 36 and 51 mJ, respectively, while it is 27 and 39 mJ for
1064 nm.
3.5. The comparison between spark and laser ignition
The spark ignition energy was obtained from the high voltage
and current between the two electrodes. The voltage and current
varying with time was measured and recorded by the oscilloscope
as shown in Fig. 10, wave 1 for voltage and 2 for the current.
Typical value of peak voltage and current in our experiment is
about 6 kV and 200 mA. And the gap between cursor a and b is the
duration of ignition process.
Fig. 11 shows the probability of ignition in gasoline–air mixture
for spark ignition. The probability changes depending on spark
energy and mixture composition. The ignition of gasoline–air
mixtures in different equivalence ratio was investigated. For
 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351
Fig.8. Absorbed energy and probability. (a) 532 nm of incident laser beam wave length, (b) 1064 nm of incident laser beam wave length.
Fig. 9. The minimum ignition energy.
equivalence ratios between 0.8 and 2.0, the ignition probability
could reach 100%. The maximum probability for equivalence ratios
2.2 was 36.7%. For equivalence ratios equal to 0.6, we failed to
obtain a successful ignition during our experiments. The ignition
probability increases with spark energy, which is similar to laser
ignition.
349
As shown in Fig. 12, the distribution of minimum spark ignition
energy with the equivalence ratios is also U-shape. The least
values of minimum spark ignition energy was obtained at the
equivalence ratio of 1.6 and it was 3.76 mJ. These results are in
good agreement with recent studies [14,20], which indicates that
the minimum ignition energies obtained with a laser-spark igni-
tion are greater than those measured by electrical sparks. The
reasons were attributed to the short time duration of the laser
pulse and the small focal spot size that tends to rapidly dissipate
the thermal energy due to high thermal gradient induced by the
laser spark in comparison to electric sparks [11]. Another major
difference is that the free electrons can be readily accelerated from
the metal surface of the cathode for electrical sparks, whereas the
free electrons in the gas in laser sparks must be somehow initially
produced by using an internal energy-absorption process [21].
Fig. 13 shows the flammability limit for laser and spark ignition.
It is not hard to find that the maximum probability for equivalence
ratio of 0.6 was 65% for 532 nm and 70% for 1064 nm by laser
ignition nevertheless, we failed to obtain a successful ignition with
spark ignition, no matter how much energy was deposited in the
spark. What is more, the least values of minimum ignition energy
was obtained at the equivalence ratio of 1 for laser ignition and it
was obtained at 1.6 for spark ignition. It can be concluded that
laser ignition extends the lean flammability limit, while the spark
ignition could be more favorable for rich mixture.
350 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351

Fig. 10. The voltage and current varying with time.

4. Conclusions

The incident and absorbed energy of gasoline–air mixtures

with laser-induced spark ignition were experimentally investi-
gated. The results can be summarized as follows:

(1) A large part of the energy loss was in the quartz glass, which
depends on the laser wavelength and increases linearly with
the incident energy.
(2) Within the flammable range, the probability of ignition
increases when the incident energy increases. For equivalence
ratios of 0.8–1.6, 100% of ignition is obtained. The distribution
of minimum ignition energy with the equivalence ratios is U-
shape. The least values of minimum ignition energy, both of
Fig. 11. The probability of ignition for spark ignition. 532 and 1064 nm, were obtained at stoichiometric ratio and
were 13.5 and 9.5 mJ, respectively. And it was 3.76 mJ at the
equivalence ratio of 1.6 for spark ignition. Compared with
spark ignition, laser ignition extends the lean flammability
limit.
(3) For 532 and 1064 nm, the focal intensity for 100% of ignition is
greater than 9.682  1012 and 4.348  1012 W/cm2, respec-
tively. It confirms that the initial free electrons are produced
from impurities in gas mixture.
(4) For the same incident energy, the absorbed energy by plasma
is not a constant value and only by absorbing enough energy
can the mixture be ignited. Getting the same ignition prob-
ability of mixture with identical equivalence ratio, the incident
energy of 1064 nm is twice more than the incident energy of
532 nm, while the absorbed energy values for 532 and
1064 nm are approximately equal. It indicates that significant
wavelength dependence is expected for the initial free elec-
Fig. 12. The minimum ignition energy for spark and laser ignition. trons but irrelevant for the process of absorbing energy.

Acknowledgements

This work is supported by the State Key Development Program

for Basic Research of China (Grant no.2013CB228100), National
Natural Science Foundation of China (nos. 50976100 and
51076138) and the Fundamental Research Funds for the Central
Universities (no. 2013QNA4017).

References

[1] Mohamed H. Morsy. Review and recent developments of laser ignition for
internal combustion engines applications. Renewable Sustainable Energy Rev
Fig. 13. The flammability limit for laser and spark ignition. 2012;6:4849–75.
 C. Xu et al. / Optics & Laser Technology 64 (2014) 343–351 351

[2] Phuoc TX. Laser-induced spark ignition fundamental and applications. Opt Donghua Fang(1987.09-): Current Master Students of
Lasers Eng 2006;44:351–97. Zhejiang University. Research on vehicle engineering
[3] Chintala N, Bao A, Lou G, et al. Measurements of combustion efficiency in and laser ignition.
nonequilibrium RF plasma ignited flows. Combust Flame 2006;144(4):744–56.
[4] Steven R, Michael HM, Steven D. Woodruff Dustin Mcintyre. Misfire, knock and
NOx mapping of a laser spark ignited single cylinder lean burn natural gas
engine[C]. SAE paper 2004-01-1853.
[5] Gao H, Matthews RD, Hall MJ, et al. From spark plugs to railplugs-the
characteristics of a new ignition system[C]. SAE paper 2004-01-297.
[6] Bradley D, Sheppard CG, Suardjaja IM, et al. Fundamentals of high-energy
spark ignition with lasers. Combust Flame 2004;138:55–77.
[7] Kopecek H, Maier H, Reider G, et al. Laser ignition of methane–air mixtures at
high pressures. Exp Thermal Fluid Sci 2003;27:499–503.
[8] Dhanajay KS, Martin W, Kurt I, et al. Characterisation of laser ignition in
hydrogen–air mixtures in a combustion bomb. Int J Hydrogen Energy
2009;34:2475–82.
Qiyuan Luo (1990.12-): Current Master Students of
[9] Weinrotter M, Kopecek H, Tesch M, et al. Laser ignition to hydrogen–air
Zhejiang University. Research on vehicle engineering
mixtures at high pressure. Int J Hydrogen Energy 2005:319–26.
and automotive electronics.
[10] Beduneau JL, Kim B, Zimmer L, Ikeda Y. Measurements of minimum ignition
energy in premixed laminar methane/air flow by using laser induced spark.
Combust Flame 2003;132:653–65.
[11] Lee TW, Vievk J, Stuart K. Measurements of minimum ignition energy by using
laser sparks for hydrocarbon fuels in air: propane, dodecane, and jet—a fuel.
Combust Flame 2001:1320–8.
[12] Ternel C, Lecordier B, Cessou, et al. Laser-induced spark ignition of Propane-air
and isooctane-air. In: Proceedings of the European Combustion Meeting,
Belgium; 2005.
[13] Lawes M, Lee Y, Mokhtar AS, Woolley R. Laser ignition of iso-octane air
aerosols. Combust Sci Technol 2008;180:296–313.
[14] McNeill DH. Minimum ignition energy for laser spark ignition. Proc Combust
Inst 2005;30:2913–20.
[15] Phouc TX, White FP. Laser-induced spark ignition of CH4/air mixtures. Jian Ma (1990.12-): Current Master Students of Zhejiang
Combust Flame 1999;119:203–16. University. Research on vehicle engineering and com-
[16] Ma JX, Alexander DR, Poulai DE. Laser spark ignition and combustion bustion characteristic of biomass fuel.
characteristics of methane–air mixtures. Combust Flame 1998;112:492–506.
[17] Virginie T, Philippe G, Denis B. Ignition study of acetone air mixtures by using
laser-induced spark. J Hazard Mater 2012:372–8.
[18] Syage JA, Fournier EW, Rianda R, et al. Dynamics of flame propagation using
laser-induced spark ignition: ignition energy measurements. J Appl Phys
1988;64:1499–507.
[19] Lim EH, McElroy A, Ronney R, Syage JA. In: Chan SH, editor. Transport
phenomena in combustion. London: Taylor & Francis; 1996.
[20] Turcu ICE, Gower MC, Huntington P. Measurement of KrF laser breakdown
threshold in gases. Opt Commun 1997;134:66–8.
[21] Meeks JM, Craggs JD, editors. Wiley; 1978.

Xie Yang (1990.01-): Current Master Students of Zhe-

jiang University. Research on vehicle engineering and
Cangsu xu(1963.09-): Ph.D. Zhejiang University; Dep-
spray atomization.
uty Director of Power Machinery and Vehicular Engi-
neering Institute of Department of Energy Engineering
in Zhejiang University. Research interests: laser igni-
tion; spray atomization; biomass fuel.