Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>



        Combustion in a spark ignition engine begins with a spark initiated in the air-fuel mixture.
After a delay, a flame forms which propagates across the combustion chamber burning the fuel-air
mixture as it goes. The pressure and temperature increase up to a certain maximum and then
decrease. The flame reaches the other end of the combustion chamber and gets quenched signifying
the end of the combustion process.

        An enormous amount of theoretical and experimental research has been carried out on the
subject of combustion of homogenous, premixed, fuel-air mixtures. Basic studies have been carried
out in constant volume bombs, shock tubes, and other apparatus that simulate the actual working
conditions. Studies have been carried out in piston-cylinder assemblies, in rapid compression
machines that simulate the combustion processes only, and in actual engines.

         In those engines where the process of combustion follows the scenario presented above, that
is, a process involving the propagation of a flame and the burning of the mixture as it goes along,
the combustion process is termed as normal. Abnormal combustion is said to occur when the
process of flame propagation is not normal, and there is a premature burning of the charge, and
other phenomena take place. Some examples of abnormal combustion are detonation or knock, pre-
ignition, and after- or post-ignition.

         Ignition, by implication, is merely a prerequisite of combustion so the study of combustion
must begin with the consideration of the phenomenon of combustion to establish a criterion on
which to decide whether, in a particular case, ignition has occurred or not. In terms of its simplest
definition, ignition has no degree, intensively or extensively. Either combustion of a medium is
initiated or it is not initiated. It is, therefore, sensible to consider ignition from the standpoint of the
beginning of the combustion process that it initiates.

        Normal combustion can be regarded, in effect, as a zone of burning, propagated through a
medium by means of heat transfer and diffusion, in a manner that constitutes a wave in the broad
sense of the term. The zone of burning is the reaction zone and the propagation of the reaction or
burning is the combustion wave. The flame can also be stationary; that is, it remains in one position
with respect to the datum while the combustible mixture (of the fuel and air) flows into it, like in a
gas burner of a boiler or gas turbine. As seen earlier, however, the general concepts of a reaction
zone and burning velocity can be retained.

        In the presence of a heat sink, like a solid surface, the dynamics of combustion are
modified. From a phenomenological point of view, the reaction zone will not approach the heat
sink closer than a certain distance, which produces a dead space. If two heat sinks, each with its
dead space, are moved towards one another, the dead spaces will eventually combine to form a
zone of quenching, into which a combustion wave cannot propagate. For example, in a flame trap
of the wire mesh type, the gap between the wires is less than the quenching distance, which is a
critical dimension depending upon the nature of the combustible mixture, among other factors.

        The ignition process though virtually connected with the initiation of combustion; in
general, it is not associated with the gross behavior of the combustion waves. Instead, it is a local,
small-scale manifestation that takes place inside a combustion wave. Therefore, in the search for
theoretical support for the phenomena of ignition and combustion, it is more rewarding to look at
processes occurring inside the wave. One such treatment of the problem requires, a hypothesis,
concepts of excess enthalpy and minimal flame; among other things, it demonstrates the existence
of the quench distance. It is assumed that a spherical shell of burning gas enclosing and emanating
from a sphere of burnt gas can approximately represent a spherical combustion wave of minimum
diameter. The propagation of this or any wave is simply the transport of energy across the wave
front in the direction of propagation. Two contributions to the problem exist, namely,

(a)    the diffusion of the chemical reactants and the products of the reaction and
(b)    the conduction of heat from the burned to the adjacent unburned gas.

       Furthermore, if it is now assumed, with some justification, that the diffusion transport can
be neglected, in comparison with the conduction of heat, it can be shown, from combustion wave
theory that the burned gas inside the wave has enthalpy, that is, energy in a particular state, in
excess of the ambient level given by the following formula

                                         h = (k/Vb)(Tb - Tu)                                     .. (1)

where h = enthalpy per unit area of the wave,
     k = coefficient of thermal conductivity,
    Vb= burning velocity
     Tb= temperature of the burned gas
     Tu= temperature of the unburned gas

       For a spherical flame of surface area d2, where d is the diameter of the sphere, the total
excess enthalpy is given by

                                             Htheo = d2h                                        .. (2)

        Now, if the hypothetical flame has a diameter less than the minimum, the volume of the
enclosed sphere of burned gas will be too small to provide the required amount of excess enthalpy
to allow the propagation to continue. Thus, the reaction will cease and the flame will become
extinct. This minimum diameter is clearly related to, and must be less than, the quenching distance.

Energy Basic Requirements for Ignition

        Self-ignition is the limiting case in which the systems have been brought to the state - by
expenditure of energy - such that no further internal energy, intended solely for the ignition process,
is required to initiate combustion. The distinction between energy supplied to a system that is not
self-igniting, to bring about ignition, and the work done on the system to make it self-igniting, is
subtle and essentially practical since it is convenient to distinguish between the two methods of
ignition in practical applications. There is no fundamental difference between the two processes.

        An ignition process obeys the law of conservation of energy. It is treated as the balance of
energy between that provided by an external source, that is released by chemical reaction, and that
dissipated to the surroundings by thermal conduction, convection, and radiation and mass transfer
which, in turn, are affected by experimental conditions. It has already been stated that Htheo given
in Eq. 2 gives the total excess enthalpy required to cause the flame to be self-sustaining and
promote ignition.

         It is, therefore, reasonable to suppose that the basic requirement of the source of ignition is
that it supplies this energy within a small volume compatible with the dimensions of the minimal
flame in a time short enough to ensure that only a negligible amount of energy is lost over and
above that required to establish a flame. Thus, the rate of supply of energy, or power, is as
important as the total energy supplied.

Spark Ignition

       A small electric spark of short duration is the ideal requirement for ignition. However, the
energy content need not be high. The actual energy required for the spark to cause ignition of a
combustible mixture is not yet truly concepted.

        A spark is caused by applying a sufficiently high voltage between two electrodes separated
by explosive gas in the gap. It is possible to pass small electric sparks through the gas (in the gap)
without producing ignition. When the spark energy is increased, that is, when the voltage across the
electrodes is raised above a certain critical value (below which a spark may not even occur) a
threshold energy is eventually obtained at which the spark becomes incendiary in the sense that a
combustion wave propagates from the spark through the volume of gas. This minimum ignition
energy is a function of experimental variables such as the parameters of the explosive gas and the
configuration of the spark gap.

        Paschen's Law states that the critical voltage (at which spark would occur) is a function of
the product of the dimensions of the gap and the gas pressure. Also, the manner in which voltage is
raised to the critical value, configuration and condition of the electrodes and the nature of the
combustible mixture are all-important in relation to the energy required.

Capacitance Spark

        In Fig. 1, which shows a part of the circuit used to study spark ignition effects (given by
Blaue et al1), when the condenser is charged to a voltage V equal to the critical or break down
voltage of the gap, the condenser will discharge as a spark across the gap. In the absence of any
resistance or inductance in the circuit a total energy of ½CV2 in the condenser will be dissipated at
the gap and, except for losses, will appear as energy in the simple, brief spark available for ignition.
One important loss is due to quenching at the electrodes.
    3rd Symposium. On Combustion, p 363.
        From a series of experiments in which the composition, pressure, and temperature of the
explosive gas were held constant and the length of the spark gap was systematically changed,
curves of the maximum ignition energy versus distance between the electrodes were obtained. Two
typical curves obtained are shown in Fig. 2.

        One curve corresponds to a series of experiments in which the electrode terminals were
tipped with stainless steel spheres of 1.5 mm diameter. In the other series, the electrodes were
similarly tipped and in addition were flanged by glass plates.

        The curves for the glass plate electrodes take a rather sharp vertical turn around 0.08-inch
(2-mm). For the other type of electrode too, the curve moves up at this point but more gradually.
The energy requirement increases and ignition becomes impossible eventually. This is due to the
quenching effect. The critical distance, wherein ignition is suppressed completely, is the quenching
distance. Experiments have shown that its value is substantially independent of the mode of
ignition. Thus, the same, or nearly the same, value of the quenching distance is obtained from
experiments in which the explosive gas is enclosed in a rectangular channel bounded by two
parallel plates and ignited at one end by a pilot flame. The material of the wall has not been found
to affect the quenching distance; evidently glass metal are equally effective as heat sinks since
thermal conductivities of solids exceed those of gases by orders of magnitude. In the present
experiments the use of flanges made from an electric non-conductor such as glass was indicated
because in this way the sparks remained centered between the electrodes. Metal plates were also
used occasionally but the sparks had a tendency to cross at random anywhere between the plates
and frequently crossed from one plate edge to the other, thus causing ignition even though the plates
were within the quenching distance. When the metal plates behaved, the results were not different
from those obtained with glass-flanged electrodes. Since a glass surface is usually conductive,
irregular discharges and particularly corona discharges were also observed with glass flanges.
Coating the glass with a trace of paraffin wax effectively eliminated this source of error. In contrast
to the sharp vertical turn of the curve corresponding to the glass-flanged electrodes, the other curve
in Fig. 2 rises gradually as the electrode distance is decreased below the quenching distance. The
quenching effect is not so marked and it can be compensated by an increased supply of energy so
that ignition can be obtained even with very small electrode distance. It is noteworthy, however,
that the beginning of the rising part coincides with the quenching distance of the glass-flanged
electrodes; that is, the quenching effect, although much weaker, extends over the same gap length.

        Similar experiments performed on 8.5% methane-air mixtures with three different electrode
configurations, namely, both electrodes flanged by glass plates, negative electrode flanged by glass
plate and positive electrode tipped with 1.5 mm sphere, and vice versa, indicate that at electrode
distances larger than the quenching distance, the size and shape of the electrodes do not
significantly affect the value of the minimum ignition energy, whereas within the quenching
distance the influence of these factors is pronounced; that is, the actual minimum ignition energy
for gaps smaller than the quenching distance may vary appreciably for different electrodes. See Fig.

       The minimum ignition energy and the quenching distance increase with decreasing
pressure. Beyond the quenching distance, the minimum ignition energy is seen to be virtually
independent of electrode distance over a considerable range of distances and pressures. Outside this
range the minimum ignition energy is seen to increase with increasing electrode distance. From Fig.
4 it is seen that when pressure was decreased from 1 to 0.2 atm, the minimum energy required rose
by a factor of 10.

        Figure 5 shows the minimum ignition energy versus distance between plate electrodes when
extended to very large spark energies. It is seen that the quenching distance does not diminish when
large sparks are used, but on the contrary, it increases. The fact that the electrodes are to be placed
apart for reducing energy requirement above the quenching distance is that large sparks produced
turbulence, which caused an increase in heat loss to the plates and thus outweighed the additional
energy. The plates must be separated further in order to obtain ignition. Also, minimum energy
requirements may not be substantially constant at higher spark gaps as shown above.

        When the minimum ignition energy for air and various hydrocarbons is plotted it is seen
that the minima of the energy curves for the various compounds occur at nearly identical energy
values. The minima shifts to richer than stoichiometric mixtures as the number of carbon atoms in
the fuel increases. For methane, the minima occurs at an equivalence ratio of 0.9 whereas for
heptane it occurs at 1.8 and hexene and benzene at 1.75.

        It is observed that the value of Htheo agrees well with actual enthalpy values for high
temperature, fast burning mixtures of fuels with oxygen or oxygen-enriched air whereas the low
temperature, slow-burning fuel-air mixtures show considerably lower experimental values of
enthalpy. This may be due to the fact that the model considers the transport of thermal energy in the
flame and does not allow for transport of chemical energy by the inter-diffusion of reactants and
reaction products; this effect appears to be more significant in slow burning mixtures than in fast
burning ones.

        For high temperature, fast burning mixtures, the Htheo varies as d3 whereas for slow burning
mixtures the slope decreases to somewhat lower than d2. The curves of minimum ignition energy
versus quenching distance, when both quantities are taken on the log scale are independent of the
type of fuel and proportion and pressure of the mixture. See Fig. 6.

       Minimum spark energy and quenching distance decreases with increase in initial
temperature (or temperature of the unburned mixture).

Effect of Electrode Configuration

        The minimum energy continues to decrease with decrease in electrode gap if the electrode is
smaller. Large electrodes produce greater quenching at or below the quenching distance and this
required higher minimum energy.

Effect of Electrode Material

        The material of the electrode has an effect, the effect being that ignition energy for electrode
gaps larger than the quenching distance was found to vary with different materials and increased
after any change to a material with higher melting point. Thus, energy required increased each time
the material was changed from cadmium to aluminum, gold and platinum. It is believed that not all
the energy released at the gap participates in the ignition process. An amount that varies with the
material is lost to the electrodes, possibly causing slight evaporation.

Effect of Series Resistance

         By imposing a resistance in series with the spark circuit, some energy will be dissipated
there during discharge, and total energy required, that is, circuit energy stored in the capacitor, is
likely to be greater. But the energy required at the gap, calculated from the gap current and voltage,
decreased appreciably with increase in the resistance of the circuit. Rose and Priede2 were able to
reduce the minimum ignition energies of hydrogen-air mixtures by this method. By this, the
discharge period is increased and this reduces shock wave formation. Gap geometry and material
have also been studied.

Effect of Stray Inductance

        An inductance L will cause an inductive energy given by


where i is the current and will affect the discharge current. If we put

                                               i = dQ/dt

where Q is the charge, then

                                    L d2Q/dt2 + R dQ/dt + Q/C = 0

where R is the resistance and C is the capacitance.

                                 If R2 < 4L/C discharge is oscillatory

                               If R2  4L/C discharge is non-oscillatory

        For very small values of R, a small, stray inductance will cause an oscillatory discharge.
Introducing a series resistance will ensure a non-oscillatory discharge when inductance is
deliberately held to a minimum. As seen above, the larger the resistance the longer the discharge
time. However, discharge continues only while the spark gap remains conductive; a situation that is
maintained while the voltage is above the breakdown potential of the gap but depends on the
residual ionization below this value. A simple capacitance spark is bright but of short duration; it
occurs mainly but not entirely, in the gas. The spectrum of the material of electrode is also present
to some extent.

    7th Symposium on Combustion, p 436-445 and 454.
Induction Spark

        If a current i flowing through an inductance L (Fig. 7) is interrupted, the collapse of the
current will cause, in the circuit, an induced voltage given by

                                                  L di/dt

where di/dt is the rate of collapse of current.

        Arcing will occur at the interrupter contacts. The energy given by

                                                  ½ Li2

originally stored in the magnetic field, will be divided between the arc-heat at the contacts and other
losses in the system. Thus, the interrupter contacts can be used as spark electrodes and the system
could ignite the surrounding combustible mixture if the minimum ignition energy requirements are
met. This, however, is not a sure method of ignition. An obvious method of improvement is by
using an induction coil in place of a simple coil, a circuit breaker in the primary circuit and a spark
gap in the high voltage secondary circuit. This usually introduces a capacitance in the circuit and the
discharge no longer consists of an induction spark alone.

        An induction spark is less bright than a capacitance spark but of a longer duration.

Induction Coil Spark

        The induction coil spark has a discharge that is generally described as having two
components, namely, capacitance and inductance. In the circuit, as shown in Fig. 8, the discharge
occurs thus:

         On completion/breaking of the primary circuit of the transformer, an electromagnetic force
(emf) is generated in the secondary circuit proportional to the rate of change of current in the
primary. This charges the condenser in parallel with the spark gap or the self-capacitance of the
spark circuit. If and when the breakdown potential of the spark gap is reached, discharge occurs,
initially in the form of a brief capacitance spark of energy, which is given by

                                             ½ CV2 - losses

        However, if a current, i, has been established in the primary circuit, total energy to be
dissipated is equal to the work done, that is, if L is the inductance, then

                                         Total energy = ½ Li2
        Therefore, ignoring resistance, there will be a residual electromagnetic energy given by

                                             ½ Li2 - ½ CV2
after the capacity discharge has occurred. It is also likely that the spark gap will remain ionized for a
brief period after completion of the capacity component favoring continuation of the discharge of
the residual energy, at lower current and voltage, until the voltage falls to a value too low to
maintain discharge. Thus, in the induction coil discharge, the capacitance component is always
present. If the condenser or self-capacitance of the secondary circuit is not charged up to the
breakdown potential of the electrode gap, no discharge will occur. But once the capacity spark is
obtained, conditions at the gap may be sufficiently favorable to allow a proportion of the residual
electromagnetic energy, arising from the work done in establishing a current in the primary circuit,
to be discharged as an induction component, sometimes referred to as an arc or flame. Two distinct
phases of breakdown have been given, namely,

(1)     a phase of high current density and high electric field and

(2)     a phase of comparatively low current, a low electric field and high gas temperature; the so
        called induction component which, due to the high temperature and longer spark duration,
        can be responsible for appreciable erosion of the electrodes.

        This causes pronounced lines in a spectrum corresponding to the materials of the electrodes;
the fainter gas lines are due to the first, or capacitance phase.

       The division of energy between the two phases can be arranged at will; the capacitance
component, ½ CV2, can be changed by varying C and/or V, the breakdown voltage. Altering the
gap changes V. If C and V are large enough the inductance component may be negligible. If either
C or V is further increased, without increasing ½ Li2, spark will fail owing to insufficient energy. A
change in L or i will change the inductance component.

         Thus, the minimum energy required for discharge is ½ CV2. However, this is not a practical
minimum since the inductance component will not normally be negligible, that is, the primary
circuit energy ½ Li2, the inductance, will be greater than ½ CV2, the capacitance.

        In general, the total circuit energy required for ignition will be greater for an induction coil
spark than for a purely capacitance spark. This is also true in respect of a spark obtained from a
magneto, which consists of an electric generator and an induction coil, which are combined for
convenience and compactness.

        The discharge of current in the secondary gap of a basically induction coil system can have
a high degree of asymmetry: this apparent rectification varies in magnitude and direction with the
length of the gap: the reversal of direction indicates a polarity effect that depends upon the
dimensions of the gap. Over 50% of the original energy in the primary circuit may appear as direct
current (DC) in the spark discharge; this is part of the so-called "inductance" component. Its other
contribution is that induced by residual oscillations or decay of current in the primary circuit.
However, for specific values of the spark gap and constants for the circuit, including the degree of
inductive coupling between the primary and secondary circuits, the unidirectional component may
contain all the energy still in the primary circuit, after the initial, very brief, capacitive discharge has
occurred. That is, there will be no residual oscillations in the primary circuit and the inductance
component will be wholly unidirectional. Spark discharge has sometimes known to fail at a
particular gap setting; changes in the gap setting or polarity reversal are required to restore the
spark. Hence, the importance of polarity in certain circumstances.

Effect of Gas Movement on Spark Ignition

         In a combustible mixture of gas and air, laminar and turbulent flows have opposing effects
on the requirements of ignition though experimental results do not entirely support this. One report
showed that for laminar flow, the likelihood of ignition increased with increase in gas velocity. This
is attributed to the tendency of the initial flame to escape with the aid of gas flow from the
quenching effect of electrodes. On the other hand, when the gas flowed turbulently, ignition tended
to be suppressed, probably due to increase of diffusivity of gases in the eddies.

         Sometimes, turbulence is deliberately introduced to promote better mixing; the degree of
which can affect requirements of ignition and subsequent combustion. For example, the combustion
chamber of a spark ignition engine usually has a squish area under the cylinder head in order to
introduce turbulence. See Fig. 9. Usually, this has a beneficial effect on the subsequent combustion
initiated by the spark. In these conditions, provision of an adequate spark is usually considered to be
of secondary importance. Whether such turbulence, deliberately introduced to promote better
mixing or a change in flow characteristics, facilitates ignition or not, depends on whether it makes
the mixture ratio in the vicinity of the spark gap more favorable for ignition than that in the absence
of turbulence. Thus, turbulence can have an effect, favorable or not, in ignition, owing to a change
in the local mixture ratio. However, the subsequent total combustion of the entire mixture is another
matter. That is, a proportioning of the mixture in relation to the geometry of the combustion
chamber could favor ignition. Thus, by definition, ignition is the initiation of combustion but not the
completion of the combustion of the whole charge.

Drawbacks of Electric Sparks

        Electric sparks are very hot and fast-acting ignition sources. Because the discharge time of
an electric spark is very short (of the order of 10-8 to 10-7 s), the energy that is imparted to the gas at
the end of the discharge period is highly concentrated, so that a very steep temperature profile with
a very high temperature at the center is established. In this initial stage of flame development, the
chemical liberation is insufficient to maintain such steep temperature profiles, so that the profile
broadens and the temperature at the center decreases within a period of time which depends on the
physical and chemical properties of the gas, and provided that the discharge energy is sufficient, the
profile develops to that of a minimal flame and thence continues to propagate as a combustion wave
and the temperature in the center becomes approximately the flame temperature.

        The process of spark ignition depends on many parameters such as energy, peak voltage,
duration of discharge, geometry of the spark gap, and its location relative to the particular geometry
of the compressed charge.

       However, the conventional spark plug system is quite ineffective when igniting a lean
mixture below an equivalence ratio of about 0.8 for gasoline and air.

        Some studies have been carried out on high-energy ignition systems. Such systems give an
increased peak power to the spark and also increase the duration of the spark.

       The development is based on the premise that by proper control of the manner in which the
plasma is generated by the spark discharge, leading in particular to the prolongation of its duration,
more unburned mixtures can be exposed to it, an effect that can be greatly enhanced by the use of
squish whereby a significant amount of mixture can be caused to pass by the spark gap and be
mixed with the plasma, generating thereby in effect, a system of distributed ignition sources.

         Tests have shown on single as well as multi-cylinder engines that increasing the gap width,
its projection and the duration of the spark (with duration enhanced from 1 to 2 ms to 5 to 10 ms),
the mixtures 10-15% leaner than those ignited by the conventional spark plug) can be ignited.
According to Aiman3, based on the ignition system of Johnston and Neuman4, the amount of
exhaust gas recirculation, which a single cylinder engine tolerated without misfire, increased when
the duration of the spark was increased at a given arc current. This may be due to multiple
opportunities for ignition or due to increased ignition energy.

        The use of multiple ignition points (usually limited to two spark plugs) as compared to a
single spark plug has also improved the combustion characteristics of the spark ignition engine.


1.       Battery ignition system where the high voltage is obtained with an ignition coil (coil
         ignition system).

2.       Battery ignition system where spark energy is stored in a capacitor and transferred as a high
         voltage pulse to the spark plug by means of a special transformer (capacitive discharge
         ignition or CDI system).

3.       Magneto ignition system where the magneto - a rotating magnet or armature - generates the
         current used to produce a high voltage pulse.

        The combustion process is initiated by a spark, which is provided by a magneto or battery
and coil. The spark occurs between the simple electrodes of a spark plug; the spark is triggered off
by the timed operation of a contact breaker in series with the primary winding of a coil or magneto.
The circuit constants are so chosen so that a collapse of the primary current is sufficient to ensure a
spark discharge between the spark plug electrodes. They are fitted on all production cars and require
maintenance but is adequate for the purpose. An improved system would be one with reduced
maintenance, increased reliability, and extended range of satisfactory operation of the engine.

     GMR-2230, 1976
     SAE Paper No. 750348
        Within limits normally encountered in engine operation, provided the energy of the spark is
sufficient, it has little or no influence on establishment of combustion although excess energy can
be harmful for the spark plug electrodes because it causes erosion of the electrodes and can lead to
preignition. The energy for spark in the circuit provided by conventional systems varies
between about 10 to 100 millijoules (mJ) with energy of approximately 1 mJ stored in the
self-capacity of the plug and cable, and thus available for immediate release at the spark plug
electrodes. In a single cylinder engine test, 0.2 mJ was found to be sufficient.

        Other researchers have estimated that the normal total 20 mJ is divided into 2.5 mJ released
during the capacitance phase of the spark discharge for 1 microsecond (s) and 17.5 mJ for the
induction phase for 1 millisecond (ms). Normally, ignition can be attributed to the capacitance
component, although, in cases of more difficult sparks caused, for example, by wet or fouled spark
plugs, the induction components assists. The conventional automobile engine ignition system
delivers sufficient energy based on a minimum average spark gap of about 0.8 mm, over the whole
engine operating range.

       About 0.2 mJ of energy is required to ignite a quiescent stoichiometric mixture at normal
engine conditions by means of a spark.

        For substantially richer and leaner mixtures, and where the mixture flows past the
electrodes, the energy required may be an order of magnitude greater (about 3 mJ).

       Conventional ignition systems deliver between 30 to 50 mJ of electrical energy to the spark.
Due to the physical characteristics of the discharge modes discussed above, only a fraction of the
energy supplied to the spark gap is transmitted to the gas mixture.

       Radiation losses are small throughout.

        The end of the breakdown occurs when a hot cathode spot develops, turning the discharge
into an arc; heat losses to the electrodes then become substantial.

        The breakdown phase reaches the highest power level (about 1 MW), but the energy
supplied is small (0.3 to 1 mJ). The glow discharge has the lowest power level (about 10 W) but the
highest energy (30 to 100 mJ), due to its long discharge time. The arc phase lies in between.

       Energy supplied by a conventional system is not invariable. Now, the total circuit energy is
given by
                                             ½ Li2

where i is the current in the primary coil at the time it is interrupted by the contact breaker. During
the period when the contacts are closed, current increases from 0 to i. If this period is less than the
time constant for the growth of the current (which is constant for the circuit) the current will be less
than 63% of its maximum. For a period equal to three times the time constant, the current will have
grown to 95% of its maximum value. Thus, a requirement for the development of almost the
maximum current and hence almost the maximum energy is a sufficiently long period between
closing and opening of the contacts, a condition met only at low engine speeds. This sets an energy
limit for low engine speed, because of the maximum current the contact breaker points can safely
interrupt. At high engine speeds, the period during which the contacts are closed can be too short
for growth of the current to be sufficient to provide adequate energy or voltage, so misfiring may
occur. Although it is desirable that the duration of closure of the contacts be as long as possible, it is
limited by the need to have the contacts open for the period comparable with the time constant for
spark discharge. If a magneto instead of a battery provides the electrical energy, the problem at high
speed is alleviated automatically since the electrical output of the magneto is greatest at high
speeds. However, at low speeds, a much reduced and possibly an insufficient amount of electrical
energy is available. This is a well-known limitation of the magneto

        The introduction of solid-state circuitry for spark ignition does not violate the basic
concepts of energy requirements. The efficiency is increased while total energy dissipated may be
reduced. The reliability of the system and repeatability of the operation may be improved. The
simplest design merely assists the conventional contact breaker to interrupt the primary current
while other features remain unchanged. In more advanced designs, the contact breaker is
eliminated. The fullest application of solid-state circuitry affords, additionally, a form of spark
discharge that is probably more favorable for ignition in adverse conditions. However, a relatively
simple, transistor-assisted contact breaker system can offer a two-fold advantage. Because the
current interrupted by the CB point is much reduced, an appreciable increase in primary current is
possible. Since current appears to the power of 2 in the relation ½Li2, it would permit a
considerable reduction in primary inductance, L, for a given amount of energy which, in turn,
would reduce the time constant for the growth of the current in the circuit. Thus the limitations of
energy at high engine speeds can be alleviated. Another modern device, the surface discharge plug,
is based on the principle which permits the plug to fire even when contaminated with deposits to a
degree that would virtually cause the short circuiting of a normal type of spark plug

        The voltage required to breakdown the resistance of a plug gap decreases with increase in
temperature of the electrode. In some engines, incorrect polarity (positive polarity of the hotter,
central electrode of the spark plug) can cause a 35-40% increase in the breakdown voltage. The
energy content of the capacitance component of the spark, ½li2, is raised. If either the available
energy is insufficient to provide this amount plus losses or the voltage, V, is, for some reason,
attainable, there will be no spark

          The voltage requirement for sparking shows the important difference between controlled
laboratory experiments and the actual working ignition system of an engine. In the laboratory,
difficulties of obtaining breakdown voltage of the spark gap are first overcome, in readiness of the
experiment, and a selection, by trail and error, of the correct capacitance, C, the minimum energy in
terms of ½CV2, is readily provided. In a practical, engine ignition system, C is fixed, and two
aspects of V must be considered. First, the required breakdown voltage will vary over a period of
time with, for example, the electrode gap, temperature and cylinder pressure. Secondly, the
attainable voltage will depend upon conditions of the spark plug and associated circuits. Therefore
it is feasible that a failure to spark could arise not necessarily as a result of insufficient ½Li2 energy
in the primary circuit to meet minimum energy requirements (½Li2 + losses) but merely because,
perhaps, owing to insulation leakage, the required breakdown voltage could not be attained. Thus,
energy would play a subordinate role to voltage.
        A spark can arc from one plug electrode to the other only is a sufficiently high voltage is
applied. When breakdown of the resistance of the gap occurs, ionizing streamers then propagate
from one electrode to the other. The impedance of the gap decreases drastically when the streamer
reaches the opposite electrode, and the current through the gap increases rapidly. This stage of the
discharge is called the breakdown phase.

        It is followed by the arc phase; where the thin, cylindrical plasma expands largely due to
heat conduction and diffusion and, with inflammable mixtures, the exothermic reactions, which
lead to a propagating flame, develop.

        This may be followed by a glow discharge phase, where, depending on the details of the
ignition system, the energy storage device, e.g., the ignition coil, will dump its energy into the
discharge circuit.

         The Breakdown phase is characterized by a high voltage (about 10 kV), high current (about
200 A) and an extremely short duration (about 10 ns). A narrow (about 40 m diameter) cylindrical
ionized gas channel is established very early. The energy supplied is transferred almost without loss
to this column. The temperature and pressure in the column rise very rapidly to values up to about
60,000 K and a few hundred atmospheres respectively. A strong shock or blast wave propagates
outward, the channel expands, and, as a result, the plasma temperature and pressure fall. Some 30%
of the plasma energy is carried away by the shock wave; however, most of this is regained since
spherical blast waves transfer most of their energy to the gas within a small (about 2 mm diameter)
sphere into which the breakdown plasma soon expands.

        A breakdown phase always precedes arc and glow discharges; it creates the electrically
conductive path between the electrodes. The Arc phase voltage is low (<100 V), though the current
can be as high as the external circuit permits. In contrast to the breakdown phase where the gas in
the channel is fully dissociated and ionized, in the arc phase the degree of dissociation may still be
high at the center of the discharge, but the degree of ionization is much lower (about 1%). Voltage
drops at the cathode and anode electrodes are a significant fraction of the arc voltage, and the
energy deposited in these electrode sheath regions, which is conducted away from the metal
electrodes, is a substantial fraction of the total arc energy (heat loss is about 45%). The arc requires
a hot cathode spot, so evaporation of the cathode material occurs. The arc increases in size due
primarily to heat conduction and mass diffusion. Due to these energy transfers, the gas temperature
in the arc is limited to about 6000 K; the temperature and degree of dissociation decrease rapidly
with increasing distance from the arc axis.

        During Glow discharge, currents are less than 200 mA; there is a large electrode voltage
drop at the cathode (300 to 500V); a cold cathode and less than 0.01% ionization are typical.
Energy losses are higher than in the arc phase (about 70%) and peak equilibrium gas temperatures
are about 3000 K. Figure 11 shows variations of current and voltage with time for conventional
spark ignition system.

        The minimum ignition energy required to ignite a pre-mixed fuel-air mixture depends
strongly on mixture composition. Figure 10 shows a typical set of results on the minimum ignition
energy as a function of the equivalence ratio under quiescent conditions5. The curve shows a
minimum for slightly richer than stoichiometric mixtures; the minimum energy required for
successful ignition increases rapidly as the mixture is leaned out. While the initial plasma kernel
growth (up to 10 to 100 s) is not greatly affected by the mixture strength, the inflammation process
and the thickness and rate of propagation of the resulting flame are strongly affected. Because the
chemical energy density of the mixture and flame temperature decreases as the mixture is leaned
out, the flame speed decreases and the flame becomes thicker. Thus more time is available for heat
losses from the inflammation zone; less energy is available to offset these losses, and the rate of
energy transfer into the zone decreases. The consequence is that, as the mixture is leaned out, the
plasma must grow to a larger size before inflammation will occur, substantially, more energy must
therefore be supplied to the discharge.

       A set of rules for the design of an ignition system of a petrol engine cannot be made.
Fundamental principles of energy requirements for ignition have been set forth in respect of
systems for such engines.

        High-energy ignition systems are quite feasible. Provision of extra energy may be beneficial
in certain cases where conditions adversely affect ignition.

       Generally, energy contained in a spark, in excess of that required to bring about ignition,
can be detrimental to the spark plug and the engine performance. In conventional spark ignition
engine ignition systems, it causes rapid wear of contact breaker points because of the increased
primary current they have to interrupt. The choice of optimum energy, sufficient to allow for a
degree of diversity, yet not enough to cause unacceptable penalties in other directions, is a matter of
expediency or opinion6.

         Namazian et al7 studied the process of flame initiation by using spark photographs at
intervals of 0.01ms (about 0.08 crank angle degree) after spark initiation. They observed a kernel
(high temperature zone around the discharge) immediately following the spark discharge, which
fills up the entire electrode gap within 2 degrees after the discharge while the spark is no longer
visible. Subsequent emerge of a flame structure similar in appearance to the fully developed flame
is observed a few degrees after the spark.

       A simple criterion for the successful ignition is that the rate of heat release in the small
flame zone surrounding the spark should exceed the total rate of heat loss.
       This implies that the characteristic dimension of the flame kernel should equal or exceed
the quenching distance.

  Ballal, D.R. and Lefebvre, A.H., “Influence of Flow Parameters on Minimum Ignition Energy
and Quenching Distance,” Proceedings of Fifteenth International Symposium on Combustion, p
1473-1481, The Combustion Institute, 1974.
  Hurtley, D., "Ignition Part I", Automobile Engineer, 59(1969): 96 and "Ignition Part II",
Automobile Engineer, 59(1969): 148.
    SAE Paper No. 800044.
        Assuming the kernel to be a sphere of diameter d, minimum energy deposition to achieve
ignition is

                       E '(min)  C p T  d 3 / 6
Minimum ignition energy in air at 1 atm, 20C

Fuel                           E’ (10-5J)
Methane                        33
Ethane                         42
Propane                        40
n-Hexane                       95
Iso-Octane                     29
Acetylene                      3
Hydrogen                       2
Methanol                       21

       The quench distance and distance to obtain the optimum gap are taken equal. Ballal and
Lefebvre found d (gap) = 5d (quench)

        Low pressure experiments are not easily applied to an engine. In an engine, if the velocity
past the plug is high, the kernel may detach from the plug and move down stream before it
develops into a full-fledged flame front. In many cases the effective point of ignition is not the
spark plug location. Such high flow velocities are not typical, but can be produced by high swirl.
Under some conditions, the plug may act as a flame holder and the flame may rotate, thus
producing an “apostrophe” shaped burned gas volume, the narrow end being at the spark plug.
For more typical cases, the spark is located on axis and the flame kernel is only slightly distorted
by the local turbulent velocity. This distortion changes the surface to volume ratio of the kernel
and the wall surface area in contact with the kernel. Both these effects change the heat transfer.
Such fluctuations in heat transfer affect the kernel growth rate, especially for lean mixtures. This
leads to CBCV of the start of rapid combustion.

        In a piston engine, the cylinder gas turbulent intensity is closely related to large-scale
velocity patterns in the cylinder. These patterns are initiated by the flow as it enters the cylinder
from the intake port. Swirling flows around the cylinder axis and vertical vortex motions called
tumble are two common structures. These large-scale patterns are modified by piston speed and
combustion chamber shape as they undergo compression. It is decay of these large flows which
provides the major source of turbulence at the time of ignition. In the case of low swirl, tumble
motions dominate. Tumble motions break down under compression more readily than do swirl
motions, and thus give high turbulence intensity at the time of ignition. Creation of a definite
swirl pattern, however, reduces cyclic variations. Low swirl rates allow the less repeatable
tumble motions to dominate, giving large cyclic variations. Thus the desire to create high burning
velocities may lead to flows which cause cyclic variability because of the flow interaction with
the flame kernel.

       From the basic experiments it can be concluded that the conditions which are most
conducive to spark ignition and which lower the minimum spark energy are:
   1.      Low burning velocity
   2.      High initial temperature
   3.      High mean reaction rate
   4.      Low volumetric heat capacity,  Cp
   5.      Low thermal conductivity
   6.      High total pressure
   7.      Nearly stoichiometric mixture
   8.      Low turbulence intensity
   9.      Electrode separation distance close to quench distance

        In spark ignition, homogeneous-charge engines, the lean limit of flame propagation can
    be extended by:
    1.      Increasing mixture homogeneity
    2.      Decreasing charge dilution
    3.      Increasing compression ratio
    4.      Decreasing engine speed
    5.      More central spark plug location
    6.      Use of multiple spark plugs

         In a quiescent mixture, minimum ignition energy is less than 1mJ. An order of magnitude
more energy is required for flowing mixtures. Conventional S.I engine ignition systems deliver
30 to 50 mJ of energy.
         A larger initial kernel size is important because it reduces the surface to volume (area to
volume) ratio. Thus the larger the radius the lower th e area to volume ratio. The larger kernel
surface area may also be affected by a larger range of turbulent eddy scales, thus increasing the
initial flame speed.

        If the same amount of source energy were delivered by an electric current, over a time larger
than the time of development of a minimal flame, the temperature at the core would drop below the
flame temperature, the heat liberation in the reaction zone would not attain a balance with the
outflow of heat into the preheat zone, and the flame would become extinct. On the other hand, if the
current flow were continued for a longer period the temperature profile ultimately would become
sufficiently broad, and the temperature in the core sufficiently high, so that the heat liberation in the
reaction zone overbalances the outflow of heat and ignition occurs.

        Energy in a wire of resistance R carrying current i for a time t is given by iRt.

         If the length of the wire does not exceed half the quenching distance for a particular
mixture, it will remain within the sphere of the minimal flame and the significant energy will be
that of the whole wire and not the energy per unit length. According to one investigator, the energy
required in the wire to achieve an ignition probability of 50% increases with the period of heating
and the wire diameter. It was noted that sometimes ignition was obtained with energies greater than
that required to melt the wire, hence it was concluded that the flow of current, sometimes continued
after fusion of the wire until ignition occurred. This is feasible, as the circuit is complete until the
wire actually collapses into droplets after which electric cores form and are extinguished as the
droplets separate. Thus under certain conditions, attainment of ignition does not indicate the
threshold of ignition energy, but merely shows that the wire has melted.

        The system is based on the concept that the combustible mixture extracts initially
instantaneously (up to 1 ms) from the source the minimum energy required for ignition and that the
remainder of the energy supplied is wasted.

        Ignition by heated metal strips has also been investigated. The temperature of the metal
strips was raised electrically until the surrounding mixture was ignited. When ignition temperatures
were recorded, the catalytic effect of certain metals became apparent. Convection current in the gas
tended to increase ignition temperature, indicating that when the gas flowed over the heated strip, it
received less heat than it would have received under quiescent conditions.

       Three independent quantities characterize the ignition threshold of a slow source:

(1)    the total heating time or the time during which the current flows called the critical heating

(2)    the total energy delivered during this time called the critical source energy which defines the
       current strength and

(3)    the temperature Tc in the core at the end of the heating period called the critical source

         Figure 9 shows the general form of the relationship between the three quantities.
         The minimum ignition energy, denoted by hc, corresponds to a very short (zero) value of the
critical heating period. The corresponding critical source temperature has a value lower than the
source temperature, Tb. As the critical heating period is increased (and the current strength is
correspondingly decreased) the critical source temperature decreases and the critical source energy
increases. Initially these changes are large and become gradually smaller. The temperature curve
becomes asymptotic which is inherent in Arrhenius’ law that at low temperature changes in a few
degrees of temperature produce large changes in the rate of chemical reaction.
                      IGNITION BY FLAME OR HOT JET
        Research into the effects of size and temperature of flame on their ignition characteristics
has been carried out but there appears to be no experimental evidence of threshold ignition energy
such as exists in spark ignition system. This may be due to the presence of too many influential
factors. For example though it is possible to control a flame to a given energy dissipation per unit
time, as for a heated wire the time required to stabilize the flame is likely to exceed the critical
ignition time. For a turbulent flame introduction of kinetic energy would complicate the assessment
of energy. It is known, however that a turbulent flame or jet entering a combustion chamber, an
appreciably leaner fuel air mixture than can be performed by a spark.
        If the threshold ignition energy defined by spark ignition were distributed in space and time,
as in the energy of a flame it is very unlikely that ignition would be achieved. Therefore if threshold
energy exists for ignition by a flame, it is likely to be higher than that for a spark. However since it
is known that under certain conditions, a flame can ignite a mixture when a spark will not, some
features of the flame must be particularly favorable for the initiation of combustion. It is likely that
the volume is relatively large, the energy presumably sufficiently dense, and the area of contact
between the source of heat and the reactants large. A degree of turbulence in the flame would
promote a favorable increase in the mixing of the flame with the combustible medium although too
much turbulence might hinder combustion. Also a flame is not only a source of heat but also
contains a mixture of reactants and products of combustion, it has a high concentration of free
chemical radicals, which may mean that energy is available in a particularly advantageous, form.

        Turbulence in a hot igniting jet which raises the ignition temperature and small eddies due
to break up of the jet, is likely to be less efficient agents than large volumes of hot gas in a laminar
flow. A hot air jet is better than a hot nitrogen jet for a stoichiometric fuel air mixture. Also the
richer the mixture, up to the fuel rich limit, the more readily it ignites with hot air. This suggests
that an igniting jet containing oxygen favor the ignition of a fuel rich mixture. On the other hand,
the presence of hydrogen in a relatively inert gas jet drastically reduces the ignition temperature of a
methane air mixture and the most ignitable mixture approaches the lean limit. The reason suggested
is that first hydrogen ignites with the surrounding air and then the resulting flame triggers the
explosion of the methane air mixture. Methane appears to inhibit the first stage so that a rich
mixture with methane becomes less ignitable. Similar effects have been noted with carbon
monoxide present in the jet.

         The ignition system using flame jet ignition consists of a divided chamber with its volume 2
to 3 percent of the clearance volume of the engine8,9 with one or more sharp edged orifices of cross
section area 3 to 5 mm per cm of pre-chamber volume to produce the flame jets. The overall
pressure built up in the pre-chamber is carefully maintained at a sub-critical level so that the jet is
essentially subsonic in order to prolong as much as possible the process of partial oxidation of the
rich mixture where the combustion was started (by means of a spark plug in the pre-chamber) as it
is ejected into the main charge containing a large amount of excess air. The equivalence ratio in the
pre-chamber varies between 1.4 to 2.5 while that in the main chamber it is of the order of 0.5.

        The jet comes out at high velocity and the ensuing turbulence it creates shears the flame
apart so that in effect it is temporarily extinguished. As a consequence, a large number of small size
turbulent kernels of flamelets, or active particles are seeded throughout the charge. After a short
induction period, rapid combustion of the lean mixture in the main chamber is thus initiated at a
large number of distributed ignition sites.

        The lean misfire limit has been claimed to be raised to an air-fuel ratio of 33:1 with a
definite improvement in the fuel economy and a significant decrease in the fuel octane requirement.

    Gussak SAE 750890
    Gussak et al SAE 790692
Ignition is believed to be due to the action of the methyl radicals in addition to the hydrogen atoms
mentioned earlier, which enhances the chain branching mechanism. Explanations to the phenomena
are based on the thermal theory10.
        Figure 13 & 14 show some designs of combustion chamber arrangements to produce a hot
jet which would ignite the charge in the main chamber.

        A flame or hot jet, whether laminar or turbulent, is probably a less efficient source of
ignition than one in which dissipation of energy is confined to a small volume and a short duration.
But there are several features of the flame/hot jet and both appear to be capable of igniting
combustible mixture under circumstances in which other ignition systems fail. A threshold energy
concept probably applies but a carefully devised experiment would be required to prove the point.

         The foregoing is neither authoritative nor exhaustive but defines quantitatively the problem
of ignition by a flame or hot jet, placing the role of ignition in its proper perspective. Evidently, this
type of ignition system warrants further study even at steady flow conditions, let alone at pulse
operation conditions required in the internal combustion engines.

       The ignition of a relatively lean air-fuel mixture by a flame - in particular, a burning jet - is
not simply a matter of energy. At least, it is partly attributable to other features of such a flame that
are more favorable to combustion than is an electric spark.

                                PLASMA JET IGNITION
         In the plasma jet igniter, the spark discharge is confined to a recessed cavity provided with a
discharge orifice, while the electrical power supply is augmented by the addition of a condenser that
helps discharge at a relatively low voltage and high current through the spark generated in a
conventional manner by a high voltage, low current ignition system10,11. The circuit consists of a
conventional high voltage ignition coil, which is used to produce an electric spark that closes the
circuit by the ionized passage it creates. This causes a condenser, charged from 900 to 1200 volts,
to shorten, forming high temperature plasma. Stored energy of up to 10 joules can easily be
employed, but typically only 1 or 2 joules are required. The high temperature plasma is created so
rapidly that the cavity is pressurized, causing a supersonic jet of plasma to be issued through the
orifice and penetrate into the charge10.
         Experiments conducted in a constant volume bomb where the combustible mixture was
initially at rest, and at atmospheric pressure and room temperature, have revealed the following:

1.       The plasma jet entered the combustion chamber in the form of a turbulent flame, which was
         embedded in a blast wave by a hemispherical shock front.

2.       The gas dynamic effects of the blast wave were dissipated by the time combustion started,
         after a delay of about 1 ms so that ignition took place in the turbulent zone of the plume.

     Dale and Oppenheim SAE 810146.
     SAE Paper No. 830479
3.     The depth of penetration of the jet was solely a function of its initial velocity; it could thus
       be controlled by the amount of energy deposited in the cavity, as well as its size and that of
       the exit orifice.

4.     In direct contrast to spark ignition, which produced a laminar flame, which later became
       turbulent, here combustion was initiated in the form of a turbulent flame, which upon
       leaving the plume tended to acquire a laminar character. As a consequence, the normal
       burning speed, which was initially quite high, decreased monotonically as the flame kernel

5.     It was found that the most effective feedstock for ignition was hydrocarbons initially in the
       liquid state. This was believed to be due to the action of the hydrocarbon atom, which was
       in abundance in the plasma created from such feedstocks.

6.     Plasma jets were shown to be capable of igniting gaseous mixtures below the normal
       flammability limit.

       Studies on single- and multi-cylinder engines have shown that with plasma jets, lean
mixtures of the order of 18:1 could be ignited. According to Ref. 11, equivalence ratios as low as
0.54 can be ignited.

        The drawback is that plasma jet igniters require more electrical energy (1 J) as compared to
conventional igniters (50 mJ) per pulse. As a result they also suffer from high electrode erosion
rates. More work still needs to be done to obtain data on the performance, fuel economy and
emission characteristics. Its main advantage is its extremely short discharge time: about 20 s,
which makes it capable of igniting lean mixtures.
        Figure 15 shows the schematic of a plasma jet ignition. Figure 16 (a) to (d) show
comparisons of plasma jet (PJ) system with standard spark (SS). PJ system have an extended lean
limit (25:1) compared to SS (22:1), produce higher power (Fig 16 (b) & (c)) and lower sfc (16 (b) &
(c)) and combustion duration is shorter (16 (a) & (d)).
                          PHOTOCHEMICAL IGNITION
        Norrish12 studied the combustion of air-fuel mixtures by photochemical methods. These
methods are based on the phenomenon of photolysis, a method of established significance to
chemical kinetic studies of oxidation reactions. Figure 1013 shows a photochemical igniter. Since
the igniter is effective only with windows transmitting in vacuum, ultra-violet radiation, it was
concluded that ignition is, in effect, caused by the action of oxygen atoms. This is caused by the
dissociation of oxygen atoms, which in turn is caused by radiation below 245 nm (nanometers),
absorption of 180 nm being most efficient in this respect. It was found specifically that for ignition
of a hydrocarbon-air mixture the critical concentration of oxygen atoms was of the order of 1014
atoms per cubic centimeter. It was established, moreover, that the energy requirement to initiate
combustion is essentially independent of the air-fuel ratio of the mixture. The most interesting
feature of this ignition system is that it is evidently capable of initiating combustion under
conditions similar to the plasma jet system with about the same expenditure of energy. However,
the plasma is, in this case, physically separated from the mixture by the window, guaranteeing that,
unlike in the case of the plasma jet, its effect is solely chemically kinetic in nature.

                               MICROWAVE IGNITION
        Ward14 studied the possibility of applying microwave energy to a burning hydrocarbon fuel-
air mixture in a conventional internal combustion engine in order to stimulate the burning of a lean
mixture. He showed that typical flame front electron plasma properties and combustion chamber
geometries are such that significant conversions of microwave energy to electron energy occur. In
turn, the electrons will give up a substantial portion of their excess energy to excitation of internal
energy levels of molecules, which are known to accelerate chemical reaction rates. In this was
microwave energy could be used to excite the relatively cool flame front molecules to more stable

       Ward and Tu15 carried out a theoretical study on the effects of microwave in treating the
mixture at the flame front with particular application to engines. Effective transfer of microwave
energy to the flame front electron energy occurs. The feasibility, therefore, of efficient microwave
heating of the flame front electrons in engines was established. These conclusions corroborated
flame experiments carried out later13.

   10th Symposium on Combustion, p 1-18, 1963.
   SAE Paper No. 810146
   Journal of Microwave Power, 12(3), 187-199, Sept. 1977.
   Combustion and Flame, vol. 32, p 57-71, 1978.
                                    LASER IGNITION

        The ignition of lean mixtures and diluted mixtures (diluted with exhaust gases) becomes a
problem. By raising the spark energy and the spark duration the spark is still tied to the relatively
cold chamber wall resulting in slow initial combustion wave development. Using the energy burst
from a focused laser beam permits the spark location to be moved away from the chamber wall and
the use of higher spark energies without heating the protruding ground electrodes to temperatures
high enough to cause pre-ignition.

        According to Hickling and Smith16 who carried out studies using several hydrocarbon fuel-
air mixtures in a bomb at elevated temperatures and pressures, the laser spark will ignite very lean
mixtures. Ai-fuel ratios of up to 31:1 ( = 0.43) were successfully ignited.

        Dale et al17 operated a single cylinder engine using energy bursts from a focused carbon
dioxide laser as the ignition source operating at 16 m wavelength. Carbon dioxide laser has the
advantage of being far more efficient than those operating in the visible and near infrared
wavelengths and also breakdown can be achieved with lower pulse energies. The minimum energy
required in the spark burst of the laser to ignite the mixture seems to be dictated by the energy
required to obtain breakdown in air at the same pressure. Raising the spark energy above this
minimum level is desirable to produce a steady running engine. They were also able to ignite a
mixture containing up to 16% recirculated exhaust gas, provided to reduce nitric oxide, which
increased if a laser system was used as igniter.

        The laser beam was focused at a small point within the mixture to achieve breakdown,
generating thereby a plasma kernel that acts as the ignition source. Dale et al focused the laser beam
near the center of the combustion chamber.
        Figure 17 compares cylinder pressure traces between laser, plasma jet and standard spark.
The laser and PJ systems show stable running at 22.5:1 A/F, The SS shows no firing.

     SAE 740114, Transactions
     SAE 780239
                                  PUFF-JET IGNITION

         The nature of the plume of hot gases, which expand from the cavity of the plasma jet
igniter, has been shown to be similar to atmospheric thermals. In addition to this, the mixing of a
plasma jet plume with the ambient gas has been shown to be very rapid due to the turbulent nature
of the plume.

        The puff-jet ignition concept described here involves the rapid mixing of a turbulent plume
created by injecting a small volume of combustible gas such that a thoroughly well mixed, near
stoichiometric, turbulent “puff” is formed in the region of the spark plug electrodes. The rapid
mixing of the plume results in a near stoichiometric mixture in the vicinity of the spark gap, thus
producing charge stratification in the combustion chamber.

        Pitt et al18 carried out experiments in a combustion bomb. The fuel injected was methane
gas. The charge to be ignited was a mixture of methane and air. They were able to ignite mixtures
of equivalence ratio 0.67 with a low energy spark. Other advantages in using this technique include
reduced delay and improvement in burn rate. Further improvement in delay period was possible if
hydrogen was injected instead of methane without any improvement in burn rate. The system
compares well with plasma jet igniter devices with lower energy consumption and reduced
electrode erosion.

        The system was later tried in a methane-fueled single cylinder engine19. The small fuel
quantity (about 1% of the total fuel charge) was admitted under pressure by a fast acting valve. A
turbulent puff of gas that rapidly mixed with the surrounding lean mixture was directed towards the
spark gap to create a local region of stoichiometric mixture, which could ignite and produce a
rapidly growing ignition kernel. This ignition kernel was similar to that produced during the early
phases of plasma jet ignition.

         The technique required less than 100 mJ of electrical energy compared to 1 J for the plasma
jet ignition system. The electrodes were found to be less eroded. While it had features similar to the
ignition and combustion characteristics of the direct injection stratified charge system, it did not
require bulk cylinder swirl flows to augment the mixing and combustion rate of the pilot fuel. This
should allow for simpler combustion chamber design.

       Fisher et al20 have found that by using a puff-jet igniter, output power increased when used
under lean mixture conditions. They found it suitable for natural gas because it burns more slowly
and is more difficult to ignite than conventional fuels. Hence it is useful for lean methane-air
mixtures. It produces less cycle-to-cycle variations in power output.

   Combustion Science and Technology, vol. 35, p 277-285, 1984.
   Pitt et al, Combustion Science and Technology, vol. 38, p 217-225, 1984
   SAE Paper No. 860538

To top