Experimental analysis of a double spark ignition system

B.Hnatiuc(1), S.Pellerin(2), E.Hnatiuc(1) and J.Chapelle(2)
(1) Univ.Techn. “Gh.Asachi”, Catedra de Bazele Electrotehnici, Bd.D.Mangeron nr. 51 – 6600 Iasi, Roumanie
(2) LASEP – Centre Univ. de Bourges, BP 4043, 18028 Bourges cedex 2 – France
Abstract. The spark that ignites the combustible mixtures is a glow discharge produced between the electrodes of a spark
plug, connected to the secondary of a coil at the high voltage. A good combustion require a frank spark, in a volume as
large is as possible and with a maximum of energy. We propose a solution to increase the plasma volume and present
electrical discharge parameters as function of inter-electrode distances, pressures in the test-reactor and the electrical
pulses width of the power supply.
Keywords: Ignition, Spark discharge, Power supply, Combustion engine
PACS: 52.00, 52.70, 52.80
The spark that ignites the combustible mix in an automobile engine, is a glow discharge produced between the
electrodes of a candle, connected to the secondary of a coil to high voltage. The ignition and the combustion in the
best conditions suppose to have a frank spark, in a volume as large is possible and with a maximum of energy.
The classic ignition uses a contact breaker and a coil boosting supplied at 12-13V that provide in secondary 15-
20kV. This level of high voltage is required to breakdown the spark between electrodes. This could happen only if
the high voltage exceeds a threshold value imposed by the dielectric constant of the mixture.
The purpose of this study concerns the conception and the realisation of a plasma ignition system for a
combustion engine room, increasing the plasma volume and the electrical power injected into the discharge in order
to have a quick and efficient combustion ignition. We have been interested to simultaneously produce two sparks at
high pressure (≤9atm), by a high voltage pulsed power supply. This power supply does not allow the transition of the
spark to an electric arc. The preliminary experiments have been done in air.

The experimental device uses a third auxiliary electrode #, not connected to the power supply, placed between
the main electrodes ! – high voltage electrode and ” – ground electrode of a typical spark plug supplied by a pulsed
high voltage power supply ) through the medium of a coil (. It is presented on Figure 1, and has been partially
described in [1]. A such geometry assure the existence of two spark discharges $ respectively between the
electrodes ! and #, called “first discharge” (referenced by 1), and between the electrodes # and “, called “second
discharge” (referenced by 2). The double spark is produced after an horizontally direction that could also help the
combustion of air-combustible mixture [2].
The pulsed high voltage supply allow to provide in the secondary of the coil a high voltage up to Ud=50kV at the
frequency of fd=100Hz with an adjustable pulse width τd between 0.05 and 6 msec. Furthermore, to study the
breakdown of each spark, the experimental device has used a time relay (opened ton≈15ms and closed toff≈10s). The
measurements have been taken in this way in order to eliminate any luminescent post-discharge inffluence due to the
metastable atoms persistance or any others species with important life time (typicaly few miliseconds [3]) into the
Few conclusions of the first experiments realized at high pressure with dry air are the following:
– The order of the spark breakdowns is always the same: firstly it occurs the spark 1 and then the spark 2.
– Once the sparks started the discharge voltage decreases and his tendency is to follow the minimum value to hold
the discharges; it is possible to observe also that the discharge voltage oscillations increase with the value of the
pressure into the experimental device so the instantaneous values of the electric field will increase with the
– At low pressures (p < 2atm) there are parasite discharges, D.B.D. type, produced around the ceramic insulation of
the main high voltage electrode !. These perturbations produce few difficulties for the electrical parameters
– For one pressure the sparks become instable and will stopped if the distance d2 between the electrodes ” and # is
constant increased.
– For one inter-electrodes distance, the sparks become instable to the variation of the pressure. The spark established
between the electrodes ” and # is more sensitive to this variation than the other spark.
– The utilization of the auxiliary electrode # permits to increase the maximum breakdown distance between the
electrodes, at atmospheric pressure from dmax=23 mm to dmax=d1+d2=28 mm. In this case the electrical power
injected into the sparks is from (2÷10)W for energy up to 60mJ [4].
– The sparks breakdown has been done for a maximum absolute pressure of 9 atm at a maximum distance between
electrodes dmax≈d1+d2=2+2 mm=4 mm.
– The sparks have a red-violet color at atmospheric pressure, and becomes blue and more shining when the pressure
– During the breakdown of the sparks all the energy accumulated in the magnetic field of the coil is consumed, after
a transient process to the gas, the secondary coil resistance and the shunt.
– There are three successive frequencies on the records of discharge voltage: The first one at ≈8kHz is due to the
breakdown and the power supply corresponding to an energy charge regime. The second corresponds to the
physical behavior of the sparks. The third one, like a relaxation phenomenon at ≈2.5kHz is influenced by the coil
and electrical circuit parameters. We have considered that the plasma is physically equivalent with a diode in series
with a low (dissipation) resistance and the oscillations appear because of the energy transfer into the
electromagnetic field [5] and the coil characteristics [6].
– It is possible to obtain multiple successiv sparks during the same electrical pulse (N2, N1 > 1), especially at
intermediary pressure. The mean discharge time <τ
i/Ni by spark is not significally modified; but on result, the
total discharge time τ
i during one electrical pulse is increases compare to the case with only one spark.

The presented double spark system seems well adapted for the ignition of combustible mixtures and it could
assure a better combustion by its two simultaneously discharges, therefore a bigger plasma volume for ignition. The
total distance between the electrodes has been increased from 0.7 mm to about 2 mm. Furthermore, the duration of
the two discharges is much lower than the electrical pulse widths τd: this phenomena allows to generate under
intermediary pressures successiv ignitions in each electrodes gap.
The experiments already done proved good results in real conditions with a BRIGGS & STRATON engine for upper
breakdown pressure (breakdown at 7 – 8 atm pressure). Therefore, we have used a controller unit, based on
AT89C2051 microcontroller, to ensure synchronisation of the ignition with the maximum pressure into the cylinder
obtained during the compression time. In these conditions, we have verified that the purposed double sparks system
provides bigger energy than with the classic system.
This research work has been developed in the frame of an Integrated Actions Programme (PAI Brancusi no.
7/2003 – 2004) and has been supported by the Francophone Universitary Agency (Post-Doc bourse accorded by
1 B.Hnatiuc, S.Pellerin, E.Hnatiuc and J.Chapelle, Czechoslovak Journal of Physics (2006) [Soumis]
2 Activity report from the division of Combustion Physics 1999 – 2000, Department of Physics at Lund Institute of
Technology, Lund, Sweden
3 N.Gherardi, V.Puech, “DBD filamentaires et luminescentes”, Atelier “Traitement de surface par plasma à pression
atmosphérique”, 1-2 decembre 2003, Aspen , France
4 B.Hnatiuc, S.Pellerin, M.Brassart, J.Chapelle, “Système d’allumage par plasma pour moteur d’automobile”, FRAPOL’04,
16–19 juin 2003, Bourges, France
5 B.Hnatiuc & S.Pellerin, “Electrical modelling and diagnostic of a luminescent discharge at high pressure”, 9th IC
Optimization of Electrical and Electronic , Brasov (Roumanie) – 20/22 mai 2004
6 B.Hnatiuc, “Système d’allumage par plasma pour moteur d’automobile”, Stage Postdoctotal (AUF) –01/10/2003-31/07/2004,
LASEP-Université d’Orléans, France

Опубликовано в рубрике Documents