Simultaneous Application of Optical Spark Plug Probe and Head Gasket Ionization Probe to a Production Engine 930464

The optical spark plug probe and ionization head gasket probe developed at Sandia Laboratories were applied to one cylinder of a production multicylinder automotive gasoline engine. The purpose of this application is to eventually study combustion phenomena leading to high emissions under cold start and cold idle conditions. As a first step in studying cold start combustion and emissions issues, diagnostic instrumentation was simultaneously applied to a production engine under steady state idle, road load and an intermediate load-speed condition. The preliminary application of such instrumentation is the subject of the present paper.

The spark plug probe was redesigned for ease of use in production engines and to provide a more robust design. The two probes were geometrically oriented to obtain radial line-up between the optical windows and ionization probes. Data were taken simultaneously with both probes at the three load-speed conditions mentioned above. In addition, the production cylinder head was modified to obtain a substantially higher tumble swirl level and the same tests repeated. Correlations between flame arrival times at the optical windows and corresponding ionization probes were attempted. Literature indicates that the shape, growth rate and displacement of the early spark kernel may be responsible for cycle to cycle variation. Data presented in this paper do not show any correlation between flame arrival times at geometrically corresponding optical (spark plug) and ionization (head gasket) probes. On a cycle by cycle basis, the shape of the early flame kernel (or its displacement) does not bear any similarity to the shape towards the end of combustion.

Flame arrival times were also determined as a function of load-speed and tumble swirl level. Data show that flame kernel speed and flame arrival times at the head gasket are strongly affected by load-speed conditions. Flame velocities (and arrival times) appear to be unaffected by high tumble swirl under all load-speed conditions tested. No direct measurement of turbulence generated by high tumble swirl was made in this paper.

Cylinder pressure measurements were made simultaneously with measurement of flame arrival times and a heat release rate analysis was done. The 0-2 percent heat release duration was arbitrarily selected for comparison with flame arrival times at the spark plug windows. The 0-90 percent heat release duration was likewise arbitrarily selected for comparison with flame arrival times at the head gasket ionization probes. The comparison shows surprisingly good correlation. It also confirms that high tumble swirl head did not have a measurable effect on heat release durations, just as it did not have a measurable effect on flame arrival times.