Different Velocity Data Analysis for Flows Near a Spark Plug in the Combustion Chamber of a Spark Ignition Engine

Flow near a spark plug is important for early flame kernel development (EFKD) and combustion efficiency. Velocity data have been measured by a laser Doppler velocimetry (LDV) for three different positions near a spark plug within a ported single cylinder optical spark ignition (SI) engine with a heart-shaped combustion chamber and a compression ratio of 8.9. LDV measurements have been performed under the wide-open motored conditions with an engine speed of 1,000 rpm conditions and maximum data collection rates of 22 kHz per channel.

This work examines the mean and turbulence flow fields as interpreted through ensemble, cyclic, discrete wavelet transformation (DWT) analysis and the energy cascade as analyzed through continuous wavelet transformation (CWT) for flows near a spark plug. The results in this paper show that the frequency of the first level DWT can be used as a cutoff frequency for filtering because the first level DWT satisfies within 5 % relative error the condition that the ensemble average of the individual cycle mean should be same as the conventional ensemble-mean velocity, and it is near the maximum frequency in frequency spectrum of the ensemble averaged velocity as Liou et al. approached [10]. Wavelet analysis is appropriate for analyzing the flow fields in the combustion engine because it gives information about the transient events in a time and a scale plane. Large-scale high energy (defined below) found with CWT corresponds well with the peak locations of the mean velocity