Application of the optical flow method for the experimental analysis of turbulent flame propagation in a transparent engine

This paper illustrates the analysis conducted on high-definition, high sampling rate image sequences collected in experiments with a single spark ignition optically accessible engine. Images are first processed to identify the reaction front, and then analyzed by an optical flow estimation technique. The results show that each velocity component of the estimated flow field has an ECDF very similar to the CDF of a Gaussian distribution, whereas the velocity magnitude has an ECDF well fitted by a Rayleigh probability distribution. The proposed non-intrusive method provides a fast statistical characterization of the flame propagation phenomenon in the engine combustion chamber.