Le, M., Furui, T., Nishiyama, A., and Ikeda, Y., "Application of High-Speed PIV Diagnostics for Simultaneous Investigation of Flow Field and Spark Ignited Flame inside an Optical SI Engine," SAE Int. J. Engines 10(3):2017, doi:10.4271/2017-01-0656.
High speed, time resolved Particle Image Velocimetry (PIV) diagnostics was applied to an optical SI engine to study the interactions between in-cylinder flow field and flame development. Optimisation and certain adaptations have been made to the diagnostic setup to enable time-resolved, simultaneous measurements of both PIV data and flame tomography imaging from the same original captured image set. In this particular study, interactions between flow and flame during lean-burn operating conditions at various tumble strength have been investigated and compared to a standard stoichiometric operation. Diagnostics were performed for both the vertical plane (x-y) and the horizontal plane (r-⊖) of the combustion chamber with a particular focus in the pent-roof area. Some major differences in the tumble flow-field prior to ignition has been observed between the lean and stoichiometric conditions. Moreover, lean flames show a high degree of convolution and distortion during the early growth period but start to round off in later development stages for higher tumble condition. In terms of flow-flame interactions, the strong stoichiometric flame induces a very high turbulence energy for a wide region of unburned charge all around its flame front. While the lean flame also induces high turbulence in the unburned region near its flame front, the size and scale of these regions are very small in comparison, and are limited to only certain sections of its flame front. However, in the horizontal plane, lean flames under higher tumble condition induces relative high turbulence energy for a wider region all around its flame front, similar to the stoichiometric flame case. These could be the explanation for lean-burn at higher tumble condition having improved CA50 timing and lower combustion variations. The captured data is also useful for engine development efforts to utilise the flow-flame interaction for more stable lean combustion.