Toward quantification of the information content in flow variables under cycle-to-cycle variations

Cycle-to-cycle variations (CCV) pose significant challenges to the performance and reliability of internal combustion engines (ICE). This study investigates the applicability of information entropy as a diagnostic tool to quantify and characterize CCV in ICE subject to exhaust gas recirculation (EGR). Particle image velocimetry (PIV) and flame imaging were used to measure the flow field and flame behavior. The concept of information entropy is related to flow and flame evolution by calculating the Shannon entropy of the turbulent kinetic energy and flow velocity components to assess their sensitivity to CCV. Shannon entropy was applied to the flow velocity components and the turbulent kinetic energy on the tumble plane, where the horizontal component was identified to correlate with high-speed (HC) and low-speed (LC) combustion cycles both, with and without EGR. The turbulent kinetic energy was found to be major driver of HC cycles without EGR, however, it plays a subordinate role with EGR. The conditional Shannon entropy was then applied to assess local fluctuation of the turbulent kinetic energy during ignition from cycle-to-cycle under the influence of the direction of the horizontal velocity component and the presence or absence of the flame, revealing the areas where turbulent kinetic energy fluctuates strongly between different cycles. The results have potential for both computational and experimental studies of ICE, since the use of information entropy as a diagnostic tool allows for early prediction of HC and LC cycles.