Experimental and simulative investigation of flame–wall interactions and quenching in spark-ignition engines

Autor: Eberhard Schutting, Dirk Linse, Helmut Eichlseder, Dominik Suckart
Rok vydání: 2016
Předmět:
Zdroj: Automotive and Engine Technology. 2:25-38
ISSN: 2365-5135
2365-5127
Popis: Flame–wall interactions are a crucial aspect concerning the design and optimisation of a spark-ignition engine. To improve the understanding of the phenomenon, flame–wall interactions are investigated by highly resolved wall heat flux measurements. For this purpose, a turbo-charged, direct-injected spark-ignition engine was equipped with eight surface thermocouples and operated at five measuring points to examine the influence of speed, load, charge motion and equivalence ratio on the quenching process. A cycle-resolved analysis is utilized to extract the wall heat fluxes during flame–wall interaction which are subsequently used to calculate the quenching distances, as well as the normalised wall heat fluxes and Peclet numbers. To correctly estimate these values, a simulative methodology based upon a 3D-CFD in-cylinder flow and a 1D flame calculation with detailed chemical kinetics is employed. The 3D-CFD in-cylinder flow was combined with a 3D-FE heat conduction and 3D-CFD coolant flow simulation to properly predict the wall temperatures and thereby the near-wall flow state. The findings show that the quenching distance is proportional to the laminar flame thickness at first order. Hence, the engine load is found to be the main parameter influencing the quenching distance. The analysis of the quenching distances, normalized wall heat fluxes and Peclet numbers reveals that flame–wall interactions in spark-ignition engines exhibit strong similarities to laminar premixed flame–wall interactions. In addition, two correlations for calculating the quenching distance are proposed. These insights provide a deeper understanding of flame–wall interactions in spark-ignition engines and can be used to develop or adapt turbulent combustion models.
Databáze: OpenAIRE