Validation of a les Spark-Ignition Model (GLIM) for Highly-Diluted Mixtures in a Closed Volume Combustion Vessel

The establishment of highly-diluted combustion strategies is one of the major challenges that the next generation of sustainable internal combustion engines must face. The desirable use of high EGR rates and of lean mixtures clashes with the tolerable combustion stability. To this aim, the development of numerical models able to reproduce the degree of combustion variability is crucial to allow the virtual exploration and optimization of a wide number of innovative combustion strategies. In this study ignition experiments using a conventional coil system are carried out in a closed volume combustion vessel with side-oriented flow generated by a speed-controlled fan. Acquisitions for four combinations of premixed propane/air mixture quality (φ=0.9,1.2), dilution rate (20%-30%) and lateral flow velocity (1-5 m/s) are used to assess the modelling capabilities of a newly developed spark-ignition model for large-eddy simulation (GLIM, GruMo-UniMORE LES Ignition Model). The model accounts for all the main physical phenomena governing flame kernel growth, including electric circuit and over-adiabatic thermal expansion, which are included in the LES formalism of ECFM-LES combustion model. In the first part of the study the agreement of the simulation results against measurements is carried out to assess a validated non-reacting condition and converged flow statistics. Then, combustion simulations are carried out, and ignition events are repeated at random timings to replicate flow variability at ignition. Finally, optical comparison is carried out for simulated enflamed volume against high-frequency Schlieren images for all the cases, and measurements of flame radius growth are presented. The agreement of the quantitative flame measurements, as well as the qualitative resemblance of flame development, indicate the use of the presented GLIM ignition model as a valuable model for multi-cycle engine simulations, with particular relevance on unstable and CCV-affected conditions