Modeling the Effects of the Ignition System on the CCV of Ultra-Lean SI Engines using a CFD RANS Approach

Abstract

Cycle-To-Cycle Variability (CCV) must be properly considered when modeling the ignition process in SI engines operating with ultra-lean mixtures. In this work, a strategy to model the impact of the ignition type on the CCV was developed using the RANS approach for turbulence modelling, performing multi-cycle simulations for the powercycle only. The spark-discharge was modelled through a set of Lagrangian particles, introduced along the sparkgap and interacting with the surrounding Eulerian gas flow. Then, at each discharge event, the velocity of each particle was modified with a zero-divergence perturbation of the velocity field with respect to average conditions. Finally, the particles velocity was evolved according to the Simplified Langevin Model (SLM), which keeps memory of the initial perturbation and applies a Wiener process to simulate the stochastic interaction of each channel particle with the surrounding gas flow. The particles diameter was also evolved, according to both the energy transfer from the electrodes and the local stretched laminar flame speed. The proposed methodology was assessed against experimental measurements from a pent-roof high-tumble single-cylinder SI engine, equipped with an electrical circuit able to provide both standard and enhanced discharge events. Promising results were achieved in terms of predicted IMEP and its CoV.

A CFD ignition model to predict average-cycle combustion in SI engines with extreme EGR levels

“…Equations for mass ( m p ) and radius ( r p ) variation are solved for each particle according to Refs. 23,27…”

A methodology to initialize tumble flow fields for fast 3D-CFD simulations of pent-roof SI engines

“…In this framework, computational fluid dynamics (CFD) is of strong interest for engine research, as it allows extensive parametric investigations under a wide set of operating conditions. In this context, strong effort is put into developing fast CFD methodologies for modeling internal combustion engines (ICEs), to speed up the optimization process in industrial research [9,10].…”

Numerical investigation of the spark discharge process in a crossflow

“…In the past few decades, many SI models [10][11][12][13][14][15][16][17][18][19][20][21] have been developed for engine CFD simulations, and the level of modeling detail varies between the models. Spark channel is an electrically conductive path that deposits electrical energy into surrounding gas via Joule heating.…”

“…Dahms et al [13] further took the turbulent fluctuation effect into account for modeling the wrinkled spark channel in highly turbulent flows. Sforza et al [21] also introduced velocity perturbation using a simplified stochastic model to reproduce the cycle-tocycle variability of spark channel evolution. On the other hand, Pyszczek et al [18] employed a numerical model that calculates the motion of a three-dimensional (3D) curve to confine the shape of the overall spark channel to be smooth while keeping constant distances between particles.…”

“…By solving an electrical circuit model, Duclos and Colin [10] monitored the voltage between the electrodes and determined a restrike when it exceeds the breakdown voltage. Other models [14,15,18,21] adopted the same modeling approach. Without a circuit model, Dahms et al [13] imposed a limit on the channel length by elongation and invoked restrike when it exceeded 10 mm.…”