Tuning the Parameters of ECFM-3Z Combustion Model for CFD 3D Simulation of a Two Valve Engine fueled with Ethanol 2016-36-0383

This paper presents the adjustment of an extended coherent flame model - three zones (ECFM-3Z) by the analysis of its variables α and β for application in a 3D computational fluid dynamics (CFD) model of a two valve engine fueled with ethanol. The engine is aspirated and uses direct injection of ethanol as a single injection. High frequency pressure gauges are used at the cylinder, intake and exhaust manifolds in order to perform a three pressure analysis (TPA) and this data were used as boundary conditions for the CFD model. In the tests the engine speed varied in the range of 1500 to 4000 rpm, and throttle position in the range of 20% to 100% (WOT), although wide open throttle (WOT) was used for adjustment of the combustion model parameters.

The computational model uses the ES-ICE module of PROSTAR for moving mesh generation, implementation of boundary conditions and transient analysis of global variables. The geometry was constructed at SOLID WORKS and represents only the internal volume of one cylinder along with intake and exhaust ports. A small length of intake and exhaust manifolds internal volume is used in order to take into account the pressure acquisition signal on the manifolds as boundary conditions. The ECFM-3Z is a combustion model developed to simulate both, premixed or diffusion combustion, and is recommended for simulation of direct injection combustion at spark ignition engines, as long as there is residual amount of fuel not evaporated in the cylinder. This model requires lots of experimental data for calibration of its empirical variables, for example α and β which have default values of 1.6 and 1.0, respectively.

The results show that the default values of α and β lead the in cylinder pressure peak to twice the experimental values for all of the cases analyzed. The same is observed for the heat release rate and burning speed. The reduction of α has the effect of reducing the in cylinder pressure, and its adjustment to around 0.5 makes the numerical in cylinder pressure as close as possible to the experiment, for β with default value. On the other hand, for reducing the in cylinder pressure and heat release rate one must increase the value of β, and the best value for this parameter is 1.3 with α equal to 0.6 for 2000 rpm.