A Quasi-Dimensional Charge Motion and Turbulence Model for Diesel Engines with a Fully Variable Valve Train 2018-01-0165
With the increasingly strict emission regulations and economic demands, variable valve trains are gaining in importance in Diesel engines. A valve control strategy has a great impact on the in-cylinder charge motions, turbulence level, thus also on the combustion and emission formation. In order to predict in-cylinder charge motions and turbulence properties for a working process calculation, a zero−/quasi-dimensional flow model is developed for the Diesel engines with a fully variable valve train. For the purpose of better understanding the in-cylinder flow phenomena, detailed 3D CFD simulations of intake and compression strokes are performed at different operating conditions with various piston configurations.
In the course of model development, global in-cylinder charge motions are assigned to idealized flow fields. Among them, swirl flow is characterized by an engine swirl number that is determined by both developments of the swirl angular momentum and the moment of inertia. The generation of swirl angular momentum during intake is estimated from the intake mass flow and instantaneous stationary swirl number. The latter is obtained from virtual flow bench simulations in consideration of valve phasing. When modeling swirl losses during compression and expansion, the effects of wall friction, turbulent conversion as well as piston motion are taken into account. Furthermore, a sub-model describing the flows induced by piston motion including axial and squish flows is set up. In conjunction with the charge motion model, a quasi-dimensional turbulence model is developed based on the k-ε turbulence model. Turbulence production rate and dissipation are determined through sub-models. Inflow turbulence is modeled as local shear in the cylinder entrance area. TKE out of axial flows is transferred latish from the axial kinetic energy. Besides, turbulence productions from squish and swirl flows are approximated using idealized flow fields. Dissipation is estimated in a zero-dimensional sub-model by means of a newly developed turbulence length scale, which takes account of all the influences of in-cylinder flow, inflow and back squish flow.
The results from the performed validations demonstrate that the proposed flow model accurately predicts the temporal change of in-cylinder flow quantities and also responds correctly to the variations of piston configuration as well as operating conditions such as engine speed, charging pressure and valve actuation. Further application of the model in other Diesel engines is feasible by tuning certain model parameters.