Three-dimensional Modeling with Monte Carlo–probability Density Function Methods and Laser Diagnostics of the Combustion in a Two-stroke Engine

A Monte Carlo–probability density function (pdf) method based on a Lagrangian particle technique to calculate the transported pdf of the thermochemical (scalar) properties has been implemented in the finite volume code KIVA-3 to model three-dimensional premixed transient combustion in a two-stroke engine. The main advantage of the pdf method is that chemical reactions in turbulent flows can be treated exactly, without model assumptions. Turbulent transport is handled by the mixing model of Curl. The pdf model and the computational fluid dynamics (CFD) code are coupled by a hybrid solution scheme. For each time step, the flowfield is interpolated to match the particle system. The density calculated from the thermochemical variables is then returned to the CFD code. The thermochemical system is simplified by a reaction scheme based on the intrinsic low-dimensional manifold (ILDM) method using a table with a few reaction progress variables. The aim of this work is to show that the Monte Carlo technique coupled to an ILDM method can model technical systems in three spatial dimensions, including the prediction of flame front species such as radicals. The results are compared to measurements of flame propagation using two-dimensional tracer laser-induced fluorescence imaging in an optically accessible two-stroke engine.