Application of a “Presumed p.d.f.” model of turbulent combustion to reciprocating engines

Abstract

This paper reports the results of an attempt to apply a “presumed p.d.f.” model of turbulent combustion using a computer code for reciprocating engines. The first step was to introduce in the CONCHAS code a model in which turbulence controls the combustion, the chemistry assumed to be very fast. We have shown that a mixing length can be used but only if o i) the turbulence integral time is related to the curl of the velocity and not to the deformation rate, ii) the geometrical shape of the piston and the engine is such that sufficiently large velocity gradients are generated. In the second step, a model based on a “presumed shape for the p.d.f.” on the fuel mass fraction, which assumes a single reaction but with a finite rate, was introduced into the numerical code. The modeling constants were determined in such a way that very realistic results were obtained for a moderate range of conditions and geometries. However, this model does not solve the problem: first, the turbulence model and the chemical model can be improved; in addition, a basic discussion can show that the temperature fluctuations have not here been linked with the species fluctuations and the presumed p.d.f. model must be further improved in order to include this feature. Nervertheless, even at this stage of the study, the model appears realistic enough, so that it can already be used by engines manufacturers for a better design.