Development of an integrated reduced fuel oxidation and soot precursor formation mechanism for CFD simulations of diesel combustion

In this reported work, a reduced chemical mechanism of surrogate diesel fuel was developed for diesel engine simulations. The aim here was to employ an appropriate reduction scheme to create a compact yet sufficiently comprehensive model which can accurately account for in-cylinder diesel combustion and soot precursor formation processes.

The Combustion Engine Research Center (CERC) mechanism of Chalmers University of Technology was used as the base mechanism since this is the most extensively validated and applied mechanism. The reduction scheme involved firstly identifying and then eliminating unimportant species/reactions in the ignition and soot precursor formation processes through the computed production rates and temperature sensitivity coefficients using CHEMKIN-PRO software.

Subsequently, reactions were assimilated based on the quasi-steady state assumption (QSSA). The final reduced mechanism, which consists of 109 elementary reactions with 44 species, was first validated under 48 shock tube conditions. Ignition delay (ID) periods predicted by the reduced and base CERC mechanisms were found to be in good agreement, although percentage errors of up to 20% were observed.

Further validation was performed by incorporating the reduced mechanism into multi-dimensional CFD code, ANSYS FLUENT through a plug-in chemistry solver, CHEMKIN-CFD. Here, simulation results of combustion characteristics and soot production profiles were compared against data from an experimental study on a heavy-duty, direct injection diesel engine.

Simulated peak pressures in cases with short and long ID periods were identical to those recorded from experiments and the maximum ID offset was maintained to within 1 crank angle degree. Spatial and temporal evolutions of in-cylinder soot were also captured successfully in both cases. Significant qualitative relationship between input parameters and soot evolution was elucidated from this study.

The implementation of the reduced mechanism has achieved a 38% reduction in computational runtime when compared with that of the base mechanism.