Development of integrated chemical kinetic mechanism reduction scheme for diesel and biodiesel fuel surrogates for multi-dimensional CFD applications

This thesis describes the research undertaken to formulate a systematic chemical kinetic mechanism reduction scheme to generate compact yet comprehensive chemical kinetic models for diesel and biodiesel fuels, for multi-dimensional Computational Fluid Dynamics (CFD) applications. The integrated mechanism reduction scheme was formulated based on the appraisal of various existing mechanism reduction techniques. It consists of five stages including Directed Relation Graph (DRG) with Error Propagation method using Dijkstra’s algorithm, isomer lumping, reaction path analysis, DRG method and adjustment of reaction rate constants. Consequently, a single-component diesel surrogate fuel model with only 79 species (i.e. n-hexadecane (HXNv2)) and a multi-component biodiesel surrogate fuel model (i.e. methyl decanoate/methyl-9-decenoate/n-heptane (MCBSv2)) with only 80 species were successfully derived from their respective detailed mechanisms, which contain thousands of species and elementary reactions. Here, both auto-ignition and jet-stirred reactor (JSR) conditions were applied as the data source for mechanism reduction. An overall 97 % reduction in mechanism size in terms of total number of species as well as an average 97 % reduction in computational runtime in zero-dimensional (0-D) chemical kinetic simulations was achieved. Both HXNv2 and MCBSv2 were also comprehensively validated in 0-D simulations in terms of ignition delay (ID) timings and species concentration profiles. Good agreement between the predictions and measurements was obtained throughout the test conditions.\ud \ud Subsequently, HXNv2 and MCBSv2 were integrated into the OpenFOAM-2.0.x solver to simulate spray combustion in a constant volume combustion chamber. The simulation results were validated against the experimental data of no.2 Diesel Fuel (D2) for diesel combustion and Soy Methyl Ester for biodiesel combustion. It was found that MCBSv2 was able to capture the combustion and soot formation events reasonably well. However, further refinement on HXNv2 was essential to improve the complex soot formation predictions. Fuel blending was then suggested to match the diesel fuel kinetics and compositions. As a result, two different versions of multi-component diesel surrogate fuel models were produced in the form of MCDS1 (HXNv2 + 2,2,4,4,6,8,8-heptamethylnonane (HMN)) and MCDS2 (HXNv2 + HMN + toluene + cyclohexane). All the fuel constituent reduced mechanisms and the integrated mechanisms were extensively validated in 0-D simulations under a wide range of shock tube and JSR conditions. Successively, the fidelity of the multi-component diesel surrogate fuel models was evaluated in two-dimensional spray combustion simulations. The computations were compared with the experimental data of D2 fuel. MCDS1 was found to be useful for simulations with less aromatic chemistry effects. In contrast, MCDS2 was a more appropriate surrogate model for fuels with aromatics and cyclo-paraffinic contents. Following that, fidelity of MCDS2 and MCBSv2 was further assessed in three-dimensional internal combustion engine simulations. The performance of the surrogate models was compared under the same operating conditions in a light-duty, direct injection diesel engine. The computed peak pressure and heat-release rate for biodiesel combustion were lower than diesel owing to the advanced ignition timing. The soot formation of biodiesel was also found to be 1.4 times lower than diesel due to oxygenated effects. Overall, the integrated reduction scheme proves to be an attractive approach for large-scale mechanism reduction to reduce the computational time-cost as well as to expedite multi-dimensional CFD computations