Auto ignition tabulation of N-Heptane in ECFM-3Z Combustion Model M. Ban ∗, 1 , P. Priesching 2 , N. Duic 1 1 Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb, Zagreb, Croatia 2 AVL-AST, Graz, Austria Abstract Recent efforts of implementing autoignition tabulated data to AVL's code "FIRE" including cool flame ignition will be presented in this paper, focused on n-heptane as fuel. Current use of n-heptane in combustion simulations did not include the phenomenon of cool flame ignition, and taking it into account could improve simulations of combustion in CI engines, also with wider application spectrum. The methods and ideas behind implementing the cool flame data into the ECFM-3Z model will be presented, as well as a comparison between the temperature fields calculated with ECFM-3Z model on a simple 100 cell mesh and a standalone application using Chemkin package. ∗ Corresponding author: marko.ban@fsb.hr Proceedings of the European Combustion Meeting 2009 Introduction Recent rapid advances in computer power lead to increased use of computational tools in engine design, significantly reducing the costs of simulations compared to laboratory engine experiments. In addition, there has been significant improvement in the physical sub models used in engine simulations, and the enhanced accuracy has made the use of computational tools advantageous for generating a better understanding of the transient physical and chemical phenomena that occur in internal combustion engines. The goal of this study was to improve the prediction of diesel fuel auto ignition processes using tabulation approach to include the cool flame ignition phenomenon. Flame development, power output and emissions formation are determined by the process of auto ignition in diesel IC engines and is dependent on chemical and physical processes. The first kind of processes is pre- combustion reactions of the fuel with air and residual gases, high temperature combustion and emissions formation. The main physical processes include atomization of liquid fuel, evaporation of fuel droplets and turbulent mixing of vapor with air. Rather than trying to simulate the complex behavior of diesel fuel itself the replacement fuel of choice is n-heptane due to its cetane number of approximately ~56, which is similar to that of ordinary diesel fuel. Current diesel auto ignition model included in AVL code "FIRE" used a tabulated data acquired by running SENKIN calculations varying following initial parameters: temperature, pressure, air excess ratio and recirculating exhaust gas mass fraction. These values were used to simulate the exact ignition moment using a precursor variable in extended coherent flame combustion model. However, existing data provides only the main ignition delays, which is sometimes not accurate enough e.g. when running simulations in a low temperature region without taking cool flame phenomenon into account. Chemistry Background And Numerical Approach When studying the complex chemical mechanisms, it is possible to get a comprehensive insight of the chemical kinetics behind the phenomena of auto- ignition. N-heptane skeletal mechanisms (that include the main species and reactions) consist in general of 20- 80 species with less than 250 reactions (Rente et al., 2001, Tanaka et al., 2003, Liu et al., 2004). These can further be simplified to 4-40 steps, but this approach (done by mathematical transformations) can cause the loss of physical meaning of the individual species (Peters et al., 2002). Initial tabulations for n-heptane were performed using small Golovitchev mechanism (Rente et al., 2001), but since the ignition delay acquired proved to be under predicted on the whole domain, it was rather used to determine the initial data simplifications to perform the tabulation using more complex mechanism. The detailed n-heptane mechanism (of Curran et al., 1998) is intended to cover the entire range of conditions from low-temperature (600-900 K) pyrolysis and oxidation to high-temperature combustion. Several methods are used to reduce the chemical mechanisms to the size (skeletal or reduced models) appropriate for reasonable computation, based on sensitivity analysis, and others (the Quasi-Steady-State Assumption (QSSA), the Intrinsic Low-Dimensional Manifold (ILDM) approach or the Computational Singular Perturbation method (CSP)) (Valorani et al., 2007) . Also, one could base the survey on whether the mechanism simplification method is based on reduction of reactions (Bhattacharjee et. al, 2003) or reduction of species (Lu, 2005 and Pepiot, 2005.). Recent studies show that, using auto-ignition delay as an optimization criteria, the Curran model could be reduced to 170-180 species (Najm et al., 2006), and some show improvement using even more reduced mechanisms (67 species and 265 reactions, Hewson 1997). The chemical mechanisms used were: • n-heptane: