CFD Simulation of Methane Jet Burner L. Perković ∗,1 , M. Baburić 1 , P. Priesching 2 , N. Duić 1 1 Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb, Croatia 2 AVL-AST, Graz, Austria Abstract Methane jet flame simulations have been performed in order to evaluate conserved scalar chemistry (CSC) approach in modelling combustion process inside cylindrical burner. Two combustion regimes were included: non-premixed and premixed. For both of them it was necessary to generate look-up tables in pre-processing stage. For non- premixed regime two combustion controlling parameter approaches were used: scalar dissipation rate and normalized reaction progress variable (nRPV) approach. For premixed regime only nRPV approach was used. As an addition to CSC module, discrete transfer radiation module (DTRM) radiation model was also included. Simulation results were compared with experimental data. ∗ Corresponding author: luka.perkovic@fsb.hr Proceedings of the European Combustion Meeting 2009 Introduction Conserved scalar chemistry model is based on pre- tabulated chemistry (look-up tables). For fluid flow solver, AVL FIRE (v2008) was used. Look-up tables were generated in separate applications. Constraining values for interpolation from look-up tables were variables resolved from computational domain. Those are tracking scalars: mixture fraction mean (Z), mixture fraction variance (Z var ) and RPV (in case when tables were reparametrised with RPV). The non-premixed tables for diffusion flames were generated with stationary laminar flamelet model (SLFM) [6] in CSC solver [8]. These tables were then re-parametrised with properly selected (normalised) reacton progress variable (nRPV) in order to generate SLFM-RPV tables. This is known as presumed conditional moment (PCM) closure approach. Tables of freely propagating adiabatic premixed flames were generated with an adopted 1-D PREMIX solver [9]. These tables have also been re-parametrised with nRPV in PCM procedure and as a result one gets FPI-RPV tables. This is known as flame prolongation of intrisic low-dimensional manifold (FPI) [12] approach. In SLFM combustion controlling parameter is scalar dissipation rate which is resolved from flow field turbulence parameters k and ε. In both SLFM-RPV and FPI-RPV controlling parameter is RPV, which in this case was sum of CO2, CO and H2O, as suggested from [5]. CHEMKIN II libraries [10] were used for chemical kinetics and species properties evaluations. K-ε turbulence model was used and turbulence/chemistry interaction was accomplished via the presumed β-PDF. Interpolation from β-PDF tables was done via tracking (active) scalars resolved from computational domain: Z and Z var . In case where RPV approach is used, nRPV is used as a parameter, otherwise scalar dissipation rate was used. In order to capture radiative heat transport discrete transfer radiation module (DTRM) [14,15] was used with wighted sum of grey gases model [16]. Numerical solution was compared with experimental results from published by Sandia National Laboratories. Comparison criteria was temperature and main species mass fractions in centerline axial and two radial directions. Fuel to air ratio is in this work described by mixture fraction Z, defined as: F F O m Z m m = + () where m F is mass flow rate on fuel, and m O on oxidizer side. Experimental configuration Experimental configuration consisted from a central fuel jet, a pilot jet and a co-flow air jet in a concentric annular arrangement (Figure 1). Figure 1. Methane flame configuration The fuel was composed of 25% methane (CH4) and 75% air by volume and had temperature 294 K. The surrounding pilot had an equivalent equilibrium composition to methane/air at Z=0.27, with the temperature 1880 K. The co-flowing air was held at 291 K. The flame operated at Re=22400 with a small degree of local extinction (Sandia flame D). The bulk velocities were 49.6 m/s for the fuel, 11.4 m/s for the pilot and 0.9 m/s for the air. Velocity profiles are presented in Figure 2.