J. Warnatz

Turbulent non-premixed combustion in partially premixed diffusion flamelets with detailed chemistry

The formulation of an extended laminar flamelet model of non-premixed turbulent combustion is developed. The microscopic elements in the turbulent ensemble are taken as laminar diffusion flamelets that are distinguished by the extent of stretching to which they are exposed and, additionally, by the degree to which their reactants are premixed. Extensive laminar flamelet computations are performed using a detailed mechanism for the methane-air system consisting of more than 250 elementary reactions of 39 chemical species. The degree both of flow field stretching and of partial premixing of reactants is varied systematically. The agreement of the numerical results with experimental data available from the literature is found to be excellent. First numerical simulations of turbulent flames on the basis of the extended flamelet model are performed. The results show that partial premixing of laminar diffusion flamelets is essential for the prediction of turbulent flame structures.

Modeling of Nitrogen and Oxygen Recombination on Partial Catalytic Surfaces

In connection with recombination coefficients derived from experimental data described in the literature, a reaction scheme including detailed rate expressions for O and N atom recombination on the surface of re-entry vehicles is established, consisting of elementary reaction steps. To validate the reaction mechanism derived, surface chemistry and fluid mechanical processes are coupled assuming a one-dimensional stagnation flow field. A quantitative agreement is achieved between recombination coefficients resulting from the numerical computations and those calculated from experiments. The temperature dependence of the recombination coefficient is explained by elementary reaction steps. Furthermore, the reaction scheme established is implemented in a two-dimensional Navier-Stokes code computing the re-entry flow around a simple geometry to show the importance ofa detailed modeling of surface reactions.

Ignition processes in carbon-monoxide-hydrogen-oxygen mixtures

Ignition processes in the carbon monoxide-hydrogen-oxygen system are simulated by solving the corresponding conservation equations (i.e. conservation of mass, energy, momentum and species mass) for one-dimensional geometries using a detailed reaction mechanism and a multi-species transport model. An additional source term in the energy conservation equation allows the treatment of induced ignition, and a realistic model for the destruction of reactive species at the vessel surface is used to treat auto-ignitions in static reactors. Spatial discretization using finite differences and an adaptive grid point system leads to a differential/algebraic equation system which is solved numerically by extrapolation or backward differencing codes. Minimum ignition energies are calculated for various mixture compositions and radii of the external energy source. Ignition limits are computed, and a sensitivity analysis shows the rate-limiting reactions.

Experimental and numerical investigation of the ignition of methane combustion in a platinum-coated honeycomb monolith

This paper represents an experimental and numerical study of the ignition of catalytic combustion of methane in a cylindrically shaped honeycomb monolith coated with platinum. The objective is the achievement of a better understanding of transient processes in catalytic combustion monoliths. In the experiment, cold methane/oxygen/argon mixtures are fed into the monolith, which is placed in a furnace used to heat up the monolith until ignition occurs. The ignition process is monitored by thermocouples and mass spectroscopy. In the numerical study, the time-dependent temperature distribution of the entire catalytic solid structure and the two-dimensional laminar flow fields of the single monolith channels are simulated. The latter predict the gaseous velocity, species concentrations, and temperature based on a boundary-layer approximation. A multistep heterogeneous reaction mechanism is used, and the surface coverage with adsorbed species is calculated as function of the position in the monolith. The heat balance for the solid structure is coupled with the single channel simulations by axial wall temperature profiles, representing the temperature boundary condition in the single channel simulation, and by heat source terms, derived from the gaseous heat convection and chemical heat release in the single channels. The procedure employs the difference in timescales of the temperature variation of the solid, which is on the order of seconds, and of the flow, which is on the order of miliseconds. Experimentally determined and numerically predicted ignition temperatures, as well as time-varying monolith exit temperatures, and fuel conversion during ignition are compared for several CH 4 /O 2 ratios. At the conditions applied, ignition starts at the rear end in the outmost channels.

Simulation of laminar methane-air flames using automatically simplified chemical kinetics

Abstract The method of intrinsic low-dimensional manifolds to simplify chemical kinetics is applied to laminar methane-air flames. The procedure is based on a mathematical analysis of the reaction system. Neither steady state assumptions for some species, nor partial equilibrium for reactions have to be specified explicitly. The only requirements to the scheme are a detailed reaction mechanism and the number of degrees of freedom desired for the simplified scheme. All necessary information on the ther-mechanical state (species concentrations, temperature, density, etc.) is then given as function of a small number of reaction progress variables, associated with the degrees of freedom. Therefore, less equations (for two or three reaction progress variables instead of for 34 species in case of CH4) have to be solved, thus drastically reducing the computational effort compared to calculations using detailed chemistry. Subsequent use of a tabulation procedure, where all information is stored, guarantees an eff...

Kinetic model of an oxygen‐free methane conversion on a platinum catalyst

In many metal‐catalyzed conversion processes of hydrocarbons at atmospheric pressure a carbonaceous overlayer quickly builds up at the catalyst covering nearly the whole surface. However, the metal still remains catalytically active. Several models have been proposed over the years to explain the crucial role of the carbonaceous overlayer during the conversion of hydrocarbons. The model presented here contemplates adsorbate effects, which means that surface carbon modifies the dehydrogenation activity of Pt. A hydrocarbon reaction mechanism on platinum, including C1 and C2 species, is established. The mechanism is based on elementary reactions offering the opportunity of using the same mechanism for a wide range of applications. It is also applied to extended simulations of higher pressures and smaller flow velocities revealing increased C2H6 yields under these conditions.

The Hydrodynamic Structure of a Methane-Air Tulip Flame

Abstract The flame propagation and structure of a stoichiometric methane-air flame in a closed vessel is investigated. For the parameter range examined a tulip flame is formed. A qualitative agreement with experiments is obtained for different length to width ratios of the vessel. The results indicate an enhanced tulip effect with increasing vessel length.

Simulation of homoepitaxial growth on the diamond (100) surface using detailed reaction mechanisms

Abstract One-dimensional reactive-flow simulations of a hot-filament CVD-system including detailed surface reaction mechanisms for homoepitaxial diamond growth are carried out. A growth model for the diamond (100) surface based on elementary chemical reactions steps is introduced. This surface reaction scheme includes the incorporation of the CH 3 radical in the diamond lattice. Homoepitaxial growth on the (100) surface is modelled for a wide range of experimental reactor parameters. The experimental growth rates are compared with simulations for two different surface reaction schemes. It is found that the scheme based on growth at monoatomic steps on the reconstructed (100) surface is more realistic. It shows qualitative agreement with experimental data, whereas the more simple mechanism for an unrealistic unreconstructed (100) surface cannot explain the surface temperature dependence of the growth rate correctly.

DETAILED KINETIC MODELING OF SOOT FORMATION DURING SHOCK-TUBE PYROLYSIS OF C6H6: DIRECT COMPARISON WITH THE RESULTS OF TIME-RESOLVED LASER-INDUCED INCANDESCENCE (LII) AND CW-LASER EXTINCTION MEASUREMENTS

The results of calculations of the main parameters of the soot formation process (τ, k f, SY, and r m) carried out with the use of the detailed kinetic model of soot formation are compared with the experimental measurements of these parameters by the continuous-wave (CW)-laser extinction technique and by the time-resolved laser-induced incandescence (LII) method during C6H6 pyrolysis behind reflected shock waves. The detailed kinetic model of soot formation that is developed incorporates the gas-phase mechanisms of acetylene pyrolysis and the mechanisms of formation of polyaromatic hydrocarbons, polyyne molecules, and pure carbon clusters. It combines the H abstraction/C2H2 addition and polyyne pathways of the soot formation process. The formation, growth, and coagulation of soot precursors and soot particles are described within the framework of the discrete Galerkin technique based on an error-controlled expansion of the size distribution function of heterogeneous species into the orthogonal polynomials of a discrete variable (in particular, the number of monomers in the heterogeneous particle) that makes it possible to preserve a discrete character of any elementary transformations of heterogeneous particles and to describe them as elementary chemical reactions for the heterogeneous particles of all sizes. The comparison of the calculations with the experimental measurements of the induction time τ, observable rate of soot particle growth, k f, and soot yield SY by the CW-laser extinction method in the pyrolysis of benzene/argon mixtures in shock-tube experiments clearly demonstrates that the coincidence is quantitatively good for all the main parameters of soot formation. A particular difference between the values of the mean soot particle radius r m experimentally measured by the time-resolved LII technique and calculated with the help of the detailed kinetic model is observed at the low and high temperatures. The results presented demonstrate the current level of the predictive capabilities of the detailed kinetic model of soot formation and the reliability of the time-resolved LII technique for the quantitative determination of the soot particle sizes.

Experimental and theoretical investigation of CO oxidation on platinum: Bridging the pressure and materials gap

Optical IR-visible sum-frequency generation (SFG) surface vibrational spectroscopy was applied for in situ detection of chemisorbed CO during heterogeneous CO oxidation on a polycrystalline platinum catalyst. The substrate temperature was between 300 and 700 K at a total pressure of 20 mbar. Experiments were carried out under laminar flow conditions in a well-defined stagnation point flow geometry to allow for a detailed comparison with numerical reactive flow simulations. These were carried out to investigate the interaction among surface heterogeneity, catalytic surface reactions and gas-phase processes, and their coupling by molecular transport. To reproduce the experimental results, the model for the Pt foil is based on two different adsorption sites: � 80% A and 20% B sites. The activation energy for desorption of a CO molecule on a clean surface was found to be 183 kJ/mol with a pre-exponential factor of 3 • 10 19 s � 1 on the A sites and 220 kJ/mol with a pre-exponential factor of 5 • 10 21 s � 1 on the B sites. A strong dependency of the desorption energy on the CO coverage was found on the A sites, dropping to 71 kJ/mol on a CO covered surface. These values match the heat of adsorption on Pt(111) and Pt(311) reported by King et al., suggesting structural similarities of these surfaces to a Pt foil. B sites were found to be effectively blocked in the presence of oxygen. But since no significant CO2 production was observed at low temperatures, we conclude that B sites are not the active sites for the CO2 formation. This finding could be explained by the strong bonding of oxygen to those sites.