G. De Soete

Overall reaction rates of NO and N2 formation from fuel nitrogen

From measurements carried out on flat premixed hydrocarbon/oxygen argon (or helium) flames, into which small amounts of ammonia, or cyanogen are added, overall reaction rates of formation of NO and N2 are determined. From similar measurements effected on nitrogen-diluted ethylene/oxygen flames, an overall rate of prompt NO formation is obtained. The discussion of these rate constants indicates that the relative importance of HCN molecules as intermediates in the fuel NO mechanism increases according to the following sequence of primary fuel nitrogen compounds: ammonia, cyanogen and molecular nitrogen; this last is found to behave like a true fuel nitrogen compound in the early flame stages. Experimental values of the total yield of nitric oxide obtained from the added nitrogen compounds have been determined; they are found to be in good agreement with yields calculated by numerical integration of the empirical overall reaction rates of NO and N2 formation, showing almost the same dependence of the NO yield on temperature, initial fuel nitrogen concentration and oxygen concentration.

Heterogeneous N2O and NO formation from bound nitrogen atoms during coal char combustion

Using a fixed bed, batch combustion method, heterogeneous kinetics of NO and N 2 O formation from char-nitrogen have been investigated on highly devolatilized coal chars, in the temperature range between 500 and 1300 Kelvin. Using a conventional steady state technique (adsorption/desorption equilibrium conditions), the heterogeneous NO and N 2 O reduction on bound carbon atoms, have been determined. Both N 2 O and NO are mainly obtained as desorption products from oxidized (−CN) sites; only a few percent of the released nitrous oxide is produced by reactions involving hydrocyanic acid as a reactant. The fractions of char-nitrogen converted into NO and N 2 O during char oxidation, are roughly proportional to the degree of char burnout, showing the charabound carbon and nitrogen atoms to be oxidized in proportions almost identical to their respective concentrations in the char. Due to simultaneously occuring heterogeneous reduction of NO and N 2 O onbound carbon atoms, the final fraction of char-nitrogen converted into NO and N 2 O, is smaller than one (0.3 to 0.7, depending on char type, oxygen concentration and particle size). The rate constants of the different adsorption and desorption reactions have been determined in the transient state, created by the sudden suppression of the oxygen in the gaseous reactants. The small values of the oxygen covered sites fractions, as well as the first order with respect to the gaseous reactants, indicate that the overall NO and N 2 O formation is controlled by oxygen adsorption; below 800 K, this adsorption evolves in the kinetic regime; at higher temperature it appears to be controlled by pore diffusivity.

The influence of isotropic turbulence on the critical ignition energy

Theoretical considerations are presented which indicate that the increase of spark-ignition energy under turbulent-flow conditions is mainly due to an increase in transport phenomena resulting from turbulent heat diffusivity. In order to account for the relatively small diffusion times involved in ignition phenomena, turbulent diffusivity should be considered as time dependent. For diffusion times (t) much smaller than the characteristic lifetimes of the turbulent eddies (defined as time scale/turbulence intensity), the turbulent mass diffusivity Dt becomes equal to u12t/2. This turbulent diffusivity acts on the ignition energy in two ways: o 1. By increasing the flame-front thickness according to the proportionality et/e= (u12t/2Dm+1) it causes a relatively important increase in the minimum critical volume of flammable mixture that must be heated up by the ignition energy; 2. It increases the fraction ξ of the spark energy that is available for ignition, the complementary fraction (1-ξ) of the spark energy being wasted beyond the critical volume. This secondary effect of turbulent diffusivity is rather small, but depends on the spark duration θ and on the energy release function. Calculations of the fraction ξ as a function of θ are presented under both laminar and turbulent conditions. Experimental evidence of the effect of the spark-duration time and the turbulent diffusivity on ignition energy is obtained from work carried out with flammable methane/air and propane/air mixtures for different turbulent-flow conditions.

Stability and propagation of combustion waves in inert porous media

An experimental study was performed on the reverse combustion of premixed methane/oxygen/nitrogen mixtures propagating in sand of calibrated grain sizes. The propagation velocity of the combustion zone relative to the solid was measured and temperature profiles were obtained. For each mixture employed, there was a minimum pore size for which flame was quenched, mainly due to the fact that the probability of chain breaking at the walls begins to exceed the probability of chain branching. Temperature profiles are found to be quite different for the solid phase and the gas phase, due to the heat released in the latter. A general scheme of heat transfer in the preheat zone and the combustion zone makes it possible to predict the propagation velocity and the final temperature of this zone as a function of thermal and kinetic parameters such as heat conductivity of the porous medium, heat-transfer coefficient, heat of combustion, and burning velocity.

The Effect of Aging on Nitrous Oxide N2O Formation by Automotive Three-Way Catalysts

Abstract Three series of Pt Rh three-way catalysts were aged in the exhaust of an engine running on a dynamometer stand for various lengths of time. Their activity was thell measured as a function of temperature at stoichiometry and as a function of the air-fuel equivalence ratio at constant temperature (450°C). A peak of N2O formation was observed close to the catalyst light-off temperature. Since the light-off temperature increases as a function of aging time, a shift in the N2O peak formation was observed towards higher temperatures. The magnitude of the N2O peak decreased simultaneously. At 450°C a minimum of N2O formation was seen close to stoichiometry with two maxima below and above this point. N2O formation was the highest on the rich side and increased progressively as a function of catalyst aging. Another series of eight catalysts was evaluated in parallel on a vehicle on a chassis dynamometer after engine bench aging. They were tested using the ECE driving cycles: ECE 15 urban with cold or hot starting and extraurban with hot starting. The amount of N2O emitted did not vary very significantly with the catalyst type but was seen to increase 2 to 4.5 times after aging, depending on the driving cycle used. This has to be related to the shift in the light-off temperature to higher values during aging, which makes the catalyst work more at temperatures where N2O is formed.

Experimental Analysis of the Initiation and Development of Part-Load Combustions in Spark-Ignition Engines

The low efficiency of spark ignition engines at partial load is mainly due to non-optimum combustion. Investigation of such combustion shows that an initial phase must be distinguished from the development phase. Unacceptable combustions may be of various types such as slow and/or late burning, misfire. An experimental method based on pressure-diagram acquisition and real-time cycle-to-cycle computer processing has been developed to characterize the effects of such combustions on engine-running efficiency and smoothness. The effects of the most significant parameters--ignition, combustion-chamber geometry, mixture turbulence and composition--have been investigated with various industrial combustion chambers, leading to guidelines for optimization. To further the study of engines, fundamental data on flame initiation have been obtained through laboratory experimentation using laser tomography.

Effects of geometrical and aerodynamic induced flame stretch on the propagation of spark fired premixed flames in early stages

During early stages of propagation in the spark ignition engine, flames may be subjected to relatively strong stretch effects tending to rapidly increase their surface but also decreasing the burning velocity. Both geometrical and aerodynamical stretch may arise, the former due to fast growth of flame radius, the latter related to the existence of tangential, flow velocity gradients. An experimental study is presented on the effect of both types of stretch on flame speed and surface growth of free expanding, spark ignited flames of ethylene/air mixtures, based on fast laser shadowgraphy records. Pure geometrical stretch effects are studied in gases at rest. Due to the existence of a spark imposed initial flame speed value, the adjustment of the latter to the value imposed by the relative stretch (dS/S dt=2 dR/R dt) is delayed and causes almost periodic, progressively attenuated fluctuations of the flame speed in early propagation stages, roughly corresponding to the relationship Vp=Vpo[1−α exp (−νt)] [sin2(2πνt)]0.5, where Vp and Vpo are the instantaneous, respectively the steady state values of the flame speed defined as the flame radius growth per unit time. α and ν are characteristics of the flammable mixture. As a general consequence the volumetric combustion rate (dvb/dt, with vb, being the volume of the combustion products) is decreased. Aerodynamical stretch effects are studied on flames ignited in the stagnation point of concentric opposed jet flow. In very early stages however, when flame radius is still small, aerodynamical stretch is always accompanied with the former described geometrical stretch prervailing often in that period. At later stages the flame speed approaches to an asymptotic value controlled by pure aerodynamical stretch and fairly well forecasted by the theoretical expression recently proposed by P. Clavin,1 at least for Lewis numbers close to unity. The effect of this combined geometrical/aerodynamical flame stretch on volumetric combustion rate in early stages may result as well in an increase as in a decrease of this parameter. Local flame extinction by aerodynamical stretch has been observed too.