Wolfgang Leuckel

A Model for Calculating Heat Release in Premixed Turbulent Flames

Abstract A unified reaction model, being valid in all turbulent combustion regimes, has been developed and tested. Based on a thoroughly validated model for the turbulent burning velocity the Kolmogorov, Petruvski and Piskunuv (KPP) theorem was applied, thus leading to the formulation of the mean reaction rate as a function of local turbulence and kinetic parameters in the flow. Numerical calculations, comprising all flame structures (0.2 Da t [25] that the blow-off limits of strongly swirling flames are determined by chemical kinetic limitation of the mean reaction rate. The results demonstrate the high performance of the reaction model and recommend its application also in complex 3-D flows, because of its simplicity and numerical robustness even in conjunction with higher order turbulence models.

Experimental investigation on the stabilization mechanism of jet diffusion flames

Two contrary concepts have been suggested in order to explain the mechanism of flame stabilization based on premixed and diffusion flamelet combustion respectively. To contribute to the understanding of which model represents the true stabilization mechanism, two different natural gas jet diffusion flames, at exit velocities between flame detachment and blow-off, were investigated. The measured profiles of gas composition and velocity around the stabilization zone were used to derive the rate of mixing and fuel burnout. The results show that, for the flames investigated, about forty to fifty percent of the total fuel flow is already mixed at a molecular level upstream of the flame stabilization zone. This mixture then reacts over a very short distance, supporting the concept of premixed combustion in lifted jet diffusion flames.

Turbulent swirling flames: Experimental investigation of the flow field and formation of nitrogen oxide

The NO emissions of a 150 kW natural gas (with and without fuel N) burner were shown to be dependent on the geometry of the gas nozzle and the swirl intensity of the combustion air. Numerous detailed inflame measurements were performed in order to understand the local conditions that influence the formation of NO in highly turbulent swirling flames. It was found, that by controlling the axial fuel gas momentum in combination with the swirl number, mixing of fuel and air can be influenced, thus reducing NO formation. Using fuels with chemically fixed nitrogen requires flames with the main heat release under fuel-rich conditions, leading to type I flames. These flames have the potential to reduce the NO emissions up to 70% compared with premixed flames. To minimize thermal NO formation, high-temperature zones with near-stoichiometric mixtures must be avoided, which may be realized by type II flames. Moreover, swirl turned out to be a strong tool for minimizing thermal NO. Under overall fuel-lean conditions, higher swirl led to faster mixing of fuel and air, thus lowering the temperature level of the flame and reducing thermal NO emissions. The experiments were performed in cooperation with the German Technische Flammen (TECFLAM) joint research program, using a standardized combustion chamber and burner. By applying identical experimental setups and different measurement techniques by the TECFLAM partners, accurate and reliable measuring data were produced on the time-mean and on the turbulent fluctuation levels.

Effects of ignitors and turbulence on dust explosions

The aim of this work is in an attempt to increase the understanding of the acting behaviour of pyrotechnic ignitors and their effects on confined dust explosions. Flame visualization has shown that pyrotechnic ignitors can initiate an explosion by instantaneous jet-like volumetric and/or multipoint ignition. Hence, the rate of pressure rise and also the apparent burning velocity will be increased to some extent, depending on the ignitor energy and the reactivity of the mixtures. The ignitor effect is more important for the early stages of flame propagation and would be more significant in small explosion chambers. Thus, for dust explosion tests with various purposes, use of pyrotechnic ignitors should be made carefully, and the ignitor effect must be accounted for in the data interpretation. Turbulence induced by dust dispersion is a dominant factor in affecting dust explosions. At different ignition delays, however, the turbulence influence will be coupled with that of ignitors. This complicates further the interpretation of explosion data measured under turbulent conditions.

Experimental Investigations on the Dynamics of Pulsated Premixed Axial Jet Flames

Abstract The understanding of the origin of self-excited pressure oscillations in technical combustion systems as well as a successful design of pulse combustors depends on the knowledge of the dynamical behavior of the used flame. In this study the mixture mass flow rate through the burner nozzle was modulated sinusoidally to measure the frequency dependent heat release rate of a premixed turbulent jet flame. The dynamical behaviour of the investigated flames was found to be predominantly a function of a Strouhal number formed with the characteristics of the corresponding steady flames and to be very similar to that of an ideal idle-time model. The discrepancy between this model and the experiments at higher frequencies could be explained by the periodical formation of vortex rings.

Reduction of NOx emission in turbulent combustion by fuel-staging/effects of mixing and stoichiometry in the reduction zone

Abstract A fuel-staged combustion system is optimized for minimum NO x emission by systematic variation of the fuel/air equivalence ratio (E.R.} and the exit momen|um of the reburn-fuel jets in the reduction zone. The fuel is natural gas (NG) doped with 3 Vol% ammonia (NH 3 ). This NG/NH 3 gas mixture is also used as reburn-fuel, in order to comply with conditions found in practical systems. The equivalence ratio in the secondary fuel-rich reduction zone is varied belween 1.04 ⪯ \Gf 2 ⪯ 1.22. Dilution of the reburn-fnel with nitrogen (N 2 ) allows a variation of the momentum of the radially injected reburn-fuel jets at constant stoichiometry to study the influence of mixing in the reduction zone on NO x emission. Concentrations of NO, NO,, HCN, NH3 and of the stable species of the C-H-O-system are measured after the reduction and after the burnout zone, respectively. Detailed infor- mation about the conversion and reduction of N-species in the reduction zone for two dif- ferent mixing conditions is obtained from local concentration measurements of the above mentioned species in the reduction zone. The experiments show that variation of the equivalence ratio in the reduction zone leads to a NO x emission minimum for tb,\Gf 2 = 1.15. For an equivalence ratio \Gf 2 close to the optimum value improved mixing iu the reduction zone results in a reduced N-species concentration (NO, HEN, NH 3 ) after the reduction zone and thus a lower NO x emission of the total system. Finally a simplified model is evaluated to calculate the penetration and the mixing length of the reburn-fuel jets, both parameters being valuable tools to characterize and optimize mixing conditions in the reduction zone,

Flame stabilization and turbulent exchange in strongly swirling natural gas flames

An experimental study has been made in the near field of a variable swirl burner under isothermal and reacting conditions in order to quantify the effect of combustion on the isothermal flow field, and to explain the mechanism of flame stabilization due to internal recirculation zones. A two colour Laser-Doppler velocimeter was used to determine turbulence properties in-/and outside of the recirculation zone. Temperature and species concentrations were measured by conventional measuring techniques. Turbulent exchange coefficients for momentum, matter and heat have been obtained by measurements and by calculations from time mean value distributions, respectively. The results show that, with strong swirl, the intensity of recirculation is reduced by the presence of combustion due to a marked decrease of the effective swirl number. Divergent burner nozzles at the burner exit lead to a radial extension of the reverse flow zone, but do not affect the reverse flow density. Reynolds number similarity of the diffusion flame is confirmed as long as reaction kinetic effects are negligible. On the other hand, blow-off occurs as a result of a reduction of local residence time in the ignition zone when increasing burner load. LDA-measurements exhibit high turbulence intensities and strong turbulent fluxes in the region of the stabilization zone. All turbulent fluxes agree with the gradient law, so that no countergradient diffusion exists under the conditions studied. The presence of combusion leads to a damping of turbulent exchange as compared with the isothermal flow of same swirl intensity.

Evaluation of Advanced Two-Phase Flow and Combustion Models for Predicting Low Emission Combustors

The development of low emission aero engine combustors strongly depends on the availability of accurate and efficient numerical models. The prediction of the interaction between two-phase flow and chemical combustion is one of the major objectives of the simulation of combustor flows. In this paper, predictions of a swirl stabilized model combustor are compared to experimental data. The computational method is based on an Eulerian two-phase model in conjunction with an Eddy Dissipation (ED) and a presumed-shapePDF (JPDF) combustion model. The combination of an Eulerian two-phase model with a JPDF combustion model is a novelty. It was found to give good agreement to the experimental data.

Basic considerations concerning the construction and usage of multiple hot-wire probes for highly turbulent three-dimensional flows

The authors discuss the use of multiple hot-wire probes in turbulent three-dimensional flows with the aid of a graphical analysis. It is shown that hot-wire sensors can be used to measure all three components of the velocity vector as long as its direction lies within the acceptability range, which can be extended to a hemisphere if a probe with at least four wires is used. The detection of flow reversal is, however, not possible with any number of simple wire sensors. A computer-aided calibration and data reduction method for a four-wire probe is described. Test measurements in an axisymmetric free jet prove the applicability of the method.

Experimental investigation of the influence of turbulence on the transient premixed flame propagation inside closed vessels

The emphasis of the present work is the systematic study of the effect of turbulent flow exerted on explosive reactions in the deflagrative range inside closed vessels, in order to gain insight into the fundamental processes governing turbulent combustion. In the experiments performed, premixed fuel/air mixtures of varied stoichiometry were ignited by electrical sparks in two pressurized vessels of different size. Using various measurement techniques, the flow and turbulence field, pressure, temperature and flame front position were measured and interpreted as a function of time. For all investigated fuel/air mixtures a universal, linear dependency between the extrapolated turbulent flame velocity and the RMS-values of the velocity-fluctuations could be evaluated. The inclination of this linear dependency turned out to be a function of the laminar flame velocity of the particular mixture. In the range of laminar flame velocity between 0.2 m/s to 3.0 m/s and for the velocity fluctuation RMS-values between 0 m/s and 2.4 m/s , the following relation was obtained: S t S l = 1 + S l 0.25 u t ′ S l (1) The dependency of the turbulent flame velocity on the turbulent length scale of the flow could not be determined, since the experimental setup used in the present study did not permit sufficiently large variations of the length scale. It can, however, be expected that a relation between flame velocity and length scale will render the above equation dimensionless.