B. Jungfleisch

Size distributions of nanoscaled particles and gas temperatures from time-resolved laser-induced-incandescence measurements

Laser-induced-incandescence (LII) signal decays are measured in sooting premixed atmospheric and low-pressure flames. Soot particle temperatures are obtained from LII signals measured at two wavelengths. Soot particle size distributions P(r) and flame temperatures T are measured spatially resolved by independent techniques. Heat and mass transfer kinetics of the LII process are determined from measured soot particle temperatures, flame temperatures, and particle sizes. Uncertainties of current LII models are attributed to processes during the absorption of the laser pulse. Implications for LII experiments are made in order to obtain primary soot particle sizes. Soot particle size distributions and flame temperatures are assessed from measured particle temperature decays by use of multi-D nonlinear regression.

Progress in characterization of soot formation by optical methods

A two dimensional in situ optical technique is used to measure absolute soot volume fractions, particle number densities, and mean particle sizes in moderately sooting laminar and turbulent diffusion flames with high spatial and temporal resolution. These data are of special interest for the development and validation of models for the formation and oxidation of soot. The technique (RAYLIX) is based on the simultaneous two dimensional detection of Rayleigh scattering and the laser induced incandescence (LII) in combination with the detection of the integral extinction from one single laser pulse. All signals are induced by a single pulse of a frequency doubled Nd–YAG laser. Besides mean particle sizes it is of special interest to derive information about the particle size distribution by the detection of the temporal decay of the LII signal. Information about the particle size distribution can then be obtained by simulating this decay using a LII model in combination with multidimensional non-linear regression. With this strategy the parameters describing the particle size distribution as well as the temperature of the surrounding gas phase are varied so that the calculated decay of the LII signal is in accordance with the measured one. In addition to these quantities also probability–density functions (PDF), correlation functions and length scales are derived from the soot volume fraction in the turbulent flames. These quantities are of special interest for the modelling of turbulent reacting flows.

Assessment of soot volume fractions from laser-induced incancescence by comparison with extinction measurements in laminar, premixed, flat flames

For the validation of detailed chemical models for soot formation, and for their application to turbulent flames, 2-D measuring techniques are necessary to derive locally resolved soot volume fractions, particle sizes, and number densities. In one-dimensional laminar flames, these particle properties are derived from Rayleigh-scattering/extinction techniques. The extinction technique, as a line-of-sight method, is not appropriate for investigation of three-dimensional, nonstationary, turbulent systems. Therefore, for turbulent flames, other measuring techniques have to be developed that can be employed in combination with, for example, Rayleigh scattering. One technique that has been applied to obtain soot volume fractions in laminar flames is Laser-Induced Incandescence (LII) [1]. The major task in applying this technique is to clarify in which way the relative LII signals should be calibrated to yield absolute soot volume fractions. In this work, the LII technique is investigated systematically in laminar, premixed, ethyne/argon/oxygen flames using a pulsed, high-power Nd:YAG laser. Therefore, the LII signals are compared with soot volume fractions obtained using extinction. The systematic investigation of the LII technique under different detection conditions (camera gate width and gate delay times) shows that there are inconsistencies between calibrated LII signals and soot volume fractions from extinction. The aim of this work is to explain these inconsistencies by a numerical simulation of the LII signal, based on mass and energy balance equations of the heated soot particles for different flame conditions. Additionally, a sensitivity analysis with respect to different parameters of the balance equation is performed. The results indicate that geometric irregularities of the soot particles probably have to be taken into account for correct prediction of soot volume fractions using LII.

Development of a measuring technique for simultaneous in situ detection of nanoscaled particle size distributions and gas temperatures

Point measurements of time-resolved LII signals have been performed in sooting premixed low pressure flames. Soot particle size distribution and gas temperature in these flames are known from independent measurements. This data is used to validate parameters of an improved LII model, where special emphasis is taken on the accurate modelling of mass and heat transfer rates. Using this model particle size distributions and gas temperatures can be estimated from time-resolved LII signals using non-linear regression. Standard numerical methods are applied. An experimental setup is presented, which allows measuring one-dimensional maps of particle size distribution and gas temperature. The technique is based on the one-dimensional and time-resolved detection of LII signals using a Streak camera.

Re-initiation of soot surface growth in premixed counter-flow flames

In order to contribute to the discussion of the active-site approach for surface growth reactions versus recovery of radicalic surface sites by reactions with species from the gas phase, experiments in counterflow, premixed flames have been performed. Soot particles are generated in premixed, fuel-rich ethyne/argon/oxygen flames counterflowing against nonsooting flames of the same components or lean carbon monoxide/argon/oxygen flames. The nonsooting flames produce a high level of hydrogen atoms combined with a variation of the concentration of other gaseous species so that surface growth by H abstraction from C−H sites at the surface of soot particles may be reinitiated in the stagnation region of the sooting flame. These flames are compared with single, premixed, sooting flat flames in which the soot volume fraction attains a final plateau f v ∞ due to the decay of surface growth. The experimental results clearly demonstrate that depending on the flow conditions in the counterflow flames, a second onset of soot formation is detected. This second onset of soot formation can be traced back to surface growth. The experimental findings are discussed by referring to results from modeling of these flames using the hydrogen-abstraction-hydrocarbon-addition (HACA) mechanism for surface growth. From this, a shift of the competing effects of surface growth and oxidation is identified to be responsible for the reinitiation of surface growth.

Two-dimensional imaging of sizes and number densities of nanoscaled particles

In this work, an optical measuring technique for nonintrusive, in-situ, two dimensional mapping of volume fractions, number densities and median radii of nanoscaled particles is presented. The method is based on the simultaneous two-dimensional detection of RAYleigh-scattering and Laser Induced Incandescence (LII) combined with the measurement of the integral eXtinction from one single Nd-YAG laser pulse and is called RAYLIX. The experimental setup of this technique utilizes a standard Nd-YAG laser, which requires an optical delay line or a modified double pulse PIV-laser. The data evaluation based on the Mie theory of scattering in the Rayleigh regime and the linear dependence of the LII-signal on the volume fraction is discussed as well as an error analysis of the technique is given. Main error sources in the case of measuring soot arise from assumptions for the standard deviation of the lognormal particle size distribution and the refractive index of the soot particles. Applications of the RAYLIX technique are presented including investigations of sooting laminar and turbulent acetylene/nitrogen diffusion flames burning in air. From this, fundamental conclusions concerning soot formation and oxidation can be drawn and consequences for modelling of soot formation processes are discussed. Furthermore, laminar diffusion flames are investigated under conditions that are comparable to exhaust gas recirculation. In these flames increasing inert gas concentration in the coflow decreases rates of soot formation. Nevertheless, increasing inert gas concentration decreases oxidation rates leading to higher soot emission levels. Promising future applications of the RAYLIX method are discussed such as online monitoring of soot particles in carbon black production and the investigation of other nanoscaled particles. Time resolved point measurements of LII signal decays in a laminar low pressure premixed flat flame are compared with model predictions. It is shown, that information about particle size distributions can be extracted from the decay of the LII signal. By providing a method to invert LII signal decays measurement of particle size distributions is possible.