Modelling and validation of soot concentration patterns of turbulent diffusion flames based on data from plug flow reactor experiments

Abstract

Since soot content and emission of technical hydrocarbon flames are important parameters for the design and application of combustion processes, quantitative prediction of soot concentration fields in turbulent flames remains to be a major topic in combustion research. The work reported here makes a contribution towards the predictability of soot formation based upon semi-empirical models. A plug flow reactor, as an experimental system to generate soot growth data under well defined conditions, is described. Measurements of soot, gas species concentrations and temperatures are implemented into a global soot growth kinetics correlation with temperature, stoichiometry and residence time being the main influencing parameters. The correlation enables the prediction of soot growth rates as a function of temperature and residence time which are in agreement with results from other experimental systems. To apply this correlation to technical flames, extensive field measurements in free turbulent jet diffusion flames including precursor hydrocarbon species were performed. Measurements of soot mass concentrations in the plug flow reactor and in the turbulent diffusion flames, both performed by a gravimetrical soot sampling method, are compared with those determined by an optical extinction technique. The values of the gravimetrical technique agree well with those of the multiple wavelength extinction method. Soot growth kinetics from the plug flow reactor, combined with a soot oxidation model from literature, was used to predict the soot concentration field in a turbulent propane diffusion flame. The computed results compare favourably with measured data and prove the applicability of global kinetic equations for soot prediction in turbulent flames.