Experimental and Modeling Studies of Small Molecule Chemistry in Expanding Spherical Flames

Abstract

Accurate models of flame chemistry are required in order to predict emissions and flame properties, such that clean, efficient engines can be designed more easily. There are three primary methods used to improve such combustion chemistry models - theoretical reaction rate calculations, elementary reaction rate experiments, and combustion system experiments. This work contributes to model improvement through the third method - measurements and analysis of the laminar burning velocity at constraining conditions.

Modern combustion systems operate at high pressure with strong exhaust gas dilution in order to improve efficiency and reduce emissions. Additionally, flames under these conditions are sensitized to elementary reaction rates such that measurements constrain modeling efforts. Measurement conditions of the present work operate within this intersection between applications and fundamental science. Experiments utilize a new pressure-release, heated spherical combustion chamber with a variety of fuels (high hydrogen content fuels, formaldehyde (via 1,3,5-trioxane), and C2 fuels) at pressures from 0.5 - 25 atm, often with dilution by water vapor or carbon dioxide to flame temperatures below 2000 K. The constraining ability of these measurements depends on their uncertainty. Thus, the present work includes a novel analytical estimate of the effects of thermal radiative heat loss on burning velocity measurements in spherical flames. For 1,3,5-trioxane experiments, global measurements are sufficiently sensitive to elementary reaction rates that optimization techniques are employed to indirectly measure the reaction rates of HCO consumption.

Besides the influence of flame chemistry on propagation, this work also explores the chemistry involved in production of nitric oxide, a harmful pollutant, within flames. We find significant differences among available chemistry models, both in mechanistic structure and quantitative reaction rates. There is a lack of well-defined measurements of nitric oxide formation at high temperatures, contributing to disagreement between chemical models.

This work accomplishes several goals. It identifies disagreements in pollutant formation chemistry. It creates a novel database of burning velocity measurements at relevant, sensitive conditions. It presents a simple, conservative estimate of radiation-induced measurement uncertainty in spherical flames. Finally, it utilizes systems-level flame experiments to indirectly measure elementary reaction rates.