Laminar premixed and non-premixed flame investigation on the influence of dimethyl ether addition on n-heptane combustion

Abstract

Several different ether components have been proposed as renewable replacements for diesel fuel in compression ignition engines. From a fundamental point of view, blends of n-heptane and dimethyl ether (DME) have been considered in this study, where n-heptane is a single-component diesel surrogate and DME acts as a representative for oxymethylene ethers. n-Heptane and n-heptane / DME blends have been investigated experimentally under laminar premixed low-pressure and non-premixed atmospheric-pressure counterflow flame conditions. For the premixed case, a series of three fuel-rich flames was studied with oxygen as the oxidizer, at a constant C/O ratio of 0.47, pressure of 4 kPa, with 50% argon dilution, and an equivalence ratio ϕ near 1.5, burning pure n-heptane and mixtures of n-heptane with DME in the ratios 1:1 and 1:4. In the non-premixed counterflow configuration, two fuel/oxygen/argon flames were studied with pure n-heptane and a 1:1 DME–n-heptane mixture as the fuel. Electron ionization molecular-beam mass spectrometry was used for both flame configurations to investigate the flame structure and determine quantitative mole fraction profiles of stable and reactive species formed in the combustion process. Particular attention was given to the changes caused by partial replacement of n-heptane by DME.Notwithstanding the experimental focus of this work, the experimental results were compared with predictions from a model suitable for both n-heptane and DME that was combined for this case from recent mechanisms available in the literature for these fuels. While the overall flame structure was not significantly altered upon DME addition, with some smaller differences mainly due to temperature effects, more prominent changes and interactive effects were found for a number of primary decomposition products, oxygenated species, and higher-molecular hydrocarbon compounds. In most cases, experimentally observed trends, but not always quantitative changes, could be satisfactorily reproduced and explained by the model.