Date of Completion

8-21-2015

Embargo Period

8-20-2015

Keywords

Counterflow, Ignition, Butanol, Non-premixed, Skeletal mechanism, Hydrogen

Major Advisor

Prof. Chih-Jen Sung

Associate Advisor

Prof. Baki Cetegen

Associate Advisor

Prof. Michael Renfro

Field of Study

Mechanical Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

Critical to the development of predictive combustion models is a robust understanding of the coupled effects of chemical kinetics and convective-diffusive transport at both atmospheric and elevated pressures. This dissertation describes a new variable-pressure non-premixed counterflow ignition experiment designed to address the need for well-characterized reference data to validate such models under conditions sensitive to both chemical and transport processes. A comprehensive characterization of system behavior is provided to demonstrate boundary condition and ignition quality as well as adherence to the assumption of quasi-one-dimensionality, and suggest limitations and best practices for counterflow ignition experiments. This effort reveals that the counterflow ignition experiment requires special attention to ignition location in order to ensure that the assumption of quasi-one-dimensionality is valid, particularly at elevated pressures.

This experimental tool is then applied to the investigation of butanol isomers for pressures of 1-4 atm, pressure-weighted strain rates of 200-400 s-1, and molar fuel loading in nitrogen-diluted mixtures of 0.05-0.25 (i.e. 5-25%). Comparison of the parametric effects of varied pressure, strain rate, and fuel loading amongst the isomers facilitates a comprehensive evaluation of the effect of varied structural isomerism on transport-affected ignition. The experimental results are simulated using isomer-specific skeletal mechanisms developed from two comprehensive butanol models available in the literature, and are used to validate and assess the performance of these models. Comparison of the experimental and computational results reveals that while both models largely capture the trends in ignition temperature as functions of pressure-weighted strain rate, fuel loading, and pressure, for all isomers both models over-predict the experimental data to an appreciable extent. In addition, neither model captures the experimentally-observed ignition temperature rankings. For atmospheric pressure, the experimental results show that the “ranking” of the isomers in terms of ignition temperature (lowest to highest ignition temperature) generally follows n-butanol≈sec-butanol<iso-butanol<<tert-butanol. At 4 atm, this ranking switches to n-butanol≈iso-butanol<sec-butanol<<tert-butanol. The tert-butanol isomer is consistently an outlier, exhibiting significantly higher ignition temperatures than the other isomers, which are closely grouped for all experimental conditions. In contrast, both models predict a large spread amongst the n-/iso-/sec-butanol isomers. The model developed by Sarathy et al. [1] predicts rankings of n-butanol<sec-butanol<iso-butanol<tert-butanol at atmospheric pressure and n-butanol<iso-butanol<sec-butanol<tert-butanol, with significant offset between n-butanol and iso-/sec-butanol. The model developed by Merchant et al. [2] (excluding iso-butanol due to erroneous reaction rate descriptions for iso-butanol breakdown) predicts rankings of n-butanol<sec-butanol≈tert-butanol at atmospheric pressure and n-butanol<tert-butanol<sec-butanol at 4 atm. While the non-premixed counterflow system is found to exhibit large sensitivities to changes in fuel diffusivity, within reasonable bounds errors in the transport model cannot account for disparities between the experimental and numerical results. Further sensitivity and path analyses reveal that significant differences exist between the fuel breakdown descriptions of the two butanol models, suggesting that further work is required to better define these pathways, particularly the branching ratios from the hydroxybutyl radicals and the breakdown chemistry of the butene isomers.

In addition, this dissertation describes experimental and computational results on the non-premixed counterflow ignition of nitrogen-diluted n-butanol/hydrogen mixtures against heated air for pressures of 1-4 atm and hydrogen molar percentages in the binary fuel blends ranging from ξH=0% (pure n-butanol) to ξH=100% (pure hydrogen). The experimental data show that hydrogen addition results in a non-linear decrease in ignition temperatures that can be broken into two regimes; the hydrogen-enhanced regime of ξH=0-40% where the addition of more hydrogen significantly decreases ignition temperature, and a hydrogen-dominated regime in the range of ξH=40-100% that displays little sensitivity to further hydrogen addition. The experimental results are also simulated using n-butanol-specific skeletal mechanisms developed from two comprehensive butanol models available in the literature. These mechanisms are used to assess their ability to predict the variation of ignition temperatures as a result of hydrogen addition to the n-butanol “base” fuel. Comparison of the experimental and computational results reveals that both chemical kinetic models capture the two-regime behavior associated with hydrogen addition, though both models over-predict experimental ignition temperatures. Further chemical kinetic analysis of the mechanisms reveals that the two-regime behavior is controlled by the production of hydroperoxyl radicals, with production via the reaction of formyl radicals and oxygen molecules dominating at low hydrogen addition levels, and production via the third body H+O2+M reaction dominating at high hydrogen addition levels.

COinS