61st Annual Report on Research 2016

For the majority of combustor systems, auto-ignition underpins the operational envelope, whether it is targeted for combustion initiation such as in diesel or HCCI engines or actively prevented as in gas turbine or spark-ignition engines. To date, there has been little experimental research examining the fundamental details of auto-ignition under highly turbulent conditions and in particular, little is known about the time-dependent coupling of the flow turbulence, scalar mixing of inhomogeneous reactants, and finite-rate kinetics that lead to the observed auto-ignition topology. The primary objective of this research is to use time-resolved (10 to 50 kHz), optical and laser-based imaging of mixing processes, temperature, and reactive species such as CH₂O and OH to quantify the auto-ignition dynamics of highly turbulent fuel jets issuing into hot, vitiated, oxidizing environments.

Our original target was to investigate vaporized n-heptane fuels, but we have altered our original course of action to look at gas-phase fuels: (1) dimethyl ether, (2) ethylene, (3), propane, and (4) n-butane. Following our initial burner design (reported on in year 1), a second burner was designed in year two that was larger and provides improved boundary condition. Pulsed fuel issues from a 4.5-mm-diameter nozzle into a 200-mm-diameter vitiated, co-flow using a series of fast-actuating solenoid valves. The duration of the fuel jet injection can be varied from less than 10 ms to nearly continuous. In the current program, the solenoid valve is opened (and the fuel will issue) for 55 ms, which is sufficient to auto-ignite and establish a stably-burning flame, but short enough such that system heating does not occur. Since system heating does not occur, boundary conditions remain constant for successive fuel injection/auto-ignition sequences. The annular co-flowing stream consists of the combustion products of lean, premixed H₂/O₂/N₂ combustion, which is generated by a series of 2500 1.2-mm-diameter holes which stabilize the individual flames. The high open area of the coflow (holes per area = 18/cm²) was designed to produce a very uniform coflow temperature distribution. Second, it is noted that the operating conditions have been specifically designed to parametrically study the influence of mixture composition and temperature individually.

In year two we have performed a comprehensive set of multi-view 20-kHz OH* chemiluminescence imaging in a series of turbulent propane fuel jets issuing into coflows ranging from 1150 to 1500K. These measurements are used to develop detailed statistics of ignition position (height and radial location) and time after injection. Two sets of experiments are performed: (1) measurements with a fixed coflow temperature and variations in stoichiometric mixture fraction and (2) measurements with a fixed stoichiometric mixture fraction and variations in temperature. For cases with a variation in temperature, significant increases in the mean ignition height and its fluctuation were observed as temperature decreased. There was little dependence of the radial dependence corresponding to the ignition location on temperature, which was surprising. The results also showed that there was a much greater dependence of ignition height on temperature variations than variations in stoichiometric mixture fraction.

This program has been used to support the Ph.D. research on one student, Rajat Saksena. This initial work has been summarized and submitted as a paper to the 10^th National Combustion Meeting, where Mr. Saksena will present the work. Finally, it is noted that the PI submitted a grant proposal to the National Science Foundation to follow on this preliminary work. The proposal was funded which allows Mr. Saksena support throughout the duration of his Ph.D. studies. Since this is one of the goals of ACS DNI grants, this aspect of the research project can be considered a success.