The work presented here discusses the effects of strong acoustic forcing on jet flames in crossflow (JFICF) and the physical mechanisms behind theses effects. For forced non-premixed JFICF, the jet fuel flow is modulated using an acoustic speaker system, which results in a drastic decrease in flame length and soot luminosity. Forced JFICF are characterized by periodic ejections of high-momentum, deeply penetrating vortical structures, which draws air into the jet nozzle and enhances mixing in the nearfield region of the jet. Mixture fraction images of the non-reacting forced jet in crossflow are obtained from acetone planar laser-induced fluorescence and show that the ejected jet fluid is effectively partially premixed. Flame luminosity images and exhaust gas measurements show that forced non-premixed JFICF exhibit similar characteristics to unforced partially-premixed JFICF. Both strong forcing and air dilution result in net reductions in NOx, but increases in CO and unburned hydrocarbons. NOx scaling analysis is presented for both forced non-premixed and unforced partially-premixed flames. Using flame volume arguments, EINOx scales with amplitude ratio for forced non- premixed flames, but does not scale with air dilution for unforced partially-premixed flames. The difference in scaling behavior is attributed to differences in flame structure. The effect of forcing on the flowfield dynamics of non-premixed JFICF is investigated using high-speed stereoscopic particle image velocimetry and luminosity imaging. The frequency spectra of the windward and lee-side flame base motions obtained from luminosity movies of the forced JFICF show a peak at the forcing frequency in the lee-side spectrum, but not on the windward-side spectrum. The lee-side flame base responds to the forcing frequency because the lee-side flame base stabilizes closer to the jet exit. The windward-side flame base does not respond to the forcing frequency because the integrated effect of the incident crossflow and vortical ejections leads to extinction of the flame base. From the PIV measurements, flowfield statistics are conditioned at the flame base. The local gas velocity at the flame base did not collapse for forced and unforced JFICF and was found to exceed 3SL. The flame propagation velocity was determined from the motion of the flame base, which is inferred from regions of evaporated seed particles in the time-resolved PIV images. The flame propagation velocity collapses for forced and unforced JFICF, which implies that the flame base is an edge flame; however, the most probable propagation velocity, approximately 2-3SL, is larger than propagation velocity predicted by edge flame theories. A possible explanation is that the flame propagation is enhanced by turbulent intensities and flame curvature. / text

The purpose of this study was to investigate the thermochemical properties of a transient supersonic flame. Creation of the transient flame was controlled by pulsing air in 200 millisecond intervals into a combustor filled with flowing methane. The combustor was designed following well-known principles of jet engine combustors. A flame holder and spark plug combination was used to encourage turbulent mixing and ignition of reactant gases, and to anchor the transient flame. Combustion created a high temperature and pressure environment which propelled a flame through a choked de Laval nozzle. The nozzle accelerated the products of combustion to a Mach number of 1.6, creating an underexpanded transient flame which burned for approximately 25 milliseconds. Qualitative information of the flame was gathered by two optical systems. An intensified charge-coupled device (ICCD) was constructed from constitutive components to amplify and capture the chemiluminescence generated by the transient flame, as well as the spatial structure of the flame at specific phases. To gather temporal data of a single transient event as it unfolded, a z-type schlieren optical system was constructed for use with a high speed camera. The system resolves the data in 1 millisecond increments, sufficient for capturing the transient phenomenon. The transient system was modeled computationally in Cantera using the GRI-3.0 reaction mechanism. Experimental conditions were simulated within the zero- dimensional computation by explicit control of the reacting gas mass flow rates within the system. Results from the computational model were used to describe the ignition process. The major limitation of the zero-dimensional reactor model is homogeneity and lack of spatial mixing. In this work a Lagrangian tracking model was used to describe the flame behavior and properties as it travels within the zero-dimensional reactor towards the nozzle. Following this, the flow expansion through the de Laval nozzle was calculated using one-dimensional isentropic relations. The computed reactor model data was then contrasted to experimental results from the ICCD and high speed schlieren images to fully describe the events in the transient supersonic flame.