An experimental study of flame stability in a directly -fueled wall cavity with a supersonic free stream.

Abstract

An extensive study of flame stability in a cavity-based fuel injector/flameholder has been performed. Flames were stabilized in cavities with two different aft wall configurations and length to depth ratios of 3 and 4. Fuel was injected directly into the cavity using two injector configurations. Fuel injected from the aft wall of the cavity entered directly into the recirculation zone and provided desirable performance near the lean blowout limit. At high fuel flowrates, the cavity became flooded with fuel and rich blowout occurred. When fuel was injected from the floor of the cavity, excess fuel was directed out of the cavity which allowed for flame stabilization at extremely high fuel flowrates; however, this phenomenon also resulted in suboptimal performance near the lean limit where the blowout point was less predictable. Images of planar laser-induced fluorescence (PLIF) of CH, OH, and formaldehyde give insight into the flameholding mechanisms. CH layers in the cavity are thin and continuous and show structure that is comparable to lifted jet flames, while broad CH zones are sometimes observed in the shear layer. OH PLIF images show that hot recirculated products are always present at the location of flame stabilization, whereas images of formaldehyde indicate that partial premixing takes place in the shear layer portion of the flame. Nonreacting measurements of the boundary layer and the free stream velocity profiles were obtained to provide necessary boundary conditions for computational modeling. Mean and instantaneous velocity profiles were determined for the nonreacting flow using particle image velocimetry (PIV). A correlation of the blowout points for a directly-fueled cavity in a supersonic flow was accomplished using a Damkohler number and an equivalence ratio based upon an effective air mass flowrate. The chemical time was formulated using a generic measure of the reaction rate, tau<italic>_c</italic> &sim; alpha/ S2L , which was found to be adequate for correlating lean blowout data from methane, ethylene, acetylene, and hydrogen flames. Blowout data was collected at a number of conditions with varied pressure and temperature and Mach numbers of 2, 2.4, and 3. The effective air mass flowrate was determined using scaling laws for compressible mixing layers, which correctly incorporated the impact of compressibility on air entrainment.