Nathan R. Grady

An Experimental Study of Homogeneous Anisotropic Turbulence in Channel Flow

A novel high bleed blown jet grid turbulence generator has been developed and tested in both a small-scale divergent duct and subsonic wind tunnel facility developed for flame kernel studies. The small-scale divergent duct is shown to produce an experimental realization of homogeneous turbulence that is comparable to traditional larger turbulence facilities. The high-bleed blown jet grid is shown to produce high Reλ turbulence, with 90 < Reλ < 298 for tunnel total mass flow bleed ratios ranging from 10% to 30%. The resultant turbulence is comparable to that achieved by active vane grids, however the blown jet grid produces such results at grid mesh Reynolds number ReM ≤ 6580, which is less than half of what is observed for similar turbulence generated by active vane grids. Turbulence generated under shared inflow conditions as the divergent duct study are evaluated in the subsonic facility configuration. The “High Contraction Ratio” anomaly is identified and analyzed, enabling measures for its removal from future turbulence studies in the subsonic facility.

Hydroxyl Tagging Velocimetry in a Supersonic Flow over a Piloted Cavity

Velocity measurements of a Mach 2 flow over a wall cavity both with and without a strut were obtained using hydroxyl tagging velocimetry (HTV). HTV is a two step process: 1) the “write” phase where water molecules present in the flow are dissociated into OH+H by an ArF excimer laser operating at 193 nm, 2) the “read” phase where after a fixed time delay the OH field is interrogated with planar laser-induced fluorescence. An 11x11 OH grid was used to obtain velocities at ~120 grid points. Single-shot profiles were analyzed to produce mean and RMS deviation velocity profiles of the flow. Such profiles were obtained for the freestream, recirculation region in the wall cavity, and in the wake behind the strut.

UV Raman Scattering Measurements of a Mach 2 Reacting Flow over a Piloted Cavity

Abstract : UV Raman scattering measurements were made in a Mach 2 supersonic air flow over a cavity piloted with ethylene fuel (C2H4). The wall cavity simulated the pilot region of a scramjet combustor. In the UV Raman system, a 248 nm KrF excimer laser beam (400 mJ/pulse, 20 ns pulse length) was used to excite the Raman scattering in the combustion zone. Raman scattered light in the 254-278 nm spectral region allows measurement of the following molecular species: CO2 (257 nm), 02 (258 nm), N2 (263 nm), C2H4 (268 nm), H2O (273 nm) and H2 (277nm). To avoid damaging the fused-silica windows on the combustion test section: 1) the laser pulse was stretched from 20 ns to 150 ns using two optical delay cavities, 2) a long focal length lens (5 meters) focused the KrF beam to a relatively large diameter (1 mm diameter) and 3) the laser energy was decreased to 100 mJ/pulse. Under these conditions, the high power pulsed laser beam passed through the side fused-silica windows without inflicting damage. Raman scattered light was collected from the top fused-silica and was focused into a 0.32 meter spectrometer.

Propagation of Premixed Flame Kernels in High Speed Channel Flows with Moderate Turbulence

Nathan Grady, Robert Pitz Vanderbilt University Brad Ochs, Tom Slais, David Scarborough, Suresh Menon Georgia Institute of Technology A premixed CH4/air wind tunnel with active turbulence generated with a vane grid design has been implemented to study turbulent combustion in high speed flows. The current experiment focused on stoichiometric kernels in M = 0.1-0.3 mean flows with turbulence from 0.4 m/s to 2.0 m/s (u’/SL = 1 and 5 respectively). Flame kernels were laser ignited using ~10 mJ/pulse from a 532 nm Nd:YAG laser, and allowed to freely propagate downstream. Combined OH/CH2O PLIF diagnostics were used to study the flame growth as a function of downstream distance/propagation time. CH2O images did not indicate any significant local extinction, so the preliminary results shown here only examine the OH PLIF images. Flame surface density was found to decrease with respect to time due to the decaying turbulence in the tunnel. However, flame brush thickness increased resulting in an increase in the consumption rate until the kernel’s expansion waves interacted with the wall. Additionally, instantaneous flame lengths determined from single-shot images were found to increase with turbulence as well.

Characteristics of Freely Propagating Premixed Flame Kernels in Supersonic Turbulent Channel Flows

A new supersonic facility was developed to study the effects of turbulence intensity and mean flow compressibility on freely propagating flame kernels in a M = 1.5 channel flow. Two devices, a passive grid of holes and an active grid of rotating vanes, were used to generate turbulence. The 5 × 5 cm test section was optically accessible on four sides and had variable divergence walls on two sides to account for Fanno and Rayleigh flow deceleration. Methane-air freely propagating flame kernels were ignited with a single pulse of a frequency doubled green (532 nm) Nd:YAG laser and imaged via simultaneous high speed OH* chemiluminescence and high speed Schlieren photography. Planar velocity statistics were measured using particle image velocimetry. The generated turbulence intensity and isotropy are compared for the baseline and passive/active grids. Turbulent burning velocity trends are presented versus turbulence intensity and equivalence ratio. The flame speeds are fit to a functional dependence on the laminar flame speed and RMS velocity fluctuation and compared to classical low speed formulations.