Date of Completion

7-24-2012

Embargo Period

7-23-2012

Advisors

Michael W. Renfro; Tianfeng Lu

Field of Study

Mechanical Engineering

Degree

Master of Science

Open Access

Open Access

Abstract

Flame dynamics of a bluff body stabilized turbulent premixed flame as it approaches lean blowoff is of interest for practical applications. It is also important to understand the flame behavior under harmonic flow perturbations as it may occur due to acoustically unstable operation of compact combustors. In this study, a harmonically excited conical flame was studied to determine its behavior under strong burning and near blowoff conditions. Chemiluminescence imaging was employed using a Photron high speed camera to characterize the phase resolved flame characteristics for a range of excitation frequencies from 50 to 400 Hz in confined and unconfined geometries. Phase-resolved measurements of stretch rate along the flame front can provide important insights for understanding of local flame behavior particularly near blowoff. Phase resolved particle image velocimetry (PIV) technique was utilized to map the different phases of the velocity field in harmonically modulated flow conditions. Processing of PIV data along the flame front identified from seed density change was employed to determine flame front location and utilize the PIV data to determine the stretch rate variations at different phases of flame oscillation. Oscillations of recirculation zone length were characterized at all external harmonic excitation frequencies accompanied by a cyclically varying strain rate along the flame front. At certain phases of the cycle, the strain rate reaches a maximum just before the phase when the recirculation zone length is a minimum. At this point, flame pinching and vortex breakdown phenomena were observed. This vortex breakdown phenomenon accompanies flame blowoff when the ratio of the minimum recirculation zone length to the convective length scale is in the range between 0.3 and 0.5.

Lean premixed flame blowoff experiments from two different experimental configurations were analyzed with the objective of determining the most appropriate time scale associated with lean premixed flame blowoff. Experimental results from combined particle image velocimetry and OH planar laser induced fluorescence reported earlier were analyzed to produce flame front conditioned strain rate probability density functions. The time scale associated with the inverse of the mean strain rate for each experimental condition is then compared with the three time scales from an opposed counter flow flame configuration using OPPDIF, a flame propagation time scale using PREMIX and an extinction time scale using a perfectly stirred reactor (PSR). In each case, the USC II chemical kinetic mechanism for C1-C4 hydrocarbons was used for detailed kinetics. It was found that the experimentally determined mean strain rate along the flame front just prior to blowoff correlates best with the extinction strain rate calculated for the same mixture using OPPDIF. The ratio of the extinction time scale from the OPPDIF to the experimental time scale obtained from the mean strain rate is found to be constant for all cases, namely unvitiated, axisymmetric, 0.15 vitiation and 0.25 vitiation respectively with a different constant value for each case irrespective of the global equivalence ratio. The other two time scales determined from PREMIX and PSR did not correlate with the experimental data as well. The good correspondence of experimental and computed results for extinction strain based time scales for the two different experimental configurations and for two different upstream vitiation levels and no vitiation suggests that premixed flame blowoff is due to attainment of a critical extinction strain rate in the mean and a larger fraction of the strain rate probability distribution exceeding this critical strain rate.

Major Advisor

Baki M. Cetegen

Download

Share

COinS