RESPONSE OF PREMIXED HYDROCARBON FLAMES WITH AND WITHOUT HYDROGEN ADDITION TO STEADY AND OSCILLATORY STRAIN RATES

Loading...
Thumbnail Image

Files

umi-umd-2912.pdf (1.77 MB)
No. of downloads: 1568

Publication or External Link

Date

2005-11-16

Citation

DRUM DOI

Abstract

The response to steady-state and oscillatory strain rates of lean, premixed hydrocarbon-air flames with and without H2 doping has been investigated analytically, numerically and experimentally. The analytical analysis provides a theoretical framework for assessing steady-state flame temperature (Tflame) response to strain rate as a function of reactant composition, through Lewis numbers (Lek), and of non-dimensional stretch defined by the Karlovitz number (Ka). An integral analysis with discrete reaction zones for each fuel has been developed, capturing the linear response of Tflame to Ka (<0.2) for steady-state CH4-H2 and C3H8-H2 lean flames as predicted by numerical models using full chemistry (GRI-Mech V3.0). This method improves predictions from single-flame zone integral analyses.

Detailed transient numerical simulations using GRI-Mech V3.0 determined the effects of H2 addition on CH4 premixed flame responses to strain oscillations. Symmetric velocity (pressure) oscillations in an opposed jet counterflow configuration were simulated for frequencies, f, ranging from 100 to 1000Hz, focusing on behavior at large Ka, near extinction. The oscillations cause flame characteristics (e.g. species mass fractions and Tflame) to vary significantly from steady-state responses. While flame responses to 100Hz oscillations closely follow quasi-steady behavior, responses for f>200Hz deviate, with phase lags and different amplitudes, from quasi-steady predictions. At f=1000Hz, the flame acts as a low pass filter, reducing amplitudes in property oscillations compared to quasi-steady analysis. For the higher frequencies, predictions indicate premixed flames can persist momentarily beyond steady-state extinction strain rate limits during oscillations.

Strain rate oscillation amplitudes required to extinguish flames, Aext, are determined numerically with the computational model and compared with experimental measurements. Simulations show that Aext increases with f due to attenuated flame response for f>200Hz. Increasing flame equivalence ratio, , and/or percentage of O2 consumed by H2, , increases reaction rates, decreasing the deviation from quasi-steady behavior and reducing the relative Aext. Counterflow premixed flame experiments with upstream speaker-imposed oscillations are also used to evaluate Aext. Phase-locked velocity measurements (with laser Doppler anemometry) assist in understanding how upstream pressure oscillations translate into velocity field (strain rate) oscillations. Measured Aext values validated trends in the computational results.

Notes

Rights