C3H8 Flames Research Articles

Extinction of Laminar Counterflow Diffusion Flames of CH4 and C3H8 Fuels with Inert Jet Impingement

Experimental results of N2-diluted counterflow diffusion flames of CH4 and C3H18 vs, air with local extinction are reponed. The local extinction was caused by inert jet impingement on flames at selected locations either from the fuel or oxidizer side of the reaction zone. This was done to simulate how local extinction affects flame extinction over a larger flame area. The results are: (1) Local extinction of both CH4 and C3H8 flames occurs at a lower strain rate when the inert jet impingement originates from the airside. (2) The global extinction strain rate for CH4 flames is insensitive to the location and number (one vs. three) of local extinction sites. (3) For C3H8 flames, one single inert jet impinging from the airside along the centerline is more effective in causing global extinction than three inert jets impinging at regions away from the centerline. This suggests that flame extinction over a larger area may depend on strategically selected smaller local extinction sites. Furthermore, since similar results of (3) are not observed in this study for CH4 flames, fuel chemistry may also play a role in the effectiveness of local extinction transitioning to global extinction. Differences between flame stabilization mechanisms with and without local extinction are discussed and the implications for turbulent diffusion flames are outlined.

Flame heat transfer in storage geometries

This paper presents measurements of the heat flux distribution to the surface of four square towers exposed to buoyant turbulent flames.The steel towers represent an idealisation of a rack storage configuration at reduced scale. Each tower was 1.8 m high and 0.3 m×0.3 m wide. The fuel was supplied from a circular gas burner at the floor. Three different gaseous fuels were used: carbon monoxide (CO), propane (C3H8), and propylene (C3H6). These fuels cover a wide range of flame sootiness resulting in distinctly different flame heat fluxes. At the same overall heat release rates the peak heat fluxes from C3H8 flames were twice those from CO flames, whereas the peak heat fluxes from C3H8 flames were 2.8 times those from CO flames. Heat fluxes were measured by thermocouples spot-welded to the back of the steel sheets. They were measured at 52 different locations. This measurement method turns out to be simple, accurate and robust in addition to being inexpensive. Formulas are provided for the flame heat flux distribution in terms of the overall fire heat release rate, fuel sootiness and separation distance between the towers. The formulas are suitable for direct use by engineering models of fire growth in storage geometries. The paper also provides additional data needed for the development of more general CFD models capable of predicting fire growth of other geometries.

Degenerate four-wave mixing measurements of methyl radical distributions in hydrocarbon flames: Comparison with model predictions

We report an investigation of the use of degenerate four-wave mixing (DFWM) to measure CH3 radical distributions in premixed hydrocarbon flames. Experimental results are compared with flame model calculations. By exciting the Herzberg β1 band near 217 nm, we monitored DFWM signals from CH3 as a function of height above a laboratory flat-flame burner at atmospheric pressure. Coherent anti-Stokes-Raman spectroscopy (CARS) of N2 was performed to obtain temperature profiles. The dependence of the DFWM signal on temperature was shown theoretically to be reduced by the short predissociative lifetimes of the CH3 upper states. A two-color FWM mixing experiment indicated that DFWM from CH3 in CH4 flames was predominately the result of population gratings. Using population-grating theory, the DFWM signal profile was corrected for varying temperature, beam absorption, and lower-state lifetimes. The results were in reasonable agreement with flame calculations and demoustrate that DFWM can provide useful relative concentration profiles of a polyatomic radical in rich hydrocarbon flames. However, we found that a weak DFWM background signal limited detection sensitivity ≈70 ppm of CH3 in CH4 flames. The background was comparable to the CH3 signal in C3H8 flames, and it obscured the CH3 signal in rich C2H2 and C2H4 flames. Strategies to alleviate the background are discussed.

Error estimates for Rayleigh scattering density and temperature measurements in premixed flames

Rayleigh scattering has become an accepted technique for the determination of total number density during the combustion process. The interpretation of the ratio of total Rayleigh scattering signal as a ratio of densities or temperatures is hampered by the changing composition through a flame, since the average Rayleigh scattering cross-section depends on the gas composition. Typical correction factors as a function of degree of reaction, fuel and equivalence ratio were calculated. The fuels considered were H2, CH4, C2H4, C2H6 and C3H8. Factors as low as 0.7 and 0.56 were found for the heaviest hydrocarbon fuel at large equivalence ratio for interpreting the Rayleigh scattering intensity as gas density and inverse temperature, respectively. This is primarily due to the presence of CO and H2 as intermediates. As CO and H2 are subsequently oxidized to CO2 and H2O, these factors approach 1.0. Conversely, the worst case, when using H2 as a fuel, occurs in the post flame zone. However, the correction factors for H2 are near 1.0 and the errors involved will, in general, remain within the expected experimental accuracy of a typical Rayleigh scattering system. Linear correlations of correction factors with equivalene ratio and with the product of equivalence ratio and fuel molecular weight were found and presented. The interpretation of Rayleigh scattering as temperature was found to have larger errors than the interpretation as density. Corrections for changes in gas composition were applied to Rayleigh scattering temperature measurements in the post flame region of CH4 and C3H8 flames with equivalence ratios of 0.75 and 1.0. The corrected temperatures were in excellent agreement with thermocouple measurements.