THE EFFECTS OF BUOYANCY AND DILUTION ON THE STRUCTURE AND LIFT-OFF OF COFLOW LAMINAR DIFFUSION FLAMES

Kevin T. Walsh, Marshall B. Long, and Mitchell D. Smooke Yale University Department of Mechanical Engineering New Haven, CT 06520-8284 Technical Monitor: Karen Weiland NASA Glenn Research Center, Cleveland, OH 44135 INTRODUCTION The ability to predict the coupled effects of complex transport phenomena with detailed chemical kinetics in diffusion flames is critical in the modeling of turbulent reacting flows and in understanding the processes by which soot formation and radiative transfer take place. In addition, an understanding of the factors that affect flame extinction in diffusion flames is critical in the suppression of fires and in improving engine efficiency. The goal of our characterizations of coflow laminar diffusion flames is to bring to microgravity the multidimensional diagnostic tools available in normal gravity, and in so doing provide a broader understanding of the successes and limitations of current combustion models. This will lead to a more detailed understanding of the interaction of convection, diffusion and chemistry in both buoyant and nonbuoyant environments. As a sensitive marker of changes in the flame shape, the number densities of excited-state CH (A2, denoted CH*), and excited-state OH (A2, denoted OH*) are measured in g and normal gravity. Two-dimensional CH* and OH* number densities are deconvoluted from line-of-sight chemiluminescence measurements made on the NASA KC-135 reduced-gravity aircraft. Measured signal levels are calibrated, post-flight, with Rayleigh scattering [1]. Although CH* and OH* kinetics are not well understood, the CH*, OH*, and ground-state CH distributions are spatially coincident in the flame anchoring region [2]. Therefore, the ground-state CH distribution, which is easily computed, and the readily measured CH*/OH* distributions can be used to provide a consistent and convenient way of measuring lift-off height and flame shape in the diffusion flame under investigation. Given that the fuel composition affects flame chemistry and that buoyancy influences the velocity profile of the flow, we have the opportunity to computationally and experimentally study the roles of fluids and chemistry. In performing this microgravity study, improvements to the computational model have been made and new calculations performed for a range of gravity and flow conditions. Furthermore, modifications to the experimental approach were required as a consequence of the constraints imposed by existing microgravity facilities. Results from the computations and experiments are presented in the following sections. BURNER CONFIGURATION The burner used in this experiment contains a central fuel jet (4 mm diameter) surrounded by coflowing air (50 mm diameter). The standard flow conditions, which have been measured and modeled extensively in normal gravity, consist of fuel composed of 65% methane diluted with 35% nitrogen by volume (denoted 65/35 in later discussion). The plug flow exit velocity of both fuel and coflow was 35 cm/s. These conditions produce a blue flame roughly 3 cm in length with a lift-off height of 5.5 mm in normal gravity. A wide range of flow conditions were examined in this study, with the fuel composition varied from 100% methane to a 40/60 CH4/N2 mixture in 5% increments, with all exit velocities held fixed at 35 cm/s.

g LAMINAR DIFFUSION FLAME : K.T. Walsh, M.B. Long, M.D. Smooke

COMPUTATIONAL APPROACH The computational model used to compute the temperature field, velocities, and species concentrations, solves the full set of elliptic two-dimensional governing equations for mass, momentum, species, and energy conservation on a two-dimensional mesh [3]. The resulting nonlinear equations are then solved on an IBM RS/6000 Model 590 computer by a combination of time integration and Newtons method. The chemical mechanisms employed were GRI Mech 2.11 [4] and a simpler 26-species, C2 hydrocarbon mechanism [5]. Flame structure was calculated over a range of flow conditions in both g and normal gravity. The results of a computed solution at standard flow conditions (65/35) and normal gravity were used as a starting point. In subsequent calculations, the value of the gravitational acceleration (g) was reduced by 10 cm/sec2 and a new solution calculated using Newtons method. Initial computations performed with different values for the gravitational constant indicate that buoyancy plays an important role in both the size and shape of the coflow laminar diffusion flame. Figure 1 shows the temperature isotherms for the 65/35 flame computed with g = 982.0 and 0.0 cm/sec2. It is clear from the figure that, as the gravitational constant is lowered, the flame becomes shorter and broader in appearance. Computations at different flow conditions were performed by using the 65/35 flame as an initial condition and varying the fuel mixture in 5% increments.

ComputedTemperature Profiles
50 45 40 35 30
y(mm) y(mm)

g = 9.8 m/s

50 45 40 35 30 25 20 15 10 5

g = 0.0 m/s

1900 K

25 20 15 10 5 -10 -5

298 K 0 5 10 -10 -5 0 5 10 r(mm) r(mm) Figure 1. Normal and microgravity temperature profiles at 65/35 computed with GRI Mech 2.11

While large variations in g are important in illustrating the differences between normal gravity and microgravity flame structure, small variations in g are important in determining whether the flame will be stable enough for meaningful time-averaged measurements. During the low-g portion of the KC-135 parabolic trajectory, the value of the local acceleration can vary by as much as 1% of Earths gravity. To simulate this effect, the 65/35 flame was computed with a DC gravitational constant of 10 cm/s2. These calculations, done prior to the experiments, suggested that the flame structure should be insensitive to such fluctuations for this coflowing geometry and fuel composition. These initial computations at standard flow conditions were performed with GRI Mech 2.11 as the kinetic mechanism, while the remaining calculations in this study used the simpler 26-species C2 mechanism, which is in better agreement with measured lift-off at this flow condition.

EXPERIMENTAL SYSTEM Several modifications to the laboratory-based experimental setup were required to make measurements on the KC-135. The burner and ignition system were housed inside a windowed pressure vessel to maintain standard atmospheric pressure. Spectrally-filtered, quantitative chemiluminescence images were collected with a f/4.5 UV camera lens and focused onto a cooled, unintensified CCD camera (Photometrics CH350). The camera/lens system was placed 50 cm away from the flame to ensure a wide depth of field, and high emission signal levels were collected with a 10 s exposure time. A color video camera (Sony XC-999) was used to give qualitative insight into flame structure and soot production, as well as monitor the stability of the flame in real time. During each low gravity maneuver on the KC-135, the combustion vessel pressure and airplane accelerometer signal were recorded simultaneously with the flame emission signal. The computer-controlled exhaust system kept the pressure inside the combustion vessel constant to better than 1%. All Rayleigh calibration was performed, post-flight, on the same optical setup. Details of this procedure are available in earlier work [2]. Emission measurements are integrated through the collection optics along the line of sight. Appropriate background images, taken for both CH* and OH* with the flame extinguished, are subtracted from the raw emission signal. Since our flame is axisymmetric and the imaging optics are configured such that the magnification changes by only 1% over the flame width, we can recover a two-dimensional, in-plane intensity distribution proportional to number density with the use of an algorithm that is equivalent to a two-point Abel deconvolution [6]. To make these measurements quantitative, further corrections and calculations are performed to account for quenching [7] and differences in the spectral profiles of the emission and Rayleigh signals [1]. These steps are not described in depth here, since the spatial location of the peak signal, rather that the absolute number density, is of primary interest in this study. RESULTS Although the gravitational acceleration produced by the KC-135 during low-g maneuvers is subject to both positive and negative unsteady forces (g-jitter), the flame anchoring region remained stable enough for careful emission measurements of both CH* and OH* for fuel compositions ranging from 50% to 100% methane. In dilute fuel blends (40%-55% N2), the g flame appeared as stable as a normal gravity flame. For richer fuel mixtures, which contained more than 75% methane by volume, the g flames produced significant soot luminescence, which could be seen to fluctuate as a result of g-jitter. In the 100% methane flame, g-jitter causes the sooty region in the flame to bounce, translate, and change shape considerably. Other g measurements performed recently, involving Rayleigh thermometry and laser-induced incandescence (LII), will be better able to quantify these fluctuations and their relation to g-jitter. Measured flame shape, as indicated by the spatial distributions of the CH* and OH* radicals, can change significantly between normal gravity and microgravity. In general, a microgravity flame is shorter, wider, and has a higher flame front curvature relative to its normal gravity counterpart. Additionally, the microgravity flame anchors closer to the burner surface. Although the OH* profile is more localized than that of CH*, similar gravity-induced changes in flame structure can be seen. Initial CH computations, performed at 65/35 with both GRI Mech 2.11 and the alternate mechanism, correctly predicted the buoyancy-related changes in flame shape and length while under-predicting the change in lift-off between normal gravity and g. This can be seen back in Figure 1, where flame onset doesnt change between 0 g and 1 g. This discrepancy prompted further calculations, over a range of flow conditions and g-levels, performed with careful attention to convergence criteria. We define the measured lift-off as the height above the burner where the maximum of OH* occurs, and similarly for CH in the computations. Since OH* is formed from ground-state CH, each maximum occurs in the same spatial location, as shown in earlier work [2]. We therefore have the ability to plot measured and computed lift-off heights, both in g and normal gravity, as a function of diluent level in the fuel stream. This is shown below in Figure 2. Since methane is lighter than air, density effects provide normal gravity flames with a higher lift-off than g flames

g LAMINAR DIFFUSION FLAME : K.T. Walsh, M.B. Long, M.D. Smooke

at a given flow condition. At 65/35, the predicted lift-off height agrees well with measurement at both 0g and 1g, in contrast to preliminary results. This shift illustrates the sensitivity of predicted lift-off height to the convergence tolerance used in the computation. Despite this success, the predicted normal gravity lift-off lacks the measured variation between 35% and 50% N2. As the fuel dilution is increased beyond 50% in 1 g, the lift-off height becomes highly under-predicted, until the code predicts a stable flame for fuel mixtures (65% and 70% N2) that do not ignite experimentally. Finally, the difference between computed normal and g lift-offs does not

Measured and Computed Lift-Off

16 14 12 10 8 6 4 2 0 0
Measured 1g Measured 0g Computed 1g Computed 0g

10

20

30

40

50

60

70

%N2 in Fuel Stream (by volume)

Figure 2. Measured and computed (26 species C2 mechanism) lift-off heights in 0g and 1g.

match the measured curves, which separate increasingly as the fuel mixture is diluted. A degree of sensitivity exists in experimental lift-off as well -- the details of burner design provide the measured onset of dilute flames with an uncertainty of one to two millimeters due to flame asymmetries. Nonetheless, the behavior of the computed lift-off as the fuel composition is varied suggests problems with this kinetic mechanism, which has been well validated in previous work at 65/35 [5]. Comparisons with other chemical mechanisms, notably the anticipated GRI Mech update, are planned in further studies. ACKNOWLEDGEMENTS The support of NASA under Grant NAG3-1939 is gratefully acknowledged. NASA Glenn personnel were extremely helpful from the early planning stages to the final nuts and bolts of flying the experiment successfully. In particular, we thank Jack Kolis, Eric Neumann, Karen Weiland, Joe Wilson, Jim Withrow, and John Yaniec. REFERENCES
1. Luque, J. and Crosley, D.R., Appl. Phys. B 63:91-98 (1996). 2. Walsh, K. T., Long, M. B., Tanoff, M. A., and Smooke, M. D., Twenty-seventh Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, PA, 1998, (in press). 3. Ern, A., Douglas, C. C., Smooke, M. D., The International Journal of Supercomputer Applications, Volume 9, No. 3, p. 167-186, (1995). 4. Bowman, C. T., Hanson, R. K., Davidson, D. F., Gardiner Jr., W. C., Lissianski, V., Smith, G. P., Golden, D. M., Frenklach, M., Wang, H., and Goldenberg, M.,: GRI-Mech version 2.11, http://www.gri.org, (1995). 5. Smooke, M.D., Xu, Y., Zurn, R.M., Lin, P., Frank, J.H., and Long, M.B., Twenty-fourth Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, PA, 1992, pp. 813-822. 6. Dasch, C. J., Appl. Opt, 31:1146-1152 (1994). 7. Tamura, M., Berg, P.A., Harrington, J.E., Luque, J., Jeffries, J.B., Smith, G.P., and Crosley, D.R., Combust Flame, 114(3-4):502-514, (1998).