Charlottesville Lecoustre Et Al
- 1. Effects of localC/O ratio and scalarEffects of local C/O ratio and scalar dissipation rate on sooting limits of sphericaldissipation rate on sooting limits of spherical non-premixed flamesnon-premixed flames V.R. LecoustreV.R. Lecoustre11 , B.H. Chao, B.H. Chao22 , P.B. Sunderland, P.B. Sunderland11 ,, D.L. UrbanD.L. Urban33 , D.P. Stocker, D.P. Stocker33 , R.L. Axelbaum, R.L. Axelbaum44 11 University of Maryland, College Park, MD ;University of Maryland, College Park, MD ; 22 University of Hawaii, Honolulu, HI ;University of Hawaii, Honolulu, HI ; 33 NASA Glenn Research Center, Cleveland, OH ;NASA Glenn Research Center, Cleveland, OH ; 44 Washington University, St. Louis, MOWashington University, St. Louis, MO This work was supported by NASA.This work was supported by NASA. 2007 Fall Technical Meeting2007 Fall Technical Meeting October 22nd , 2007
- 2. BackgroundBackground • Experiments: localcritical C/O ratio of about 0.6 has been identified for ethylene spherical diffusion flames. Agrees with the global C/O ratio for premixed flames (Haynes and Wagner, 1981, Glassman, 1988). • The critical soot formation T is 1250 – 1650 K (Glassman, 1998) in diffusion flames. • Short tres can prevent soot formation. • Microgravity offers strain-free 1D diffusion flames (Law, Axelbaum, Atreya, co-workers). • 17 sooting limit microgravity flames were identified by Sunderland et al. (2004).
- 3. ObjectivesObjectives • Investigate sootinglimits of microgravity C2H4 diffusion flames, focusing on the effects of: • local C/O atom ratio, • local T, • local scalar dissipation rate χ • This numerical investigation uses detailed chemistry to study flame structure.
- 4. Identification of SootingLimitsIdentification of Sooting Limits (a) 18% C2H4 → 27% O2 (b) 18% C2H4 → 28% O2 (c) O2 → 12% C2H4 (d) O2 → 13% C2H4 Tests performed in NASA Glenn 2.2 s drop tower. Fuel – C2H4 oxidizer – O2 diluent – N2 df at 2 s = 12 - 34 mm
- 5. Sooting Limit FlamesSootingLimit Flames Normal Flames InverseFlames FlameFlame AmbientAmbient XXC2H4,0C2H4,0 XXO2,0O2,0 ZZstst ttresres, s, s TTadad,, KK TTff 2s K2s K 1 Oxidizer 1 0.22 0.065 2.72 2390 1545 2 Oxidizer 0.6 0.21 0.102 1.63 2326 1492 3 Oxidizer 0.31 0.21 0.18 0.91 2226 1479 4 Oxidizer 0.25 0.23 0.225 0.665 2238 1498 5 Oxidizer 0.18 0.28 0.333 0.351 2306 1592 6 Oxidizer 0.17 0.29 0.353 0.33 2308 1593 7 Oxidizer 0.11 0.5 0.586 0.11 2381 1795 8 Oxidizer 0.11 0.8 0.685 0.044 2528 2057 9 Oxidizer 0.15 1 0.661 0.024 2740 2262 10 Fuel 1 0.13 0.041 0.059 1847 1581 11 Fuel 0.8 0.13 0.051 0.072 1835 1549 12 Fuel 0.6 0.13 0.066 0.086 1814 1515 13 Fuel 0.21 0.25 0.277 0.119 2274 1689 14 Fuel 0.19 0.3 0.336 0.122 2370 1736 15 Fuel 0.15 0.5 0.509 0.148 2539 1802 16 Fuel 0.12 0.8 0.666 0.279 2578 1729 17 Fuel 0.13 1 0.692 0.249 2670 1814
- 6. Numerical methodsNumerical methods •Sandia’s PREMIX code was modified to model steady-state or transient solutions of spherical laminar diffusion flames. • Detailed chemistry (GRI Mech. 3.0, 53 species, 325 reactions) and transport properties were used. • Discrete ordinates radiation model. • Ignition was via a steady state solution for a small domain (~1.2 cm) with adiabatic boundaries (Tse et al., 2001). • At time zero, the transient computation commenced over an extended domain (100 cm) with radiation.
- 7. Transient evolution offlameTransient evolution of flame Flame 10 Inverse flame Radiative heat losses @ 2 s ≈ 20 W All flames present same transient behavior after ignition Time (s) Flameradius(cm) Flametemperature(K)0 1 2 3 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Flame Position Temperature Frame 001 29 Aug 2007 Frame 001 29 Aug 2007
- 8. Scalar dissipation rateScalardissipation rate • Scalar dissipation rate χ = 2 DN2 ( dZ / dr )2 • By definition : • Simple definition for pure fuel : ZCH = YC + YH • Following Bilger, we can obtain a mixture fraction for ethylene: ( ) ( ) ( ) ( ) ( ) ( ) O OO C CC H HH O OO C CC H HH CHO MW YY MW YY MW YY MW YY MW YY MW YY Z 1,2,2,1,2,1, 2,2,2, 2 2 2 2 − + − + − − + − + − = mixtureofmass streamfuelfromgoriginatinmaterialofmass Z =
- 9. Mixture FractionMixture Fraction Mixturefraction T(K) 0 0.025 0.05 0.075 0.1 400 600 800 1000 1200 1400 1600 ZCH ZCHO Zst = 0.041 Flame 10 Fuel stream : pure C2H4 ZCH does not preserve Zst ZCHO preserves Zst, temperature peaks at Zst
- 10. Structure of sootinglimit flamesStructure of sooting limit flames Structure of flame 10 at 2 seconds. Peak χ is about 0.2 s-1 . High temperature gradients at flame inner side. Large high temperature field at flame outer side.
- 11. Critical local C/OratioCritical local C/O ratio C/O Ratio Standarddeviation(K) 0.4 0.45 0.5 0.55 0.6 0 10 20 30 40 50 60 70 80 ame 001 28 Aug 2007 ame 001 28 Aug 2007 For a given C/O ratio consider temperature of low scalar dissipation rate flames. Compute the temperature standard deviation for this C/O. C/OCRIT ratio minimizes standard deviation C/OCRIT = 0.51
- 12. Results : TResults: Tc/o=0.51c/o=0.51 vs. 1/vs. 1/χχc/o=0.51c/o=0.51 We consider the local T and scalar dissipation rate at the location where C/O=0.51 Local T ≈ 1400 K for 1/χ >0.5 s For shorter local mixing time, higher local T is required 1/X(C/O=0.51) (s) T(C/O=0.51) (K) 0 5 10 15 20 0 250 500 750 1000 1250 1500 1750 2000 Normal Inverse Tavg = 1400 K Frame 001 28 Aug 2007 |Frame 001 28 Aug 2007 |
- 13. ConclusionsConclusions An improvedmixture fraction definition suitable for C2H4 flame has been defined. Identified a critical C/O = 0.51 Flames with scalar dissipation rate lower than 2 s-1 at this location present same temperature averaged at 1400 K, regardless Zst. Flames with scalar dissipation rate greater than 2 s-1 need higher temperature at this location to form soot.
Editor's Notes
- #2 Thank you for attending my presentation. My name is Vivien Lecoustre, mechanical engineer PhD student at UMD. My topic presentation is Effects of C/O ratio and scalar dissipation rate on sooting limits of spherical non-premixed flames /I would like to present my co-authors who are professor Beei Huan Chao, from university of Hawaii, Professor Peter Sunderland (who is my advisor) from UMD, David Urban and Denis Stocker, from the NASA Glenn Research center and Professor Richard Axelbaum, from Washington University/. This project is supported by NASA. This work is about modeling results. We revisit an experimental paper which characterized 17 sooting limits spherical diffusion flames. Those flame were characterized in microgravity. You may be familiar with the temperature and scalar dissipation rate effects on soot formation. But maybe not on the C/O ratio. Sooting limits arise from a competition between the reaction of fuel pyrolysis and the reaction of soot oxidation. The former will be likely to occur when we have a large number of carbon. The latter, on the contrary, will be likely to occur when great number oxygen atoms are present, greater than the carbon atoms. C/O ratio is a useful parameter to asses the importance of one or the other reactions. The past work carried on soot formation was first established for premixed flame due to nearly constant temperature and Carbon to Oxygen atom ratio, C/O ratio, in the soot-forming regions of premixed flame, and to the decoupling effect of those two parameters. Experiments, carried out by Haynes and Wagner and Glassman using a ethylene premixed flame, show that there is a critical C/O ratio value for soot inception. Soot did not form into ethylene flames with a global C/O ratio less than 0.6
- #3 Despite the differences between soot inception in premixed and non premixed flame, the C/O ratio has been shown to be relevant to sooting limits in diffusion flame. Temperature plays an important role in soot formation. In diffusion flame, an increase of temperature leads to an increase of soot. Critical temperatures at the onset of soot formation for diffusion flame have been observed to range from 1250K to 1650K by Glassman. Time is the third important parameters in soot formation. It has been shown that short residence time within the flame can prevent soot formation. The microgravity allows the observation of buoyancy free and strain free diffusion flames, along with the observation of flame having longer residence or mixing time than buoyant flame. Moreover it allows the observation of the simplest flame possible, get rid off of all the non one dimensional effects because diffusive, convective and radiative transports are all in the radial direction. 17 sooting limits flame has been then characterized by Sunderland and al.
- #4 The objectives of this works are to investigate numerically the inner structure of the microgravity sooting limits diffusion flames, focusing on the effect of the C/O ratio, the local temperature within the flame, especially in the vicinity of the critical C/O ratio, and the local scalar dissipation rate, which is related to the mixing rate at a specific location and which can be seen as the inverse of a local characteristic mixing time. We assume that soot formation occurs when 3 conditions are meet. C/O ratio needs to be greater than a critical value, which provides enough available carbon atoms for soot formation and overcoming soot oxidation. Temperature is needed, low temperature freezing reactions. Time, given by inverse of local scalar dissipation rate, balances temperature. We now think that the C/O ratio scales with H/O ratio, which is a key parameter to specify soot oxidation, which is mainly caused by presence of OH (hydroxile radical). The flame structure is study through the use a 1-D detailed chemitry code, aiming to quantify the critical parameters values.
- #5 The 17 sooting limits spherical diffusion flames, which you have a pictures of some, have been observed in microgravity conditions. This is from experiment we did several years ago using the 2.2 s Nasa Glenn drop tower. Those flames were observed to reach their sooting limits at 2s. Soot was observed within the flame until 2 seconds. At 2 seconds and after, flames were permanently blue. The experimental process was to vary the concentration of fuel or oxidizer leading to soot suppression at 2s. Picture of four flames, Presence of bleu indicating the location of the main reaction. On some picture we see the presence of a light yellow part, presence of soot. The one with the yellow sooting limit flame. Those flames are fueled with ethylene, diluted in diverse proportion with Nitrogen, but all equivalent to 1.51 mg/s pure ethylene. The oxidizer used was a mixture of nitrogen and oxygen. Role of convection is study by investigating normal flame and inverse flames. Normal flame means that fuel is flowing out from the burner toward an environment of oxidizer (top picture). Inverse flame means that oxidizer is flowing out from the burner, flow in heading toward fuel (lower pictures). Dimension involved : the burner diameter is 6.4 mm and the flame diameters range between 12 to 34 mm.
- #6 This table shows the different characteristics of the 17 sooting limits diffusion flame considered. I apologized for all the numbers in this slkide. This is our test Matrix. Normal and inverse flame are tabulated, along with mole fraction of fuel and oxygen. Stoichiometric mixture fraction ZST, the residence time, define here as the mass of gas contained between the flame and the burner surface divided by mass the burner flow rate, The adiabatic flame temperature computed using Chemical Equilibrium with Application. (CEA) an dthe flame temperature at 2 seconds given by the code. From this table, one can appreciate the wide range of conditions covered. We consider flame with pure fuel flame (flame 1 and 10) to flame with a fuel concentration as low as 11% (Flame 8 and 7). The oxygen concentration varies from 13 % (flames 10,11 and 12) to pure oxygen. Therefore, we cover a broad range of mixture fraction, from 0.041 (flame 10) to almost 0.7 (Flame 17). Wide range of residence time is covered. The minimum is 29 ms for the flame 9 to 2.72s for the flame 1. Residence time represents the time for a fluid particle at the burner to reach the location of maximum temperature. However think scalar dissipation rate is the key. Adiabatic flame temperature ranges from 1814 K (flame 12) to 2740 K (flame 9). At 2 seconds the flame temperatures have dropped in more less extend. Max T is 2262 K and the lowest T is 1479 K, which is not the flame having the lowest adiabatic flame temperature.
- #7 The 17 flames are numerically modeled using a 1-D code to investigate their inner structure. Get accurate temperature field, species concentration at the desire time. We used a modified Sandia’s PREMIX code, with the adequate modifications to model 1-D spherical non-buoyant flames, either at steady state or time accurate transient solutions. The code uses a detailed chemistry model (GRI Mech. 3.0- 53 species, 325 reactions) and detailed transport properties. No soot models incorporated. The optically thick radiative heat losses from H2O, CO2 and CO are modeled using the discrete ordinates radiation. Ignition is modeled used the same process than Tse. The steady state solution for the flame is computed on a small domain (1.2 cm) with adiabatic boundaries condition on both sides. This solution is used as initial condition for the computation of the transient solution. The transient solution is computed over a 100 cm radius domain, with constant boundary temperatures and enabling radiative losses. This process gives a very good agreement with the experimental results. The time accurate transient computation were stopped at 2s, corresponding to flame at their sooting limit point.
- #8 Following ignition, the flame will grow with the time, expending itself. The plot here represent the variation of flame location and flame temperature for the flame 10 over a 3 seconds period, with the origin taken as ignition. The growth of the flame is observable on this chart, reading the triangle. As the flames growth, its temperature decreases, mainly due to increase of radiative heat losses, which have shown to be proportional to the flame surface area. At the sooting limit point (2 second) the flame is still present (Tf = 1581 K), the radiative heat losses are about 20 W (out of about 70 W)-1.51 mg/s C2H4 equivalent flame.
- #9 Our approach in this work seek for a characteristic local time scaling for the mixing of species and time allowed for soot formation reaction. The scalar dissipation is inversely proportional to such a time. The scalar dissipation used for our study is given by the above formula, and is equal to twice the diffusivity of N2 multiply by the square of the mixture fraction gradient. The diffusivity of N2 is justify since N2 is one of the dominant species at the vicinity of the flame. We compute locally the scalar dissipation using the local value of the N2 diffusivity given by the detailed transport model. The scalar dissipation rate is very dependant on mixture fraction Z. By definition the mixture fraction is defined locally to be the local mass of material originating from the fuel stream divided the local mixture mass. Since we are dealing with detailed chemistry, we must considered mass fraction of atoms, since fuel pyrolysis leads fuel drop mass fraction. A trivial way to define the mixture fraction is to consider Z=Yc+Yh since C and H only comes from the fuel stream. However this definition is not suitable as I will show in the next slide. Another way is to define the mixture fraction as below, Bilger like mixture fraction definition,, where we account for C, H and also O, only coming from the oxidizer stream. The suffix 1 stands for the condition at the fuel stream, 2 stands for the conditions at the oxidizer stream.
- #10 Test the accuracy of our two model, using the same way than Mahalingam. Zch is based on the sum of the local mass fraction of carbon atom with the local mass fraction of hydrogen. ZCHO accounts for Oxygen, Hydrogen and Carbon. In this graph we plot the field temperature of the flame 10, at 2 seconds, in the mixture fraction space, using both model. The flame 10 is an inverse flame with an environment of pure fuel. The temperature curve given by ZCH does not peak at the stoichiometric value of mixture fraction but is shifted on higher value of Z. Hence, this formulation of Mixture fraction is not appropriate since it does not preserve the Zst. On the contrary, the ZCHO temperature curve peaks at Zst, meaning that this formulation preserves the value of Zst. Hence this model is adopted for the computation of mixture fraction, which derived is used for the scalar dissipation rate computation.
- #11 We consider the predicted solution at 2 seconds for each 17 flame. This chart illustrates the structure of an spherical diffusion flame. The temperature, C/O ratio, Mixture Fraction and scalar dissipation rate are the value from flame 10. Those value are plotted against the domain radius, with the origin taken at the center of the burner. The burner radius is 0.32 cm. From the temperature plot, we can notice : Stiff gradient close to the burner. The outer side of the flame presents a boarder high temperature area. The inverse nature of the flame can be seen from the behavior of the C/O ratio. Close to the burner, the C/O ratio is very low, close to 0, we are on the oxidizer side. The C/O ratio grows when we increase the radius. The area where the C/O is greater than 1/3 corresponds to the fuel side. On this side, when the C/O ratio increases, the temperature decreases. If we put aside the temporal effect on soot formation, the balance between C/O ratio and T can be understand this way: Important C/O provide a high number of available carbon atoms, hence promoting soot formation by fuel pyolysis. However, we can notice that the temperature on this part of the flame is low, hence blocking fuel pyrolysis. To overcome this, we need to consider hotter part, which are located in area of low C/O ratio, in area where Oxidizer and fuel pyrolysis are on the same level. This area correspond to the area close to a critical value of C/O ratio. At the critical C/O ratio, the temperature is the maximum allowed temperature for soot formation. However since oxidation of soot equal soot formation, we do not create more soot.
- #13 From our numerical data, a Critical C/O ratio value has been identified. Its critical value is 0.51. We can study the couple effects of temperature and scalar dissipation rate. For that we plot the temperature at the critical C/O ratio location against the inverse scalar dissipation rate at the same location for every 17 flames. Using the inverse of scalar dissipation rate gives use a characteristic mixing time. The plain black symbols plots the T of normal flame, will the open symbols plots the T of the inverse flame. 2 different Temperature behaviors can be observed from this plot, depending on the value of the scalar dissipation rate. The flames with a inverse scalar dissipation rate greater than 0.5 seconds present a nearly constant temperature at the critical C/O ratio. This gives us a deep insight on mechanism of soot formation. For flame with sufficient mixing time, temperature is the controlling parameters. This temperature averaged at 1400 K. The second behavior of the flames are given by the following: When the mixing time or inverse of scalar dissipation rate is lower than 0.5 s-1, then time is the controlling parameters. Shorter time requires increased temperature to form soot. Hence for short mixing time, scalar dissipation rate is the controlling parameter. For a fixed temperature at a same C/O ratio, increasing the scalar dissipation rate will lead to soot free flame.