Structure, Extinction, and Ignition of Non-Premixed Flames in the Counterflow Configuration

Transcription

1 Structure, Extinction, and Ignition of Non-Premixed Flames in the Counterflow Configuration Ryan Gehmlich STAR Global Conference 2013 Orlanda, Florida March

2 Outline Background Developing Reaction Mechanisms for Combustion Systems Validating Mechanisms Using Ideal Flames Case Study I: Extinction and Autoignition of ethane/air/nitrous oxide flames Case Study II: Extinction and Autoignition of Lightly- Branched Octane Isomers Summary 2

3 Motivation for chemical kinetic studies in combustion Gun/Artillery Propellants Aviation Engines Power generation Rockets/Missiles Ground Transportation 3

4 Modeling combustion phenomenon Combustion modeling tools are now able to couple CFD with detailed chemistry For this to work, we need to develop validated chemical mechanisms! Validate chemical mechanisms through the use of 1D ideal flames 4

5 Reaction Mechanisms Global Reaction of Hydrogen Combustion 2 H 2 + O 2 2 H 2 O(g) + heat 5

6 A few combustion mechanisms San Diego Mechanism C1-C4 hydrocarbons, hydrogen, nitrogen oxides, JP10, heptane GRI-Mech Natural gas (including NO) USC-Mech II C1-C4 hydrocarbons, hydrogen Jetsurf 2.0 Jet fuel surrogates (i.e. n-dodecane, n-butyl-cyclohexane, etc.) Creck Modeling Group C1-C16 hydrocarbons, alcohols, esters, reference components of surrogates of real fuels Lawrence Livermore National Laboratory C1-C7 hydrocarbons, alcohols, dimethyl ether, etc. Engine Research Center, UW Madison n-heptane, n-butanol, PAH, biodiesel 6

7 Counterflow burner for combustion kinetics Laminar, opposed-flow diffusion flames can be established experimentally using this simple flow geometry Counterflow flames can be simulated by applying the equations of continuity, motion, energy, and species concentration Boundary conditions are welldefined at the duct exits Properties such as temperature and species concentrations can be modeled in 1-dimensional space 7

8 Flow Field Characteristics Oxidizer strain rate, Flow is momentum balanced such that Duct separation distance, L = 10 mm (extinction) or 12 mm (ignition) Three screens of 200 mesh ensure plug flow at the duct exit planes 8

9 Flow Visualization Oxidizer duct Illuminated by HeNe laser sheet Seeded with baby powder (corn starch), micron diameter Streamlines demonstrate plug flow at the oxidizer duct boundary Fuel duct 9

10 Numerical Simulation of Flames Digital Analysis of Reacting Systems (DARS) Basic Includes 0D and 1D reactor models, including a 1D opposed flow diffusion flame model Visualize mechanisms and species data Perform sensitivity analyses, flow analyses, and mechanism reduction Visualize species and temperature profiles, compare predictions with experiments, tune the mechanisms! 10

11 Using DARS for a 1d opposed flow reactor Current versions of the DARS GUI do not having looping capabilities for opposed flow reactors Looping can be achieved using a high level programming or scripting language and the command line tools of the DARS interface (I used MATLAB) Convergence to solutions tends to be smoother, faster, and more consistent than other commerical codes on the market Select run path Use previously generated start solution? No Yes Copy start solution to run path T j, Y i,j, p V j,l, grid settings, solver settings Chemistry set (mechanism, thermo and transport files) Write GasComposition.txt Write FlameUserSettings.txt Copy to run path: InputRedKinMec.txt InputRedKinTherm.txt Use better start solution or adjust grid/solver settings No Create folders in the run path for output files (DARS command line tools cannot do this) Convergence? Yes Run Chamble.exe within the run folder 11

12 Case Study I: Extinction and Autoignition of Ethane/Air/N 2 O Flames Improve knowledge of detailed and reduced chemical kinetic mechanism for gas-phase reactions in the ignition of gelled hypergolic propellants Gas-phase N 2 O chemistry is a subsystem of nitramine propellant combustion Data can be used to validate or improve chemical mechanisms for nitrogen chemistry in these systems 12

13 Experimental Apparatus 13

14 Numerical Computations All computations done using DARS v and 2.08 Used the latest San Diego mechanism including NO x 61 reactive species, 297 reversible reactions Some cases checked using Creck C1-C3 mechanism with NO x, v (111 species, 1,835 reactions, 2,357 including reverse) 14

15 Extinction The structure of the reactive flow-field depends on the five independent parameters Y F,1, Y N2 O,2, Y O2,2, T 1, and T 2. The experiments were conducted with T 1 =T 2 =298 K. This reduces the number of independent parameters to three. To facilitate comparison of predictions of asymptotic analysis with experimental data, the temperature for complete combustion, T st, and the stoichiometric mixture fraction, Z st, was fixed. This reduced the number of independent parameters by two, leaving only one independent parameter. The strain rate at extinction, a 2, was recorded as a function of the mass fraction of N 2 O, Y N2 O,2. 15

16 Results N 2 O/O 2 /N 2 C 2 H 6 /N 2 At a fixed flame temperature (T st ) and location (Z st ), replacing O 2 by N 2 O promotes extinction (inhibition) 16

17 Ignition Mass Fractions and Boundary Temperatures Fuel Stream Balance N 2, measured by a thermocouple below the fuel duct screens Oxidizer Stream Contains a mixture of N 2 O, N 2, and air Kept a constant mass fraction of O atoms in the oxidizer stream for varying T 2 is increased slowly until ignition occurs, all flows are constantly recalculated to retain a constant strain rate and a momentum balance 17

18 Results Autoignition temperature vs. strain rate for pure ethane-air flame 18

19 Results Autoignition temperature as a function of N 2 O mass fraction in the oxidizer stream 19

20 II. Extinction and Ignition of Lightly-Branched Octane Isomers 20

21 Motivation Previous studies on 2-methylalkane and singly methylated alkanes (such as 2-methylheptane) showed significantly different combustion behavior than their linear alkane counterparts The present study extends this to work with iso-alkanes that have methyl groups on different locations and with more than one methyl substitution 2,5 dimethylhexane (C 8 H 18-25) and 3-methylheptane (C 8 H 18-3) are important components of petroleum-based transportation fuels Octane 2,5 dimethylhexane 2-methylheptane 3-methylheptane 21

22 Experimental Conditions Mass Fractions and Boundary Temperatures - Extinction Fuel Stream A range of mass fractions of fuel from Balance N 2 Oxidizer Stream Contains undiluted air Strain rate is increased slowly until extinction occurs 22

23 Experimental Conditions Mass Fractions and Boundary Temperatures - Autoignition Fuel Stream Balance N 2 Oxidizer Stream Contains undiluted air T 2 is increased slowly until ignition occurs, all flows are constantly recalculated to retain a constant strain rate and a momentum balance 23

24 Numerical Computations Mechanism development by Lawrence Livermore National Laboratory in Livermore, CA Used two mechanisms: LLNL detailed mechanism 767 species, 3,961 reversible reactions LLNL skeletal mechanism 241 species, 1,587 reversible reactions 24

25 Results: Extinction Measured and predicted strain rate at extinction Methyl branch location makes little difference in extinction between 2- and 3- methylheptane 2,5 dimethylhexane extinguishes at lower strain rates 25

26 Results: Autoignition Measured and predicted autoignition temperature Methyl branch location makes little difference in extinction between 2- and 3- methylheptane 2,5 dimethylhexane autoignites at higher temperatures 26

27 Summary DARS 1D solvers are a useful tool in the development, validation, and reduction of reaction mechanisms DARS has proven to be an excellent tool in our arsenal fast, consistent convergence to flame solutions without too much fuss Thanks: Fabian Mauss, Lars Seidel, Karin Frojd 27