Chemistry in Combustion Processes, part 1
Exercise 4:
CFDsimulation of methane combustion with air in a turbulent diffusion flame.
The main steps of the exercise are to
-Set up and carry out the calculation
-Draw graphs of the species mole fractions (O2, CH4, H2O, H2, CO, and CO2) using contour and/or XY plots.
-Answer the questions below.
Questions
1)What is the gas composition at the outlet, and does this agree with your previous calculationsusing stoichiometry (exercise 1), equilibrium (exercise 2), and kinetics (exercise 3)?
Collect the results (mole fractions of O2, CO2, H2O, CO, and H2) from ex 1-4 in a table.
2)In the CFD simulation, what is the distance and time needed along the symmetry axis for complete combustion of methane (CO concentration ≈ the outlet concentration)?
3)Compare the distance and time (question 2, above) to the results from exercise 3 (plug flow reactor). What are the differences?
4)As compared to kinetics only (exercise 3), what is the impact of introducing fuel and oxidizer separately (=the need to mix reactants) on the burnout time of methane?
Input to the calculation
10 MW:methane flow 0.0125 kmol/s = 0.2 kg/s
Combustion air: 0.21 vol-% O2 (rest N2) → O2 mass fraction = 0.233
Air factor 1.2: combustion air flow 4.3115 kg/s
Both methane and combustion air enter at a temperature of 1273 K.
Steps for completing exercise
0. Start Fluent 6.3.26
Choose the 2D version, and press Run
1. Read in the case-file
File -> Read -> Case
Save the Case file with a new name (for example add your initials or name to the file name)
File -> Write -> Case
2. Display the grid
Display -> Grid
Press “Display”
The burner is represented by the grid as shown below.
3. Zoom in to identify the location of the methane and air inlets. Use the middle mouse button to zoom in and out.
Zoom in: While holding down the middle mouse button, draw a box (upper left corner to lower right corner) around the location where you want to zoom in.
Zoom out: While holding down the middle mouse button, draw a box (lower right corner toupper left corner).
4. Make sure your model will calculate combustion in a turbulent diffusion flame.
Define -> Models -> Species -> Transport & Reaction
Turbulence-Chemistry Interactions:
The reaction rate will be limited by the kinetic rate or by the turbulent dissipation rate; whichever is the slower.
5. Make sure your model includes the species involved in the chemical reactions.
Define -> Materials - > Mixture Species (press “Edit”)
Press “Cancel”
The following reactions are included
1: CH4 + 0.5 O2 → CO + 2 H2
2: CH4 + H2O → CO + 3 H2
3: H2 + 0.5 O2 → H2O
4: CO + H2O → CO2 + H2
5: CO2 + H2 → CO + H2O
These are defined in the reactions panel
Define -> Materials -> Reaction (press “Edit”)
To scroll through the reactions, change the ID number in the Reactions panel
Press “Cancel” in the Reaction panel and “Close” in the Materials panel.
6. Set the boundary conditions for methane and combustion air.
Define -> Boundary Conditions
Choose “inlet-ch4” and press “Set”
Define the boundary conditions for methane
Press “OK” in the Mass-Flow Inlet panel
Set boundary conditions for air (“inlet-air-1”)
Press “OK” in the Mass-Flow Inlet panel
Set an energy source term to the fluid:
Choose “Fluid” in the Boundary Conditions panel and press “Set”.
In the Fluid panel choose Source Terms and Press “Edit” to define the Energy source
The volume of the fluid is 527779.9m3 and the energy released from combustion of the methane (0.2 kg/s) is 10MW.
This energy is extracted from the gas, i.e., a negative source term in the fluid:
-10 000 000 W / 527779.9 m3 = -18.947 W/ m3
Press “OK” to close the Boundary Conditions panel
7. Define a surface monitor to display the flue gas O2 content at the outlet
Solve -> Monitors -> Surface
Press “Define”
Press “OK” in the Define Surface Monitor panel
Press “OK” in the Surface Monitors panel
8. Initialize the solution and save Case&Data
Solve -> Initialize -> Press “Init”
File -> Write -> Case & Data
9. Start the calculation
Solve -> Iterate
Set the number of iterations to 3000
And press “Iterate”
10. While waiting for the results, recap what you are calculating
11. When the simulation is finished (after 3000 iterations), save case and data
File -> Write -> Case & Data…
Some tips to help you analyze your results
12. The simulation result can be viewed in different ways.
Contours:Display -> Contours…
XY-plots:Plot -> XY Plot
The solution data (XY Plot) can also be exported (option Write to File).
The exported files can be combined for example in Excel.
13. Pathlines can be used in obtaining an estimate of the timeit takes for CH4 to be oxidized fully.
Display -> Path Lines
A pathline gives the time it takes for the gas to flow to different positions along the symmetry axis. The pathline is calculated by tracking the flight of a massless particle (=will follow the gas flow) as function of time.
To estimate the time it takes for CH4 to be completely oxidized, compare the XY-plot showing CO mole fraction and pathline XY-plot.
14. Species mole fractions at the outlet
Report -> Surface Integrals
15. Net reaction rate – determined by chemical kinetics or turbulent mixing?
Have a look at this for at least one reaction, for example using xy-plots along the symmetry axis.
Plot the different “Reaction Rates” predicted by the model to see which determines the net rate of reaction. See below for description of the different rates.
“Rate Of Reaction” is the rate at which the reaction occurs = net rate
“Arrhenius Rate of Reaction” is the rate if it is calculated using rate equations defined in the “Reactions” panel (see above); species concentrations used in the equations are the concentrations in each cell.
“Turbulent Rate of Reaction” is the rate determined by turbulent dissipation predicted by the turbulence model, i.e., the rate at which mixing occurs on a molecular level.
In the model (Turbulence-Chemistry Interaction Model: Finite Rate/Eddy-Dissipation) the net rate is determined by the slower of the Arrhenius rate or the turbulent mixing rate.
What is the implication of your finding – is the turbulence model or the chemical kinetics more decisive for the model predictions?