CFD Report on cold flow and hot flow

Thermofluids MMP830/Computational Fluid Dynamics code STAR CCM+ Hands-on Tutorial

Flow and Combustion in OEL Bluff-body Burner
PART 2 – COMBUSTING FLOW
Objective:
To develop a CFD model to study the steady flow environment and combustion process around the
Optical Engineering Laboratory bluff-body burner and to compare the simulation results with
experimental test results. The study naturally subdivides into two parts:
(i) Wednesday p.m. session 1 – Cold flow assignment: simulation of the flow of atmospheric air through
the burner
(ii) Friday p.m. session 2 – Hot flow assignment: simulation of the gaseous combustion of a premixed
air/fuel mixture consisting of air and methane
Geometry:
Use the burner geometry you have developed during session 1.
Exercise 1: Tutorials multi-species gas flows – non-reacting
To prepare you for the assignment you need to work through Star-CCM+ tutorials that illustrate the
set-up for multi-component gas flows. First, we consider mixing of two gas species:
? Use the Internet Explorer to go to the MMP830 module content on the Learn server
? Go to the item headed Introduction to CFD – Star CCM+
? Click the link to Species Mixing (cold flow) Star-CCM+ Tutorial. This connects you to a *.zip file. Save
the contents of the zip-file.
? The file SpeciesMixingSteadyFlow.pdf contains the tutorial instructions
? Open a new session of Star CCM+ and use File ? Import to load the geometry and mesh from the
dilPipe.ccm file.
? Follow the instructions in the *.pdf file, taking particular note of the Model selections in Physics 1 to
invoke Multi-component Gas flow capability and the relevant Gas Component definitions and Species
Mass Fraction boundary conditions and initial conditions to describe the gas composition and the inlet
conditions.
Exercise 2: Set up Star CCM+ burner simulation model – Mixing Flow
? Load your simulation of the OEL experimental burner into Star CCM+ (e.g filename Burner.sim)
? Use File ? Save As to SAVE THE FILE UNDER A NEW NAME: e.g. Burner-SpeciesMixing.sim
? Continua ? Physics 1 ? Models: Set up the correct Physics 1 Models following the method in the
Species Mixing tutorial Exercise 1: De-select Gas ? Segregated Flow ? Ideal Gas; Select
Multicomponent Gas ? Non-reacting ? etc.
? Continua ? Physics 1 ? Models ? Multicomponent Gas ? Gas Mixture: Set up the correct
Material Properties for air and methane following the method in the Species Mixing tutorial Exercise 1
? Continua ? Physics 1 ? Initial Conditions: Set up the correct Initial Conditions following the
method in the Species Mixing tutorial Exercise 1: throughout the CFD domain set Air mass fraction to
1.0 and the CH4 mass fraction 0.0: enter [1.0,0.0] in the Properties dialog box (left bottom)
? Setting Boundary conditions:
(i) Regions ? Body1 ? Boundaries ? FuelInlet:
(a) set the Velocity Magnitude to a value that corresponds to the flow rate of fuel measured on the OEL
experimental burner (in previous years this has been a total volume flow rate round 36.5 litres/min)
(b) set the Species Mass Fraction boundary condition of the Fuel Inlet to correspond to the fuel
composition during your experiment (in previous years the volumetric flowrates were approx. 4.5
litres/min of methane and 32 litres/min of air; please note that you need to use the molar mass of
methane and air to work out the mass fractions from the volume fractions; you should enter the
resulting values of Mass Fractions of CH4, O2 and N2, making sure that the sum of the entered values
is exactly 1.0.). If you want to make a quick start you can study the mixing of pure methane fuel by
setting Species Mass Fraction Properties dialog box set Species Mass Fraction to [1.0,0.0,0.0,0.0,0.0],
so that the only non-zero mass fraction is that of CH4, which is 1.0.
(ii) Regions ? Body1 ? Boundaries ? Air Inlet: set Species Mass Fraction to
[0.0, 0.233, 0.0, 0.0, 0.767], so that O2 mass fraction is 0.233 and N2 mass fraction is 0.767.
(iii) Repeat setting (ii) for Regions ? Body1 ? Boundaries ? Outlet
(iv) Other boundary conditions are unchanged
? Scenes: Generate a Scalar Scene to visualise the progress of the mixing process as the CFD solution
progresses by choosing Scalar ? Field ? CH4 Mass fraction. In Scalar Scene ? Displayers ?
Scalar1 ? Double click Parts. In the Edit pop-up menu click Select All and click Apply to visualise the
scalar field on all Parts
? Also regenerate some of the other Scenes including a Vector Scene and the residuals plot to track the
progress of your simulation; consider using existing Derived Parts or develop new ones to capture
interesting aspects of the mixing field; select parts to be displayed for the most effective visualisation.
? Save your simulation setup
? Before running we clear the old results: on the top menu bar select Simulation ? Clear Simulation.
? Run: Next click initialise button . To run press Run button . A new residual progress window
will appear and the program will start to run. If you bring up the CH4 Mass Fraction scalar window you
will see how the field is initialised to blue and gradually the fuel inlet boundary condition of CH4 mass
fraction 1 will cause a red area to enter at the bottom of the fuel inlet. As the iterations progress this red
fuel-rich zone will reach the gap between the outer bluff body and the conical top of the inner bluff body.
From there this will start to mix with the surrounding air and you should see a jet of air-fuel mixture
propagate upwards through the domain as the iterations progress. The Scalar and Vector windows will
show how the values are updated with the progress of the run. Run until the residuals have decreased
to values below 0.001 (convergence) and the changes in the scalar and velocity vector field have
settled down.
? Save your simulation results
? Visualise final results using other appropriate plots and animations
? Revised Fuel Inlet Species Mass Fraction: in preparation for the combusting flow model you should rerun
this model with revised species mass fraction boundary condition at the fuel inlet corresponding to
the correct fuel composition in the OEL lab ( Re-run the simulation and inspect the results.
Exercise 3: Tutorials combusting gas flows
Next, you need to work through Star-CCM+ tutorials that illustrate the set-up for reacting gas flows:
Use the Internet Explorer to go to the MMP800 module content on the Learn server
? Go to the item headed Introduction to CFD – Star CCM+
? Click the link to Combusting Burner (hot flow) Star-CCM+ Tutorial. Save the contents of the zip-file.
? The file Combustion-EBUTutorial.pdf contains the tutorial instructions
? Open a new session of Star CCM+ and use File ? Import to load the geometry and mesh from the
combustor.ccm file.
? Follow the instructions in the pdf file, taking particular note of the Model selections in Physics 1 to
invoke Multi-component Reacting Gas flow capability and the relevant Gas Component definitions,
Reaction definitions as well as instructions relating to Species Mass Fraction boundary conditions &
initial conditions and also Discretisation Order (1st order) and use of Underrelaxation Factors in the
Solver Controls.
? Run the simulation and Save the results and inspect your visualisations
? When you feel that you have learnt a sufficient amount about the necessary entries for Save and Exit
Star CCM+
Exercise 4: Set up Star CCM+ burner simulation model – Hot Flow
? Load your FIRST SIMULATION of the OEL experimental burner into Star CCM+ (prepared in
Session 1 on Wednesday e.g filename Burner.sim). DO NOT RELOAD THE SIMULATION you
made in Exercise 2 above.
? Use File ? Save As to SAVE THE FILE UNDER A NEW NAME: e.g. Burner-HotFlow.sim
? Physics 1 ? Models ? Multicomponent Gas ? Reacting Flow…: In Physics 1 unclick Ideal Gas ?
Segregated Temperature ? Segregated Flow ? Gas and select Multicomponent Gas ? Reacting
Flow ? Non-Premixed Combustion ? Eddy Breakup Combustion model ? Segregated Flow.
? Physics 1 ? Models ? Turbulence Model: The default turbulence model is the Realizable k-epsilon
model. You may try this one, but the combustion tutorial suggested the standard k-epsilon model. To
change the selection unclick Two-Layer All y+ Wall Treatment ? Realizable k-epsilon Two layer and
Click instead Standard k-epsilon Model (the latter will appear greyed out along with High y+ Wall
Treatment in the Models pop-up menu)
? Physics 1 ? Models ? Multicomponent Gas ? Gas Mixture: Make other Physics 1 selections
following the method of the above tutorial Exercise 2 Mixing Flow but now defining five gas species to
represent the simplest reaction scheme for complete combustion of methane: CH4 (methane), O2
(oxygen), CO2 (carbondioxide), H2O (water) and nitrogen (N2) (defined in that order or re-ordered
later to ensure that CH4 is the first species, O2 the second species and N2 the last species on the list)
? Physics 1 ? Models ? Reacting ? Reactions: Write down the reaction scheme for stoichiometric
combustion of CH4 and O2 to find the correct entries for the stoichiometric coefficients. Treat N2 as an
inert gas.
? Physics 1 ? Reference Values & Initial Conditions: Make other Physics 1 selections following the
method of the above tutorial Combusting Gas Flow; set Physics 1 ? Initial Conditions set the O2 mass
fraction to 0.233 (check that 21% by volume corresponds to 23% by mass) and the N2 mass fraction
0.767 and leave the other mass fractions 0.0. First set the initial temperature to 300K everywhere and
run the simulation. It is possible that your final simulations do not show evidence of combustion
reaction (i.e. temperatures doggedly remain around 300K and CO2 and H2O mass fractions are equal
to zero). In this case the initial temperature can be increased to provide the necessary ignition source.
Re-run the simulations with higher initial temperature to check that combustion happens.
? Setting Boundary conditions:
(i) Regions ? Body1 ? Boundaries ? FuelInlet:
(a) set the Velocity Magnitude to a value that corresponds to the flow rate of fuel measured on the OEL
experimental burner (in previous years this has been a total volume flow rate round 36.5 litres/min)
(b) set the Species Mass Fraction boundary condition of the Fuel Inlet to correspond to the fuel
composition during your experiment (in previous years the volumetric flowrates were approx. 4.5
litres/min of methane and 32 litres/min of air; please note that you need to use the molar mass of
methane and air to work out the mass fractions from the volume fractions; you should enter the
resulting values of Mass Fractions of CH4, O2 and N2, making sure that the sum of the entered values
is exactly 1.0.). As a shortcut you can use pure methane fuel by setting Species Mass Fraction
Properties dialog box set Species Mass Fraction to [1.0, 0.0, 0.0, 0.0, 0.0], so that the only non-zero
mass fraction is that of CH4, which is 1.0. However, you need to bear in mind that the amount of fuel
will now be more than 10 times as large as the OEL burner, so the resulting flame characteristics will
be very different from the experiment and you will need to re-run the simulations later with the correct
fuel composition.
(ii) Regions ? Body1 ? Boundaries ? Air Inlet: set Species Mass Fraction to
[0.0,0.233,0.0,0.0,0.767], so that O2 mass fraction is 0.233 and N2 mass fraction is 0.767.
(iii) Repeat setting (ii) for Regions ? Body1 ? Boundaries ? Outlet
(iv) Other boundary conditions are unchanged
? Scenes: Generate a Scalar Scene to visualise the progress of the mixing process as the CFD solution
progresses by choosing Scalar ? Field ? CH4 Mass fraction. In Scalar Scene ? Displayers ?
Scalar1 ? Double click Parts. In the Edit pop-up menu click Select All and click Apply to visualise the
scalar field on all Parts
? Also regenerate some of the other Scenes including a Vector Scene and the residuals plot to track the
progress of your simulation; consider using existing Derived Parts or develop new ones to capture
interesting aspects of the mixing field; select parts to be displayed for the most effective visualisation.
? Save your simulation
? Before running we clear the old results: on the top menu bar select Simulation ? Clear Simulation.
? Run: Next click initialise button . To run press Run button . A new residual progress window
will appear and the program will start to run. If you bring up the temperature scalar window and you will
see how the field is initialised to blue and reaction starts where the fuel is in contact with air (near the
fuel inlet). The combustion model is very “robust” and reaction does not need an artificial ignition
source (as would be necessary in real life!). The red (hot) flame front propagates upwards as the fuelrich
inlet flow reaches the gap between the outer bluff body and the conical top of the inner bluff body.
From there you will see the development of the characteristic flame pattern as the iterations progress.
The Scalar and Vector windows will show how the values are updated with the progress of the run. If
you have set up the run correctly – with a good mesh and 1st-order (default) discretisation for all
convection terms – the residuals decrease rapidly to values well below 0.001. If this cannot be achieved,
run until the residuals have decreased to values below 0.01 convergence and the scalar field has
settled down. Visualise results using other appropriate plots and animations.
? Save your simulation results
Assignment tasks:
1. Complete the above work and modify the simulation setup to include the correct details of the
combusting burner experiment in the OEL. If you have used pure methane as a fuel as a start you will
need to re-run with the correct fuel mixture and possibly revised initial conditions to achieve ignition
later! You should use your own experimental data (guidance: the volumetric flow rate of methane in
the fuel supply was around 4.5 litres/min and the volumetric flow rate of the air in the fuel supply was
around 32 litres/min in previous years).
2. Iterate until you obtain a converged solution (this can take 1000-2000 iterations for a complex flow like
this) and save your converged solutions (NB: each CFD file is large, so you need a very substantial
filestore for this; you may need to make space on your U:\ drive or use a mobile Hard Disk drive to
store your simulations).
3. Reduce the maximum residuals levels from 0.001 to 0.0001 and continue the iteration process.
4. Study the two converged solutions to satisfy yourself that all key features of the solution are
independent of the setting of the residuals level.
5. Refine the mesh and consider alternative boundary conditions (if appropriate) and repeat steps 1-4.
Compare the converged solutions on the initial mesh with those of the finer mesh to satisfy yourself
that all key features of the solution are substantially independent of the mesh.
6. When you are satisfied with your solution, write a summary report that includes the following sections:
Introduction (description of problem), Simulation Setup (table of simulation inputs and details of mesh),
Results (produce figures from saved Scalar and Vector Scenes capturing the main features of
simulation results) along with descriptions of these results.
7. The final and most important element of the assignment is to compare and contrast the CFD simulation
results with the experimental data you obtained in the Optical Engineering Lab (OEL). Add your
comments on the validity of your output (for this you may want to include consideration of the possible
effects of uncertainties in the experimental data and in the CFD simulation results).