You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 67d0f94
Browse filesBrowse the repository at this point in the historyBrowse files
docs: fix tutorial links and unknown option flagged by site check
Use ../../ for the Inc_Laminar_Diffusion_Flame images, drop the stray path
prefix from the Inc_Lid_Driven_Cavity resource link, and remove the
nonexistent RELAXATION_FACTOR_MOMENTUM option from its config snippet.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
*Figure (1): 2D axisymmetric configuration of the laminar counterflow diffusion flame. The axisymmetry is shown here vertically to align the simulation results with the usual lab orientation, but in the simulations, the axisymmetry axis is always horizontally at y=0.*
20
20
21
21
In this tutorial we show how to setup a laminar diffusion flame computation for a counterflow flame. We will use a flamelet method to model combustion, which is a tabulated chemistry approach. The table was precomputed from 1D Cantera solutions using the SU2_DataMiner code. We then reproduce the experiments of C.K. Law, mentioned in his book *Combustion Physics* (2006).
*Figure (2): Counterflow Diffusion burner from the Thomson Lab in Toronto: https://thomsonlab.mie.utoronto.ca/counterflow-diffusion-burner/.*
24
24
## Resources
25
25
@@ -31,7 +31,7 @@ The resources for this tutorial can be found in the [incompressible_flow/Inc_Lam
31
31
32
32
The mesh is created using [gmsh](https://gmsh.info/) and a respective `.geo` script is available to recreate/modify the mesh [counterflow.geo](https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Laminar_Diffusion_Flame/counterflow.geo). The mesh has quadrilateral cells, with 8324 elements and 8551 points. This very coarse mesh is sufficient to capture accurately resolve the flame for this testcase.
33
33
34
-
34
+
35
35
*Figure (3): Computational mesh with color indication of the used boundary conditions.*
36
36
37
37
@@ -123,12 +123,12 @@ The lookup table was generated using the SU2_DataMiner package, which can be fou
We will not go into the table generation now, but merely provide the results. For the constant strain diffusion flamelets, we fix the strain rate and run a series of counterflow flames for different enthalpies (we vary the inlet temperature). The resulting solutions for temperature are shown below:
*Figure (4): The 1D cantera solutions for the counterflow flames, using a variation in inlet temperature.*
128
128
129
129
We then store the 1D solutions as a 2D field with mixture fraction and enthalpy as the independent variables. The result for temperature and heat release rate is shown below:
*Figure (5): A 2D surface representation of the FGM, showing the temperature and the heat release rate as function of mixture fraction and enthalpy. Note that our enthalpy range is not very large. We can represent constant strain flames with different inlet temperatures, but for simulations with large heat losses the table has to be extended to lower enthalpies.*
133
133
134
134
Note that the mixture fraction equation is a diffusion equation and does not contain any source terms, unlike the progress variable equation for premixed flames.
*Figure (6): Residual plot showing convergence of the counterflow diffusion flame.*
154
154
155
155
The residuals for pressure and velocity have dropped by 6 orders of magnitude, which is sufficient for this testcase.
156
156
157
157
A contour plot of the temperature field in the domain is shown below. Note that only 1/3 of the right vertical edge is an outlet. The smaller outlet will force the flow to accelerate close to the outlet. This prevents backflow, which deteriorates convergence.
*Figure (7): Result visualized using paraview. Shown is the temperature field and the streamlines colored by mixture fraction.*
161
161
162
162
In the figure we have also indicated the isocontours of mixture fraction for values Z=0.4 (blue), Z=0.5 (dark purple) and Z=0.6 (red), showing that the maximum temperature occurs at $Z<0.4$. The flame looks very diffusive close to the outlet because of the blockage in the corners. Here, a low velocity recirculation zone in the top and bottom corner is mixing the fuel and air, causing the diffusive behavior.
163
163
164
164
A figure of the temperature on the symmetry axis is shown below, together with measured results from Sung et al. (1995). These results are also mentioned in the book of C.K. Law, *Combustion Physics*.
*Figure (9): Methane mole fraction as function of distance (from the fuel inlet) on the symmetry axis.*
173
173
174
174
A first sanity check is always to to check that inlet mass fraction or mole fractions are correct. The consumption of $CH_4$ in the flame is also predicted correctly
175
175
176
176
The $CO_2$ mole fraction is zero at both inlets and rises in the reaction zone.
*Figure (10): Mole fraction of CO2 as function of distance (from the fuel inlet) on the symmetry axis.*
179
179
180
180
As for $CH_4$, the $CO_2$ mole fraction in the flame zone is correctly predicted, the maximum value as well as the slope is matching with the experiment.
181
181
182
182
The lookup value for CO is much lower than the experimentally measured value. CO as well as NO are very slow reactions, and are not accurately predicted by flamelet approaches. To get more accurate results for CO or NO, a separate transport equation should be solved for them. Instead of the value for CO, we look up the source term for the transport equation for CO.
*Figure (12): Temperature as function of mixture fraction, using the values taken at the symmetry axis.*
189
189
190
190
This shows the typical Burke-Schumann- like behavior where the diffusion flame can be described by a pre-flame and a post-flame zone with almost linear slope for temperature.
Copy file name to clipboardExpand all lines: _tutorials/incompressible_flow/Inc_Lid_Driven_Cavity/Inc_Lid_Driven_Cavity.md
+1-2Lines changed: 1 addition & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -27,7 +27,7 @@ The intent of this tutorial is to demonstrate the pressure-correction solver (PI
27
27
28
28
## Resources
29
29
30
-
The resources for this tutorial can be found in the [Inc_Lid_Driven_Cavity](incompressible_flow/https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Lid_Driven_Cavity) directory in the [tutorial repository](https://github.com/su2code/Tutorials). You will need the configuration file ([incomp_pb_liddrivencavity.cfg](https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Lid_Driven_Cavity/incomp_pb_liddrivencavity.cfg)) and the mesh file ([mesh_cavity_257x257.su2](https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Lid_Driven_Cavity/mesh_cavity_257x257.su2)).
30
+
The resources for this tutorial can be found in the [Inc_Lid_Driven_Cavity](https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Lid_Driven_Cavity) directory in the [tutorial repository](https://github.com/su2code/Tutorials). You will need the configuration file ([incomp_pb_liddrivencavity.cfg](https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Lid_Driven_Cavity/incomp_pb_liddrivencavity.cfg)) and the mesh file ([mesh_cavity_257x257.su2](https://github.com/su2code/Tutorials/tree/master/incompressible_flow/Inc_Lid_Driven_Cavity/mesh_cavity_257x257.su2)).
31
31
32
32
33
33
## Tutorial
@@ -104,7 +104,6 @@ For a regular PISO algorithm, the pressure and momentum underrelaxation is commo
0 commit comments