[mjvanput]

Hi,

I'm trying to simulate a bunsen flame using the following models.

- 2D axisymmetric
- realizable k-epsilon with EWT
- partially premixed combustion model with chemical equilibrium and both oxidizer and fuel at 300 K

The bunsen flame burns at an equivalence ratio of 0.8 with fuel consisting of 80 percent methane and 20 percent hydrogen.

The geometry consists of a pipe with L = 50*D and a surrounding area with L = 100*D and height of circa 20*D. I also modeled the pipe thickness of 1 mm to include possible separation effects at the pipe exit as can be seen in the geometry.

- For the pipe inlet (velocity inlet, 3.95 m/s) I have defined progress variable 0, mixture fraction 1
- For the surrounding inlet (velocity inlet 0.05 m/s) I have defined progress variable 0, mixture fraction 0
- For the farfield I am using a symmetry boundary condition
- For the outlet I am using a pressure outlet with progress variable 1, mixture fraction 0

To initiate ignition I have tried several options, such as patching the progress variable near the pipe outlet with a value of 0.8, patching the progress variable for entire domain with a value of 1 and several combinations. However these options did not result in combustion as can be seen in the temperature contour, the temperature is far too low and the highest temperature occurs at a position far away from the pipe exit.

I also tried patching the temperature, but for some reason I am unable to patch the temperature when using the partially premixed model. Therefore I'm wondering if someone could explain how I should ignite the mixture and get a steady bunsen flame.

Thanks in advance!

[ztnuaa]

Dear Max,

When using Partially premixed models, I prefer the following workflow:

(1) Calculate without ignition by several hundreds steps. The steps depends on your case, but the principle or purpose is to get a good result of mixture fraction.
(2) Patch the entire domain with progress variable 1.0
(3) Continue calculation until it's converged.

Here is an example (the model is derived from your description.there may be some simplification). I don't know the correct result of your case, but I think this can illustrate my point.

[ajithnair]

which software is this discussion on. Is it openFoam?


Sorry Just now saw the tags.

[kjde21]

Originally Posted by ztnuaa

Dear Max,

When using Partially premixed models, I prefer the following workflow:

(1) Calculate without ignition by several hundreds steps. The steps depends on your case, but the principle or purpose is to get a good result of mixture fraction.
(2) Patch the entire domain with progress variable 1.0
(3) Continue calculation until it's converged.

Here is an example (the model is derived from your description.there may be some simplification). I don't know the correct result of your case, but I think this can illustrate my point.

How do you make the calculation without ingnition?
I want to know if you put the progress variable 1.0 in the outlet or you patch the entire model with progress variable 1.0. Thank u for your response.