[Per_Holmgren]

Hi all.

I have a very straightforward question. Lets say we have a very simple geometry of a one 2 m long tube with a diameter of 10 cm. The tube is entirely open internally.

On one side is the inlet(s), on the other the outlet.
The inlets is a flush side of flow straighteners, with a mix of air and propane entering in a central 5 cm diameter region, this mix is ignited.
The rest of the inlet area from 5 cm to the outer edge (10 cm) only supplies air. The outer tube wall have agiven temperature profile, and has zero mass flux.

As the entire geometry and flows are axi-symmetric, it makes sense to model it as such.

Now, the inner inlet region would have a radius of 25 mm, while the outer inlet region would have a outer radius of 50 mm and an inner of 25 mm.

Through the inner region flows a perfectly mixed air-propane mix of 20 l/min air and 0.75 l/min propane (normal gas).
This would correspond to a mass flux of 0.2153 kg/m2-s.

The outer region supplies air at a rate of 56 l/min, which given the geometry gives a mass flux of 0.19 kg/m2-s.

So we have lots of air (56+20 l/min) and a bit of propane (0.75 l/min) going in to our tube.

I have used the inbuilt 'propane-air-2step' mechanism, the stiff chemistry solver using Finite Rate/No TCI, volumetric reactions.

The mechanism says it reacts 1 C3H8 + 3½ O2 -> 3 CO + 4 H2O, followed by full oxidation of CO to CO2.

If all the propane eventually turns into CO2 and H2O, the outlet gases should consist of ~4% H2O, ~3% CO2, ~77% N2 and ~16% O2 (mole fractions).

My simulation always give me around 16% H2O, 11% CO2, 70% N2 and 0% O2.

All inlets have the same pressure.

And here is where I get stuck.
The outlet have the same mass flow rate as the inlets, the inlet velocities both match what one would expect, yet the chemistry is all wrong.

All the oxygen is consumed even though there is not nearly enough fuel put into the system to achieve that. I get around 4 times the H2O and CO2 mole fractions compared to what I should get.
As the mass flows are correct, and the inlet velocities are correct, all I'm left with is the chemistry.

If I do a simplified composition pdf transport solver I get reasonable values, but I need to be able to add stuff to the mechanism...

Does anyone know what could be wrong? Have I misunderstood something about 'Finite Rate/No TCI'?

Best Regards
/Per

[LuckyTran]

Your reactants (the fuel and air) are not perfectly mixed at the inlet or outlet.

When you say you have too much H2O at the outlet, is this locally or averaged over the entire outlet?

Originally Posted by Per_Holmgren

If all the propane eventually turns into CO2 and H2O, the outlet gases should consist of ~4% H2O, ~3% CO2, ~77% N2 and ~16% O2 (mole fractions).

Eventually is also assuming that the burnt products are thoroughly mixed.

Originally Posted by Per_Holmgren

All the oxygen is consumed even though there is not nearly enough fuel put into the system to achieve that. I get around 4 times the H2O and CO2 mole fractions compared to what I should get.
As the mass flows are correct, and the inlet velocities are correct, all I'm left with is the chemistry.

There is not enough fuel in the overall sense, but locally where the reactions take place, it is a different story.

[Per_Holmgren]

Thanks for the reply, much appreciated!

Originally Posted by LuckyTran

Your reactants (the fuel and air) are not perfectly mixed at the inlet or outlet.

Well, kinda. The outer ring/inlet is perfectly mixed air. The inner inlet (or burner) is perfectly mixed propane/air. Even taking only the propane-air mix inlet into account, H2O should only reach a maximum of about 14% and CO2 10% I.e. less than the total average that exits at the outlet).

Originally Posted by LuckyTran

When you say you have too much H2O at the outlet, is this locally or averaged over the entire outlet?


Eventually is also assuming that the burnt products are thoroughly mixed.

There is not enough fuel in the overall sense, but locally where the reactions take place, it is a different story.

That's averaged over the entire outlet. Looking at the radial species distribution at the outlet, there are no streaks of flue gases, just a very nice flat distribution. Though the proportions are all wrong.

However, I think I have this figured out. Maybe.
I tried opening my case on a different PC, but still got the same result. I tried resetting the entire thing (again) and recreated the mechanisms from scratch, and still the same thing.

So I saved my materials, then upgraded to the latest version of Ansys Fluent. Previously had all my saved files on a cloud server, but this time I saved everything locally instead.
I suspect the cloud server might block fluent from accessing some files properly during operation.
Sometimes I got this error message:
GENERAL-CAR-CDR: invalid argument [1]: improper list

Fluent sadly doesn't tell you what list is improper, but I suspect it was my Materials.

Anyway, after all that, I could import my materials and mechanisms, and now everything works! Horray.

Kinda funny, had I not gotten this problem (which I first stated having over 2 years ago), my Ph.D thesis might have looked quite a bit different.

So lesson learned. Don't save your active cases on a cloud server. Maybe?