[SinNi]

I am modelling 1-step CH4-air combustion using species transport model with laminar. I patched a temperature of 2500K at one region after cold run, and when i start the calculation again, the temperature straight hit the upper limit of 5000K.

It is a non-premixed combustion (stoichiometric reaction):
air inlet: 0.2134 O2
fuel inlet: 1 CH4
wall: zero heat flux
Stiff Chemistry Solver ON

I tried patching some product at the region, no change.
I tried lower the density under-relaxation factor, no change.

Kindly suggest what parameters should I be looking for.

Thank you.

[vinerm]

The combustion models in Fluent are meant to be used only with turbulent models. If you want to model laminar flow, then use species transport with Finite Rate.

[SinNi]

Hi Vinerm,

Sorry for the late reply.

I am currently using species transport model, with laminar selected as viscous model (Not sure if that is what you mean sorry). My model is a simple model of 2 inlets(1 for fuel and 1 for air) and 1 outlet. It has a mesh of around 15000 cells. I am using the built-in 1 step methane-air mechanism in Fluent.

Boundary conditions:
fuel inlet: 0.0445m/s, 298K, 1 CH4 mass fraction
air inlet: 0.062m/s, 298K, 0.23 O2 mass fraction
pressure outlet: 0
wall: 298K

Patching:
Temperature: 2500K
species: 0.01 H2O and 0.01 CO2

I increase the upper limit of the temperature, and the temperature will drop to around 3000K.

So the issue I am facing now:
1. Temperature is too high ~3000K.
2. The solution will not converge. the residual start to fluctuate after around 2000 iterations.

Any suggestion is appreciated. Thanks.

Regards

[vinerm]

You understood it correct. Species Transport with FR can be used with laminar. However, you should be sure that the flow is certainly laminar. If Re number is high, it will not be able to stabilize without the diffusivity generated due to turbulence. Therefore, check the Re number for the flow and if it is in transition or fully turbulent regime, enable turbulence. You will immediately observe the stability.

[SinNi]

My model is fully laminar as the fuel velocity is very low (0.0445 m/s) and the fuel inlet is 10.8 mm diameter. So I don't think there will be turbulence in my model.

[vinerm]

It's that the velocity is low, however, its not the velocity that decides whether flow is laminar or not. You can have laminar flow at 100 m/s and turbulent flow at 1 mm/s. You need to calculate Re number and then see if the flow could be turbulent or not. If it can be, enable turbulence in Fluent.

If it is not turbulent, then most likely the system has inherent transients or your mesh is very fine. Run a transient case or coarsen the mesh.

[SinNi]

I manage to get the temperature to around 2300K, but I would like to get some clarification before i proceed further.

Based on the paper by Westbrook for 1 step mechanism, the value for activation energy is 48.4kcal/mol which is similar to 2.027e8 J/kmol in Fluent.

However for the pre-exponential factor, it is listed as 2.119e11 in Fluent. In the paper, A is equal to 6.7e12. Is there any conversion required for A?

I can only get the reasonable result if using 6.7e12 as pre-exponential factor. Hence I am not sure if my simulation is correct?

Regards.

[vinerm]

A does not have fixed units; it depends on the order of reaction. For first-order reaction, it is 1/s, i.e., representing collision frequency. For second order reaction, it is L/mol-s. What you are dealing with is most likely first order, hence, no alteration should be required. Do note that values in Fluent may not always be as per the user requirement.