I focus on the propagation of premixed methane-air mixture flame in a closed tube. I consider the flow is laminar and use the FRC model for combustion. the tube,50*8cm,2D. I set the number of grid to 250*40. With Popi's help, I specified a energy source point to ignite the mixture. It worked. But the temp. and velocity I got were unrealistic to have more than 5000K and 100m/s, compared with our experiments. I read the relative posts in this form. Are they the causes as follows? 1) The tube is small, so I need multiple step chemical reactions. I used the the reaction methane air WD1 in my simulation. Is there detailed chemistry kinetics in CFX? How can I import it? 2) I am told the laminer flame is hard to simulate. I need to have the right transport properties implemented for viscosity, thermal conductivity and species diffusivity. Is this true? Where can I find these parameters? 3) or some other reasons? mesh quality.......... Thanks in advance!- Harry

nobody konws?

Hi,

It is not a simple question. Talk to CFX support for some pointers, but to get it accurate will be a significant amount of work.

Glenn Horrocks

I know it is difficult to get an accurate result in numerical simulation. But the result now I got is ridiculous. As we konw, the temp. of methane flame is around 2000K (my result:>5000K).

What I am eager to know is some methods to improve my results, just getting better. (model, chemical reactions.....)

Btw: in the tutorials-combustion and radiation in a can chamber, how can the reactions happen, without ignition?

Any suggestion will be appreciated!

Did you the standards, time step and grid refinement study? Check that in anyway as numerical difficulties may be the reason.

Try recalculate the temperature in the spark by thermodynamics? How large is the volume you release your energy? If its very small 5000K may be the correct result.

I creat an expert parameter "stiff chemistry", which is suggested in the forum to avoid the the high temperature. But then I find the flame propagates slowly in my simulation, compared with my experiments. It seems the flame is just around the ignition point, not from one end to another. How could it happen!!??

Now I am using a souce point to release the energy. I consider whether I can define an area to patch high temperature for igntion. Maybe it is more suitable. But I dont know how to make it. If someone knows ,plz tell me. Thank you!

Source point for energy? I would expect that the max temperature depends on the volume of the element where you placed your point. Try grid refinement! Try time step refinement!

Chemical reaction mechanisms are an art. What do you use?

Combustion tutorial: In steady flames no ignition is required. Initiation of flame propagation is setted via initialisation of products. Look at the the source terms of the EDM and BVM in order to understand, what drives the modelled flame propagation.

Dear Harry,

Did the solution converge? If it did, the maximum flame temperature is just a function of properties, and fuel/air ratio.

Where did you get the information that air-methane flame temperature is about 2000 [K].. The last time I recall, the flame temperature is a function of composition of the reactants, and products.

Would you mind elaborating about your fuel/air ratio? Is it stoichiometric, lean or rich and how much?

I am guessing is stoichiometric; therefore, I am not surprised you got 5000 [K].. You can do a back of the envelope calculation for

CH4 + 2 * 02 --> 2 * H2O + CO2

with stoichiometric fuel/air ratio and getting the properties from your preferred thermodynamics book. As you add N2, the max flame temperature will start to drop.

Hope this helps,

Opaque