CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Laminar combustion in porous media (

Mashihu May 8, 2003 09:09

Laminar combustion in porous media
I am simulating the process of the laminar combustion through porous media zones! Methane and air are premixed perfectly.Then the gas flowed into a duct(which is filled with porous media). The porosity of the porous media is 0.9. The gas will combust in the duct. I use the laminar finite-rate model and patch a high temperature zone in the duct. to improve the convergence,I use the command: (rpsetvar'stiff-chem-seg? #t) and (models-changed). Then I specify time step as 0.001s and begin to iterate. Just after several iterations,reverse flows occurs in the out-let.stopping iteration,I found the temperature of the most area in the dust is nearly I change the time step to be 0.00000001s.but the result is unchanged.

cg May 8, 2003 16:52

Re: Laminar combustion in porous media
Try the coupled implicit solver.

Mashihu May 8, 2003 20:33

Re: Laminar combustion in porous media

I must use segerated solver.Because I need modify porous media.In fluent,the energy equation is only one,but I need seperate it into two:one for gas and one for porous media. If I use the coupled solver,I can't modify the model.Currently,I just want to get a result with the model provide by fluent to examine the result of my experiment.

The segerated solver maybe solve the problem for I have a piece of material which states a example about laminar combustion that is solved with segerated solver.It use the two commands: (rpsetvar'stiff-chem-seg? #t) and (models-changed). The material is as follow: Segregated solver 1. Fractional time stepping: over a time step Dt Advance solution with no chemical source terms(only convection and diffusion) for Dt. 2.Then, advance chemistry in each cell for Dt as a constant pressure reactor where the chemical source term S = wk*Wk / r, wk is the reaction rate, Wk is the molecular weight, and r is the density. 3. Chemistry integrated with stiff ODE solver CVODE. 4. Requires unsteady solution, even for steady state! 5. Final solution depends on time step! 6. Hence, only use for unsteady reacting flows 7. Fractional step scheme is first order accurate in time 8. Hidden from gui/tui: activate with scheme commands¡* (rpsetvar ¡®stiff-chem-seg? #t) (models-changed) The exemple is as follow: Single, driven piston compresses hydrogen-oxygen-argon mixture which ignites due to heat of compression H2/O2/Ar 8 reacting species, 19 step mechanism Moving mesh, segregated solver, fractional step stiff chemistry solver

cg May 10, 2003 09:23

Re: Laminar combustion in porous media
Several things can cause the chemistry to diverge Some chemical mechanisms are very stiff and the ODE solver will diverge. You should test your chemistry: I model a plug flow reactor (PFR) with a 1 cell Fluent case run for a specified time step. The results from Fluent should match the PFR solution from another code, such as Senkin.

Mashihu May 10, 2003 20:08

Re: Laminar combustion in porous media
Hi,cg. The chemical mechnism maybe is wrong. I simulated laminar combustion in a duct without porous media. I found the reaction only can propageate downstream. Initializing the flow field with x-velocity and y-velocity as 0,patching a high temperature regin in the midst of the duct, and seting up the inlet velocity as 0, I found the reaction proceeding just in the downstream, While the upstream temperature is 300K all along. The is impossible.

All times are GMT -4. The time now is 13:28.