Learning Forum

[Nick Pester]

Hi, 

I am running a LES combustion simulation. I am running a swirling, non-premixed combustion methane case. The case I am running is a stable flame from the experiment. I have initialized my LES simulation from a converged reacting RANS case (coupled+ Least-Square Cell-Based, Pressure: Standard, First Order Upwind for all other equations (2nd order doesn't show a stable flame). I have tried a lot of things, but the flame keeps dying. I have read a few papers that you guys have published recently and made sure I have the same setup but still same issue. 

Any suggestions of what I could possibly do? 

I am using the 22r2 version, would you recommend to change version and use the 2023r2? (I cant find release notes to see if there are any upgrades to the FGM combustion model). 

 

Thank you for your help !!!

Below are my setup:

Fluent

Version: 3d, dp, pbns, pdf20, LES, transient (3d, double precision, pressure-based, 20 species pdf, large eddy simulation, transient)

Release: 22.2.0

Under-Relaxation Factors

 

Variable Relaxation Factor

-----------------------------------------

Pressure 0.75

Density 0.75

Body Forces 1

Momentum 0.75

Energy 0.89999998

Temperature 1

Progress Variable 0.80000001

Mean Mixture Fraction 1

 

Linear Solver

 

Variable Solver Type Termination Criterion Residual Reduction Tolerance

------------------------------------------------------------------------------------------

Pressure F-Cycle 0.01

X-Momentum Flexible 0.1 0.7

Y-Momentum Flexible 0.1 0.7

Z-Momentum Flexible 0.1 0.7

Energy F-Cycle 0.1

Progress Variable Flexible 0.1 0.7

Mean Mixture Fraction Flexible 0.1 0.7

 

Pressure-Velocity Coupling

 

Parameter Value

-----------------------------

Type SIMPLEC

Skewness Correction 0

 

Discretization Scheme

 

Variable Scheme

----------------------------------------------------

Pressure Second Order

Density Second Order Upwind

Momentum Bounded Central Differencing

Energy Second Order Upwind

Progress Variable Second Order Upwind

Mean Mixture Fraction Second Order Upwind

 

Below is the species model tab: 

[Essence]

Hello,

Try this out. This may help.

Please first complete the turbulence modelling and then move to reacting case. That is, first model URANS (non-reacting) case then switch to non-reacting LES. Once converged, then turn ON the reacting case and then radiation (if you need). Also, while turning ON the reacting case, I would suggest you to first run FR/ED with relax to chemical equilibrium with just one/two step(s) after turning ON the LES. Then using the EDM (relax to chemical equlibrium) as initial solution, switch to FGM and radiation.

[Nick Pester]

Hi, I will try that! Thank you!

[Essence]

You are welcome. Sure, please try it out.