CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Pre-Processing (http://www.cfd-online.com/Forums/openfoam-pre-processing/)
-   -   Simulating diffusion flames with fireFoam (http://www.cfd-online.com/Forums/openfoam-pre-processing/112952-simulating-diffusion-flames-firefoam.html)

babakflame February 8, 2013 10:38

Simulating diffusion flames with fireFoam
 
Dear Buddies
Hi
I want to simulate a diffusion flame with fireFoam. Does any body has any tutorial about this issue?
I would be very pleased if an experienced buddy help me on this object.


Thanks
bobi

babakflame February 9, 2013 04:25

more information from my work
 
Dear Buddies
Hi
I have simulated diffusion flames with reactingFoam earlier. However now I want to simulate them with firefoams using FSD combustion model. I have looked into firefoam tutorials about poolfires. when I start my simulation, I got this error message:

Create time

Create mesh for time = 0.0522381


Reading g
Creating combustion model

Selecting combustion model FSD<psiCombustionModel,gasThermoPhysics>
Selecting thermodynamics package hsPsiMixtureThermo<reactingMixture<gasThermoPhysic s>>
Selecting chemistryReader foamChemistryReader
terminate called after throwing an instance of 'std::bad_cast'
what(): std::bad_cast
Aborted

I even deleted the chemistry properties file, made the chemistry solver off,modified the reactions file and the combustion model, Nevertheless I still get this error. I would be very please if an experienced fireFoam man help me.:)
Regards
Bobi

Matze7 April 3, 2013 04:35

Hi Bobi,

the combustion models have changed in OpenFoam 2.2.0. I don't know why, but some of the combustion models, which worked fine on 2.1.1, don't work on 2.2.0.

you can try to replace

combustion model FSD<psiCombustionModel,gasThermoPhysics>

with

combustion model FSD<psiThermoCombustion,gasThermoPhysic>

Regards

babakflame April 4, 2013 03:49

Quote:

Originally Posted by Matze7 (Post 418014)
Hi Bobi,

the combustion models have changed in OpenFoam 2.2.0. I don't know why, but some of the combustion models, which worked fine on 2.1.1, don't work on 2.2.0.

you can try to replace

combustion model FSD<psiCombustionModel,gasThermoPhysics>

with

combustion model FSD<psiThermoCombustion,gasThermoPhysic>

Regards

Hi Matze

Thanks for ur reply. now, I have a question about firefoam in 2.2.x. Does supprot multistep combustion? :confused:
As u know in 2.1.x firefoam only had singlestep combustion.

Regards
Bobi

Matze7 April 5, 2013 06:11

Hey Bobi,

I've no idea if fireFoam can handle that, sorry. I'm new at simulating with fireFoam and so my experience with it is quite little. I'm still struggling to get my own simulation running :(.

Regards.

babakflame April 13, 2013 06:03

Hi matze
I I have a doubt about 0 folder. In this folder we have to create distinct folder for any reactant. however, I want to know the value for each specie in boundary field is in mole or mass fraction?

Matze7 April 17, 2013 07:41

Hi Bobi,

in my oppinion just mole fraction makes sense.

babakflame May 20, 2013 06:42

Dear Buddies
Hi

I have simulated a cold flow with fireFoam. However, when I tried to ignite combustion by increasing the whole computational domain temperature to 1600 K and then injecting the reactants; I do not have the combustion yet.I think I have set out the chemistry, combustion and reaction files, correctly (according to its tutorials in openFoam). Would somebody PLZ help me on this? I can put any of my files that is needed for investigating my case. However, I should mention that I am trying to simulate one of Sandia flames with fireFoam and with FSD (Flame Surface Density) combustion model.

Tobi May 22, 2013 07:20

Hi,

is your Diffusion flame laminar or turbulent?

Maybe you can use the flamelet-model tool :) (adverticement) :D

But I think bobi could give you a more accurate answer which solver you should use (:


PS: for turbulent flames it s possible to use the flamelet-model 2.2.x from my Homepage and for laminar flows, alberto cuoci build a new solver.

Tobi

babakflame May 22, 2013 07:38

Hi Tobi :)
I am trying to simulate DLR flame (Sandia CH4-H2-N2) with fireFoam (FSD combustion model) which is a single-step flamelet/progress variable combustion model with LES.
Ofcourse I would be so much happy :D to link the flow solver of fireFoam (LES) with your flamelet code :rolleyes:; However I do not know that how can I link these together:(:confused::(
can you hint me on this issue?:rolleyes:;):confused:
Regards
Bobi

Tobi May 22, 2013 08:20

Hi bobi,

you are still working on flamelet-model RAS reprogramming to LES?

Well as in the libOpenSmoke thread I am not familiar with LES modelling and there for I can not give you a advice or any hints to Combine both.

*sorry*

Do you have to simulate with LES ? RANS flamelet model and DLR flames are working very good.

babakflame May 22, 2013 10:10

Hi Tobi
;)

Unfortunately, As part of my thesis I should combine LES with steady flamelet.
I have succeeded this job in two different route.

First, I am trying fireFoam; Although with some progress there is still problem in igniting combustion.

Second, I am trying your code. At first I was thinking that I should change the simple algorithm to pimple plus changing the stoichiometric chi update formulation + turbulence model. After contacting Dr. Cuoci, I figured out That I just need to change the stoichiometric chi update formula plus changing turbulence model.
I am trying both route, and I hope to become successful:):rolleyes::)
Regards
Bobi

Tobi May 23, 2013 07:11

Wow - Alberto answered you :D

Hmmm so you are sucessful in the LES with Albertos flamelet libraries ?

babakflame May 23, 2013 07:34

Next Week; I will try again to build LES-Flamelet code :)

I will tell you the result of the code (compiled or not). I hope to succeed ; cause in the last try my problem was with pimple algorithm;)
Regards
Bobi

TBO August 27, 2013 11:01

Janaf Thermo crash
 
All,

As proposed by Bobi on the libOpenSmokeForum I am trying to use fireFoam to model the turbulent mixing of methane jets in a swirling air flow (without combustion involved). However I get an error in a simplified model and I don't know what causes this error. The error is the following:

--> FOAM Warning :
From function janafThermo<EquationOfState>::limit(const scalar T) const
in file /home/.../OpenFOAM/OpenFOAM-2.1.1/src/thermophysicalModels/specie/lnInclude/janafThermoI.H at line 108
attempt to use janafThermo<EquationOfState> out of temperature range 200 -> 5000; T = -4667.51


It starts with some small deviations from the required temperature range but then the deviation starts to grow and finally the model crashes. I use the following boundary conditions:

internalField uniform 500;

boundaryField
{
outlet
{
type zeroGradient;
}
burnerwall
{
type zeroGradient;
}
inletair
{
type fixedValue;
value uniform 500;
}
inletfuel
{
type fixedValue;
value uniform 500;
}

}


So using this I assume that temperature field normally should be around 500 K and stay around this temperature (since combustion is off). Does anybody know what to do with this?

Thank you

babakflame August 28, 2013 03:02

Hi TBO

I hope to provide helpful suggestions for you.
1) first probability is your CFL number. Its maximum must not exceed about 0.4
2) check your grid quality. Sometimes grid quality at downstream causes this error.
3) Check your maximum DeltaT. It must not exceed 2e-06
4) your boundary conditions does not have impact on this error.Although I suggest to try another condition for burner wall(no slip condition)

Regards
Bobi

TBO September 10, 2013 03:40

Thank you for your answer Bobi, I implemented the changes that you proposed. However the problem now is that DeltaT gets very small (1e-12), the result is that the calculation will take extremely long (actually impossibly long, it run for a couple of days on 12 processors and I reached time 0.0001, while I need to go to time 0.3 for a complete throughflow of my domain). This is already the case for a very simplified mesh of 60000 cells, I put a lot of effort in optimizing the mesh, for computional effort I already moved to a cyclic mesh.

How can I speed up the calculation and increase the timestep. In the end, I'm not interested in the transient behaviour but only in the steady state result. I am using fireFoam in OF 2.1.1 without chemistry, to model the mixing of a methane jet in air.

Regards,

babakflame September 13, 2013 11:32

Hi TBO

Sorry for late answering, I was a bit busy last days.
My best suggestion for your case is this:

Be careful that there is a coupling between maximal courant number and time step. If you want to run a LES case, you can adjust your courant number in a way that your maximum time step do not exceed 1e-05. Try raising your CFL number as I mentioned, It speeds your running time. I am sure that you can reach to a acceptable result after a few days.

Monitor your maximum time step in a log file to be confident about it. never more than 1e-05.
use this directive for monitoring your results in a log file

Code:

  mpirun -np 12 fireFoam -parallel > log


Bobi

babakflame October 9, 2013 06:19

Dear Buddies

Does anybody has ever tried to use firefoam code with a mixture of fuel instead of a single fuel. for instance CH4+H2 or C3H8+H2?

If there is any try with successful results, I would be very pleased to be informed. Cause I have some questions to ask
?

regards
Bobi

vishal_s September 16, 2014 19:04

Hello Bobi,

I am sorry.. I don't have much experience in C++ or openfoam. What is "chi"??


All times are GMT -4. The time now is 23:58.