CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Species transport and reaction (https://www.cfd-online.com/Forums/fluent/106985-species-transport-reaction.html)

snh September 14, 2012 03:06

Species transport and reaction
 
Hi

I am trying to simulate a reaction between lithium and sulphur hexafluoride using the eddy dissipation model. I would like to know if I can input specific quantity of reactant..If I want to study the consumption of 5 kg of lithium in the reaction, is there any way where I can enter the amount of reactant used?

Amir September 14, 2012 13:12

Quote:

Originally Posted by snh (Post 381736)
Hi

I am trying to simulate a reaction between lithium and sulphur hexafluoride using the eddy dissipation model. I would like to know if I can input specific quantity of reactant..If I want to study the consumption of 5 kg of lithium in the reaction, is there any way where I can enter the amount of reactant used?

Hi,

Yes it's possible but in different manner.
First, you have to specify the rate of consumption (kg/s) and in a specified time, you can reach the consumed mass (kg).
To set a rate of consumption (kg/s), you can set source term of reactants in cells which are adjacent to the prescribed location where you want to have reaction near them. you can separate these cell and add the source terms in dimension of (kg/m3.s).

Bests,

snh September 15, 2012 01:23

Quote:

Originally Posted by Amir (Post 381834)
Hi,

Yes it's possible but in different manner.
First, you have to specify the rate of consumption (kg/s) and in a specified time, you can reach the consumed mass (kg).
To set a rate of consumption (kg/s), you can set source term of reactants in cells which are adjacent to the prescribed location where you want to have reaction near them. you can separate these cell and add the source terms in dimension of (kg/m3.s).

Bests,

Hi Amir,
Thanks for the reply and your help is appreciated. I will try this out but I have no idea about the time required to consume the 5 kg of lithium (reactant). Also the reaction kinetics is not very well known. I only know that it is an instantaneous reaction. So what should I do in such case.

Amir September 15, 2012 04:07

Quote:

Originally Posted by snh (Post 381879)
Hi Amir,
Thanks for the reply and your help is appreciated. I will try this out but I have no idea about the time required to consume the 5 kg of lithium (reactant). Also the reaction kinetics is not very well known. I only know that it is an instantaneous reaction. So what should I do in such case.

Well, your chosen chemistry-turbulence model is appropriate. But regarding the amount of consumption, I think you can use a UDF, where you compute the amount of products in the whole domain via a volume integration and according to the stoichiometric reaction, you can stop the calculation when your constrain is justified.

Bests,

eng_s_sadeghi September 16, 2012 04:50

Quote:

Originally Posted by snh (Post 381736)
Hi

I am trying to simulate a reaction between lithium and sulphur hexafluoride using the eddy dissipation model. I would like to know if I can input specific quantity of reactant..If I want to study the consumption of 5 kg of lithium in the reaction, is there any way where I can enter the amount of reactant used?

Hello,
At first you should have the reaction rate. Then you can use DEFINE_NET_REACTION_RATE macro for implementing the reaction rate. Like this:

DEFINE_NET_REACTION_RATE(user_net_reaction_rate,c, t,particle,pressure,temp,yi,rr,jac)
{
int i;
for(i=0;i<n_spe;i++)
rr[i] = 1./(real)n_spe - yi[i];
}

snh September 17, 2012 03:25

Quote:

Originally Posted by Amir (Post 381884)
Well, your chosen chemistry-turbulence model is appropriate. But regarding the amount of consumption, I think you can use a UDF, where you compute the amount of products in the whole domain via a volume integration and according to the stoichiometric reaction, you can stop the calculation when your constrain is justified.

Bests,

Hi Amir,

I have one more question. The lithium that is used in the reaction is already present in the reactor in molten form and its amount is 5 kg & the oxidizer i.e. sulphur hexafluoride is injected through a nozzle. So in this case can I use the source term to specify the amount of lithium since the amount of lithium present in the reactor is not increasing with time and is decreasing as the reaction proceeds.

Thank You.

snh September 17, 2012 04:38

Quote:

Originally Posted by eng_s_sadeghi (Post 381971)
Hello,
At first you should have the reaction rate. Then you can use DEFINE_NET_REACTION_RATE macro for implementing the reaction rate. Like this:

DEFINE_NET_REACTION_RATE(user_net_reaction_rate,c, t,particle,pressure,temp,yi,rr,jac)
{
int i;
for(i=0;i<n_spe;i++)
rr[i] = 1./(real)n_spe - yi[i];
}

Hi,

Thank you for joining in. I will implement the reaction rate as mentioned by you. I would like to know if there is any way by which I can mention that 5kg of lithium (reactant) is already present in the reactor. Since I am new to fluent, I am facing problem of modeling my problem statement in a correct way in fluent.

Thank you.

Amir September 17, 2012 06:01

Quote:

Originally Posted by snh (Post 382067)
Hi Amir,

I have one more question. The lithium that is used in the reaction is already present in the reactor in molten form and its amount is 5 kg & the oxidizer i.e. sulphur hexafluoride is injected through a nozzle. So in this case can I use the source term to specify the amount of lithium since the amount of lithium present in the reactor is not increasing with time and is decreasing as the reaction proceeds.

Thank You.

Hi,

I think your concern is about pyrolysis of the lithium which doesn't have any relation to the rate of reaction; you've used a proper turbulence-chemistry model.
You have to know that FLUENT is not a technical software in reaction simulation so you have to model few parameters such as rate of pyrolyzation.
i.e., you can define a proper relation as a source term to release desired amount of reactant and let FLUENT find the rate of reaction according to the eddy dissipation model. (If you have a hydrocarbon fuel I strongly recommend you to use FDS instead of FLUENT)

Bests,

eng_s_sadeghi September 17, 2012 06:13

Hi,

Fluent is highly powerful software that can be used for solving all finite volume problems. It you see it doesn't have predefined modules for your problem, you can use it just as a high speed solver core, as I have done in my master project (solid fuel combustion in a packed bed of porous materials).

In this project, I have used fluent just as a solver and have written more than 1400 lines of complex User Defined Functions for implementing solid fuel combustion, limestone decomposition, moisture evaporation and condensation and other parameters that required for modeling this process.

I believe in Ansys Fluent and I am sure all Finite Volume problems can be solved by UDF programming in it.

Amir September 17, 2012 06:34

Quote:

Originally Posted by eng_s_sadeghi (Post 382095)
Hi,
I believe in Ansys Fluent and I am sure all Finite Volume problems can be solved by UDF programming in it.

You're kidding! :D
Have you ever seen alternative open source softwares such as openFOAM?!!
FLUENT is very restrictive in comparison to such softwares. e.g., can you implement you own pressure-velocity coupling method or discretization scheme?!

eng_s_sadeghi September 17, 2012 06:50

Quote:

Originally Posted by Amir (Post 382102)
You're kidding! :D
Have you ever seen alternative open source softwares such as openFOAM?!!
FLUENT is very restrictive in comparison to such softwares. e.g., can you implement you own pressure-velocity coupling method or discretization scheme?!

As I said, as a solver core. Do you understand?
I mean if there would not be any need to change the discretization, as in most problems there is no need of changing the solving method. The only problem would be just to add equations.
How much time does it take to change OpenFOAM solver core? I am not so crazy to do that...

eng_s_sadeghi September 17, 2012 06:55

Quote:

Originally Posted by Amir (Post 382102)
You're kidding! :D
Have you ever seen alternative open source softwares such as openFOAM?!!
FLUENT is very restrictive in comparison to such softwares. e.g., can you implement you own pressure-velocity coupling method or discretization scheme?!

I don't like discretizing problem. They are not duty of a mechanical engineering. Mathematicians should deal with it not a mechanical engineer. I hate it.
What is your experience? Discretizing, haha. You will not get in touch with the phenomena with just work with them.
I hope you understand what I mean.

Amir September 17, 2012 09:24

Quote:

Originally Posted by eng_s_sadeghi (Post 382103)
As I said, as a solver core. Do you understand?

It seems that you don't exactly know what the solver core is! The discretization schemes are not categorized as a part of solver core, they are additional compiled libraries which attached to the solver core. ("dll" files in windows and "so" files in linux)
Quote:

Originally Posted by eng_s_sadeghi (Post 382103)
I mean if there would not be any need to change the discretization, as in most problems there is no need of changing the solving method. The only problem would be just to add equations.

Not at all! you can severely save time if you use optimized code in all of its aspects.
Quote:

Originally Posted by eng_s_sadeghi (Post 382103)
How much time does it take to change OpenFOAM solver core? I am not so crazy to do that...

I'm surprising! It depends on your capabilities but if you think so, it's better for you just to use FLUENT. :D
Quote:

Originally Posted by eng_s_sadeghi (Post 382105)
I don't like discretizing problem. They are not duty of a mechanical engineering. Mathematicians should deal with it not a mechanical engineer. I hate it.

If you really think so, I'm very sorry for you.
Quote:

Originally Posted by eng_s_sadeghi (Post 382105)
What is your experience? Discretizing, haha. You will not get in touch with the phenomena with just work with them.

No this is not my experience. My field is viscoelastic fluids; lost capability of FLUENT! :D

eng_s_sadeghi September 17, 2012 09:42

OK. I don't know CFD. Goodluck.


All times are GMT -4. The time now is 16:19.