I need a help for my combustion simulation

COMSOL Moderator

Hello Olivier Thierry Sosso Mayi

Your Discussion has gone 30 days without a reply. If you still need help with COMSOL and have an on-subscription license, please visit our Support Center for help.

If you do not hold an on-subscription license, you may find an answer in another Discussion or in the Knowledge Base.

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Olivier

Did you managed to solve this one? I have simplified model from combustion (turbulent diffusion, non-premixed), which is using following models:

- Turbulent Flow, k-ε
- Transport of Diluted Species
- Heat Transfer in Fluids

Because model is turbulent, I'm using Eddy Break-Up model to define reaction rate. This is "mixed is burnt" assumption (see: www.cfd-online.com/Wiki/Combustion). This model doesn't take count of kinetic reactions in combustion. Model can be later to modify in to complex ones, like taking account of kinetics, but this is good start.

So far I have get it working only in 3D. 2D and 2D- axisymmetric are not working.

Best regards

Tero Hietanen

Francois Gregoire

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Tero,

I will have to model turbulent combustion eventually and I would like to ask you a couple of questions regarding your implementation of the eddy break up model:

1. What is your mixture? which fuel?

2. How many species do you consider (reactants and products)? Do you consider dissociation reactions of H2O and CO2?

3. Did you modify the equations in the turbulence k-e node and the heat transfer node in order to reproduce the model? Or you started from scratch with the mathematical interfaces?

4. How did you treat heat transfer along the boundaries?

Ok, a lot of questions huh?! If you have 2 minutes it would be very appreciated. Combustion is a big step and a little guidance would help a lot.

Is it not weird that your model worked in 3D but not in 2D??? Do you have an idea why?

Thanks

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Francois

I hope you have reserved lot of time to do your model! I have spend several moths to do this one and still I don't is it really doing right things...

1. and 2. So far I have used methane. Reaction is two steps mechanism:

CH4 + 1.5*O2 => CO + 2*H2O
CO + 0.5*O2 => CO2

For CO I have also kinetic reaction rate which is taking account of water vapor. Dominating reaction rate is selected by minimum of reacting rates (kinetic or mixing).

3. k-eps and heat transfer models are originals Comsol but heat transfer has heat source by reaction.

4. Heat transfer to boundaries is only by convection, no radiation (no time to test this one).

For 2D/3D problem I have no answers. Model is working but fuel is not mixing or burning. Do you have ideas?

Does these answer to your questions?

Best regards

Tero

Francois Gregoire

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

thanks for the answer.

I can't really help you, I have not undergone this model yet. Maybe you have to set initial values of products > 0 to initiate reaction?

Do you obtain realistic flame temperatures?

There are some papers about combustion in Comsol but I haven't read those yet. I can't see how we could use a PDF approach in Comsol...

I'll come back to this thread if I can work something (in a couple of months...).

Cheers

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Francois,

Initial values didn't have any effect to reaction.

Flame temperatures are still too high. This is issue what I'm solving now.

Where did you found articles from Comsol and combustion? Some reaction stuff is available but combustion I didn't found at all...

PDF and Comsol... Sounds interesting (=difficult).

Br Tero

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

I am working on this phenomenon too. I think the problem is lack of combustion model in Comsol. you can just
define the species and steps of the reaction. I am in doubt if it was sufficient?
I think we could do it faster by sharing our findings due to there is no guidance in combustion solving in Comsol.
My Email is: amajburi@yahoo.com
would be very thankful if could see your findings or the process you have done in Comsol.

Francois Gregoire

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

I have not done combustion modeling with Comsol yet, but from my readings, I think that the Eddy break-up or Eddy dissipation concept are straightforward to implement in Comsol.

But another combustion model such as the probability density function (PDF) approach seems laborious to implement... Good old Fluent is teasing me a lot for the combustion part of my project...

I'll keep you posted if I ever succeed in modeling combustion in Comsol.

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hello Amir

Sorry, I cannot provide you full model because it is my company model. I can provide you basic guidelines what physics or models are used. These can be also found from literature.

Like Francois said I don't see problem implementing combustion to Comsol by Eddy break-up or Eddy dissipation model (I have used both ways). I think this is made quite same way in Fluent (according manuals). I have not used Fluent in combustion simulations but it seems to "default"-software in all reactive flows.

In my model I have managed to solve previous problems and now I'm verifying results.

Best regards

Tero

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thank you Francois

Modeling combustion seems to be easy while utilizing Ansys CFX or Fluent, but according I need to solve Electro hydro dynamics, I need to use Comsol as it is very strong in multi physics problems.

Again I would be very thankful if could get more information or at least a tutorial for combustion solve in Comsol.

Regards
Amir

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Tero

Thank you for your reply. I compeletely understand for your limitations in sending what you have done.
I just have some questions :
In use of consentration and species migration I need to define Maxwell-stefan diffusivity matrix, also in defining parameters like density or the coefficients I have some questions.

In combustion as the temperature rises all the parameters are getting change, for example the mobility of the particles should be change as themperature rises. How I can mention this influence.

It would be very nice if I could study an example or a good tutorial. Again thanks for your care Tero.

Regards
Amir

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Amir

I can gather you some basic information but "step by step" -tutorial takes quite lot of work. I will make this as soon as I have some spare time...

Sorry, I don't have any good ideas to your other questions.

Best regards

Tero

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Great Tero

Again thank you for your care

Mayur

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

I don't know whether I am replying it correctly or not, regarding change in parameters w.r.t. temperature, you have to define all the physical properties as a function of temperature..

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thank you Mayur

The problem is I do not know how the parameters is the function of Temperature.

Mayur

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

what kind of parameters you need to define?

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thermal conductivity of the mixture, Heat capacity, particles mobility and the maxwell-stefan matrix components.

Another problem is finding the temperature for the reaction,
for example there is an empirical Temperature relativity for N2:
Cp=39.06-512.79T^-1.5+1072T^-2-820.4T^-3

I can t find the part that gives the reaction temperature to use it.

Mayur

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

i think you can look at couple of examples in comsol under chemical reaction engineering module. To define reaction heat there is a "heat source" option under convection and conduction heat transfer module.

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Amir

Here is first version of basic combustion stuff.

Br

Tero

Attachments:

Combustion with Comsol4 ver1.pdf

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Found an error. Here is fixed version.

Br Tero

Attachments:

Combustion with Comsol4 ver2.pdf

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Unfortunately I could not find same example, any way thanks for your care and guidances Mayur.

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thank you Tero for the useful PDF.

Abdul Halim

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Dear Comsol users,

Now I am trying to model a rotating machinery in AC/DC Module. It is exactly an axial flux permanent magnet synchronous generator.
I`ve finished the geometry and model of the generator. Yes, It`s like a disc, so it will be more efficient than the radial one.

These are the explanation of my model:
- Permanent magnet poles are pleced in the back yoke of rotor.
- Stator, which consist of some winding/coil, are placed between the two rotor. I`m using coreless stator one.
- There is air gap between rotor and stator.

I have adjusted the subdomain physics settings by choosing the material and the H-B equation, but I have no idea, what I have to do next.. So users, please suggest me what I have to adjust next.

- How to adjust the boundary settings?
- What function and equation that I`ve to give for the model.. In what part?
- and how to visualize the outcoming voltage of every phase of the generator??

Thank you
Halim, Zahari

Attachments:

axial flux permenant magnet generator.mph

David Gongora

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hello Tero,

I'm just seeing this forum about combustion simulation. I hope you can help me.
Now I'm also working on a combustion project to design a flat flame burner, but unfortunately my simulation doesn't goes well.

I always get this error messages:
Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution is not converged.

or this:
Failed to find a solution.
In segregated group 2:
Very ill-conditioned preconditioner.
The relative residual is more than 1000 times larger than the relative tolerance.
Returned solution has not converged.

I use a global one step reaction mechanism: CH4 + 2O2 -> CO2 + 2H2O
And I think I have input the right physics the same as you mentioned above, and input the right data.
Can you guide me for this problem? I can send you my .mph file (can be opened with v4.1 or 4.2) if you want to see it.

Thanks.

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi David,

Yes, I can check your model.

Solving has been problem also to me. Usually I have solved this by calculating first flow, then reactions and finally heat/temperature. Now new version 4.3 is giving extra problems (turbulent mixing).

Br Tero

David Gongora

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Tero,

Have you received my files?
I just try to solve in order of what you said. But again it's failed. I think I have some mistakes that I don't know where to fix them.

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Dear David

The error can be because of discrepencies in units. if you use Arrhenius equation to gain rate of reaction
just check the units for the coefficient spediaaly double check the Pre-ex factor (Af).

good luck
AMIR

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Dear Tero

I have modeled the combustion of ethane by ur guidance in 3D but for the propane with an extera step in 2D Axissymetric, I can not get the flame. It seems the total Q is very small to be distinguished although for the 0D model of the reaction the Q is high.
I can not attached the file here but I can send it by email if possible.

Special thanks
AMIR

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Amir,

In Comsol version 4.2a turbulent mixing is not taken account. Due to this in 2D and 2D-axi fuel is not mixing properly with oxygen. I have tested new Comsol version 4.3 and now fuel is mixing with oxygen but I have still problem with convergence... In 3D result are quite same between versions 4.2a and 4.3 (?).

You can clear all solutions and mesh from "edit" to get file smaller for sending.

Best regards

Tero

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thank you Tero

I have modeled the problem in 3D but again same problem occured. No reaction is plotted!
The job is exactly same as what I have done with ethane, The job with ethane works but
for propane nothing happens. I really confused. unfortunately I can not access new version of
COMSOL according that it is not very famous software here in our country and I don know how I can get the new version.

Regards
AMIR

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thank you Tero

I have modeled the problem in 3D but again same problem occured. No reaction is plotted!
The job is exactly same as what I have done with ethane, The job with ethane works but
for propane nothing happens. I really confused. unfortunately I can not access new version of
COMSOL according that it is not very famous software here in our country and I don know how I can get the new version.

Regards
AMIR

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Thank you Tero

I have modeled the problem in 3D but again same problem occured. No reaction is plotted!
The job is exactly same as what I have done with ethane, The job with ethane works but
for propane nothing happens. I really confused. unfortunately I can not access new version of
COMSOL according that it is not very famous software here in our country and I don know how I can get the new version.

Regards
AMIR

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hello Amir,

Do you have some initial values for H20 and CO2? If you are using EDM, it needs some starting values for those or else "IF" selection goes to zero. Or do you have temperature dependent reaction rate which needs "spark"?

Best regards

Tero

Amir Majbouri

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Tero

No CO2 or H2O in initial conditon, but a T0 settled as role of spark.
The problem was solved. I had made mistake in putting data for gamma.
but now have some problem in putting electric field on the flame...
again thank you for your useful guidance Tero.

AMIR

Olivier Thierry Sosso Mayi

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Tero,
I'm happy to come back on the forum, only because i did not yet have any solution of my combustion problem. As you succeeded to do it, can you Tero send me in my E-mail your procedure.
My mail is olivier.thierry.sosso. mayi@uqtr.ca
Please Tero it's very important for me, to be able to go fast. I succeeded to have good temperatures using reaction engingnering for combustion of methane, now my problem is to join it with multiphysic, laminar flow, transport of diluted species, or concentrated species, i think both can do the job, and finaly heat transfer.
Although my problem is in laminar flow and in the micro scale, i think that if i succed to do one example, it will be possible for me to do the same thing for my case.
Best regards, i am waiting for your reply

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi Olivier,

It looks like I cannot help you in this case because you are using laminar flow. This model is working only with turbulent flow, not laminar. Laminar flow needs some other type of approach which I have not studied yet. Sorry!

Procedure of this model can be found in middle of this thread (pdf-file) and here are some more details:

www.comsol.com/community/forums/general/thread/30127/

Best regards

Tero

Zari

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi,
I'm doing simulation of a combustion zone for a gas mixture. I face most of the problems you had before.
Something you may can help me is how you defined the ignition source?
I started with Chemical Reaction engineering module(re) to check if the set of reactions work well. Since the inlet temperature is lower than auto-ignition, a temporary heat source to start the exothermic reactions is needed. What do you suggest for that?

Best,
Zari

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi,

Have you check how initial temperature affect to reaction?

Br Tero

Zari

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hi,
I am here with another question!

I am still working on my combustion model and since I have a mixed fuel as inlet (6 species) then I needed to have a reaction engineering interface in addition to the CFD interfaces: Heat transfer in fluids, Turbulent flow k-e, and Transport of diluted species.

There is a question regarding the boundary condition setting for turbulent flow at outlet. The options are: 1-Pressure no viscouse stress, 2-Pressure, 3-No viscouse stress, 4-Normal stress, and 5-Velocity. Which one did you set and worked better? I receive errors when choosing 3&5, and negative pressure at some points with 1-2&4!

Best,
Zari

Tero Hietanen

Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago

Hello Zari,

I have used "Pressure, no viscouse stress" condition with small negative pressure value (-20 Pa). Minus pressure I have used to prevent back flow(?!) and give real situation to combustion chamber (exhaust fan). I have not counted problem for this boundary condition.

I have noticed that Comsol needs quite good initial conditions and due to this I solve these cases with steps: first flow, second reactions and last heat. Final solving contains all these cases.

Best regards

Tero

Discussion Forum

I need a help for my combustion simulation

Suggested Content