[Gerrit]

Hi everyone,

I am learning OpenFoam now for arround three weeks. For my project from university I want to set up a simulation where UV-light is present and where I have some particles in the flow. In the end I would like to know how much UV-light a particle received.
The simulation is steady-state (Thanks god ) so I am using until now simpleFoam as a solver.
To track the particles I looked at this overview and think for my case the solidParticle or Lagrangian.
The radiation models are a little bit more difficult, as I do not have a combustion or something like that. On this page there is an overview about the different models as well for radiation...

The trouble I have until now is:
1. Can I combine these three topics in one case?
2. Does anyone of you have some experience with this kind of thing? Is it better to use solidParicle or direct the Lagrangian?
3. Can I use one of the radiation models to model the UV-light? P1 e.g?

I hope my questions are not too dumb...

Thanks a lot!
Best regards
Gerrit

[Gerrit]

Ok so far...

As long as no one answers I will write here, what I found until now. So others can maybe follow my thoughts if they have a similar problem. Or maybe someone has a better answer for me

I started up collecting more information about particle tracking and ended up figuring out, that it seems only possible to track particles using transient solver. I don't want that (will probably take too many resources).

Now I found here a discussion about something similar. In short: there the idea is, to use a transient solver but choose time and timestep in a way it can be considered as a single step.
Hopefully I repeated that correct now....

[Gerrit]

Hi everyone,

I still did not manage to implement radiation and particle tracking - but I did not spent too much time on it.
I have a question about the solver in general. As my first step was to just have a model with flow (no particles or radiation), I used simpleFoam as a solver. My fluid is water, and my model runs quite nice. But I am not sure if, my model "accepted" that I want to use water.

transportProperties

nu              nu [ 0 2 -1 0 0 0 0 ] 1.422e-03;

rho         rho [ 1 -3 0 0 0 0 0 ] 1000;

Is that enough to tell the solver that I want to use water as a fluid?

Thanks!

[Gerrit]

Hi,

concerning the last post: That was non-sense as simpleFoam is defined as incompressible by the viscosity(density included)...

My original plan to find a solver, that can handle everything does not seem to work. There is no solver out of the box for this case(at least I could not find one)... I received a lot help until now (thanks by the way )

The "new" plan, is to split the tasks:
1. Solve the flow field with simpleFoam
2. Solve the particle tracks with something else (transient as it seems until now), based on the flow field from 1.
3. Radiation modeling

About point 2:
My original plan to combine solidParticle and simpleFoam received a lot support, but I gave up as my programming skills are not that good until now...
I choose MPPIC Foam, as I do not have a reaction occurring. In a laminar case the solver runs wihtout problems the same for the tutorial case with LES (/opt/openfoam240/tutorials/lagrangian/MPPICFoam/cyclone). But as soon as I want to run a case with k-epsilon model I get errors.

Maybe the question is stupid, but has anyone already used this kind of solver with RANS models? Or does someone has an alternative idea?

Thank you very much

Best regards
Gerrit

[Gerrit]

Hello,

my project is closed and done (at least for me).
How I did it in the end:
1. I solved the flow fields with simpleFoam and used the flow fields for a Lagrangian simulation based only on drag forces.
2. The radiation (light in this case) is modelled in Matlab and gives intensity fields

The particle tracks and the intensity fields of the light are combined as well in Matlab.

Best

[ShashiDeSilva]

Hi,

I am working on a cyclone simulation with the MPPICFoam solver. It is finely working fine with LES(smagorinsky) and laminar flow models. Furthermore I want to run it with RAS (k epsilon) model. Still i could not succeed as it it gives errors at the beginning. Is it require to change fvsheams and fvsolver files when changing LES to RAS ?
I am stuck here . I would highly appropriate any help.

[Gerrit]

Hi ShashiDeSilva,

at first: Please open your own thread. For others it is very confusing to get information, when two not connected topics are in one thread.

About your question: You are not giving a lot of information --> Nobody can help you!
Important to know would be: Is the mesh ok? Does it makes sense to use k-epsilon with your mesh?

Best
Gerrit

[ShashiDeSilva]

Dear Gerrit

Thank you for your reply. actually I just post here because you also try MPPIC solver with k-epsilon. Yesterday I found the reason. I have replaced some 0 values to very small value ( 1e-6) in the epsilon boundary condition file. Now it is working.
If possible, I will remove this thread as you suggested.

Best Regard
Shashi

[Gerrit]

Hi Shash,

Great that it works! I did not use MPPIC in the end.
Nevermind.Leave your messages here ;-)

Best
Gerrit