[iconde]

Hello Germilly,

Really interesting work, we are working in chtMultiRegionTwoPhaseEulerFoam with two regions, one fluid (with air and iron inside) and another one solid (porous sand) trying to simulate filling a sand mold with iron.

We have problems evacuating the air without outlet pipes and we are trying to find a boundary condition to fix it. have you got any idea how we can do this?

Also, have you got any "tutorial case" to see how do you prepare your solver?

Thanks in advance.

[Germilly]

Originally Posted by iconde

Hello Germilly,

Really interesting work, we are working in chtMultiRegionTwoPhaseEulerFoam with two regions, one fluid (with air and iron inside) and another one solid (porous sand) trying to simulate filling a sand mold with iron.

We have problems evacuating the air without outlet pipes and we are trying to find a boundary condition to fix it. have you got any idea how we can do this?

Also, have you got any "tutorial case" to see how do you prepare your solver?

Thanks in advance.

Hello,

There were some good online tutorials, which are mentioned in #4, #12 and #18 of this thread, but they are not available now, and I dont know why. Have you seen my comments in #18 of this thread? See if the following tutorial can help.


GB

[Tobi]

Hi all,


I just want to jump in as I was developing a similar model for calculating the heating-up of a SCR catalyst (6 m³ volume) within an exhaust pipe system. I was using pimpleFoam for my approach with some extensions. My extension was as follows:

• I defined the SCR as cellZone while using the Darcy-Forchheimer equation to model SCR catalyst as a porous media
• I added the energy equation to the solver (actually a passive scalar)
• The properties for cp, lambda and rho for the air were implemented as polynomial functions based on the VDI Wärmeatlas database
• For the cellZone, I calculated the temperature depended properties of the SCR material (cp, lambda, rho)
• Knowing the ratio of fluid/solid in the cell zone, I modified the thermophysical properties of the fluid in the cell zone in order to get some semi-solid properties

This approach is fast, easy but the main problem is, that the heat-exchange from hot air to the solid is ideal which means that 100 % of the energy is stored within the cell zone. In other words:

• The air behind the cell zone keeps a low temperature until the whole cell zone is heated up which does not represent the truth

Even though I am working with foam for 9 years now, you guys have interesting ideas. Thanks for sharing your thoughts.

We have problems evacuating the air without outlet pipes and we are trying to find a boundary condition to fix it. have you got any idea how we can do this?

Personally, I would solve such topic with a codedFixed/Mixed boundary condition. E.g. if the face is in contact with air -> let the air escape (outlet) otherwise make it a fixed wall with velocity (0 0 0).

[iconde]

Hello foamers,

Thanks for your replies.

Originally Posted by Germilly

There were some good online tutorials, which are mentioned in #4, #12 and #18 of this thread, but they are not available now, and I dont know why. Have you seen my comments in #18 of this thread? See if the following tutorial can help.

Germilly thanks a lot for your suggestions. We have seen your #18 comments and tutorial but we don´t get any solution for our problem. We are focusing in a new boundary condition, because we undertand that is where we have the main problem considering that we are capable of filling the mold region as we can see in the next image, but maybe we are in wrong way. Any ideas?

https://imgur.com/y0YnNWp

Originally Posted by Tobi

I just want to jump in as I was developing a similar model for calculating the heating-up of a SCR catalyst (6 m³ volume) within an exhaust pipe system. I was using pimpleFoam for my approach with some extensions. My extension was as follows:


I defined the SCR as cellZone while using the Darcy-Forchheimer equation to model SCR catalyst as a porous media
I added the energy equation to the solver (actually a passive scalar)
The properties for cp, lambda and rho for the air were implemented as polynomial functions based on the VDI Wärmeatlas database
For the cellZone, I calculated the temperature depended properties of the SCR material (cp, lambda, rho)
Knowing the ratio of fluid/solid in the cell zone, I modified the thermophysical properties of the fluid in the cell zone in order to get some semi-solid properties
This approach is fast, easy but the main problem is, that the heat-exchange from hot air to the solid is ideal which means that 100 % of the energy is stored within the cell zone. In other words:

Thanks for your answer Tobi, in our case, we are working with chtMultiRegionTwoPhaseEulerFoam (which is new solver in Openfoam), since it allows to simulate a heat transfer between solid and fluid regions, and define two different phases in the fluid regions (air and iron).

Originally Posted by Tobi

Personally, I would solve such topic with a codedFixed/Mixed boundary condition. E.g. if the face is in contact with air -> let the air escape (outlet) otherwise make it a fixed wall with velocity (0 0 0).

Considering that our fluid cavity is surrounded by a solid mold and has 2 phases; air and iron, the aim is to achieve a boundary in which the air tends to 0, so that the entire cavity ends up being filled with metal.

For this we are trying to manipulate the values ​​of pressure, velocity and fraction (alpha.air, alpha.iron) at the boundaries using the GroovyBC. However, we failed to reduce the pressure levels. Do you think the solution could be something which involves the mass fraction?

Regards,

Anna

[Tobi]

Originally Posted by iconde

For this we are trying to manipulate the values ​​of pressure, velocity and fraction (alpha.air, alpha.iron) at the boundaries using the GroovyBC. However, we failed to reduce the pressure levels. Do you think the solution could be something which involves the mass fraction?

Regards,
S
Anna

Hi Anna,

during my PhD. I also did some mold filling but I was too lazy to investigate into such topic too much. Right now your pictures looks like "MAGMAsoft". You meshed with snappyHexMesh without snapping, right?

For your boundary problem, I cannot give you a clear answer as I never did it.

[Tobi]

@Germilly, can you please give some statement about the following code:

+ hconv*Tf-fvm::Sp(hconv, Ts)

The Sp() term should be a stabilization term, right. I am wondering why you are using Ts here?

[Germilly]

Originally Posted by Tobi

@Germilly, can you please give some statement about the following code:

+ hconv*Tf-fvm::Sp(hconv, Ts)

The Sp() term should be a stabilization term, right. I am wondering why you are using Ts here?

Hello Tobias,

The energy equation for the solid structure (solid phase) is:


The term , which describes the heat transfer between the solid and fluid, can be written as In OpenFOAM I have:

+ hconv*Tf-fvm::Sp(hconv, Ts)

If I remember, the fvm::Sp is to make the variable to be as implicit source term (I am solving the equation for variable ).

Am I right?

There are some parts of the code that need to be improved. If you have any suggestion, I would be grateful if you can share them here.

GB

[Tobi]

Hi,

If I remember, the fvm::Sp is to make the variable Ts to be as implicit source term (I am solving the equation for variable Ts).

Not 100 % correct. the fvm:: operator do belong to the implicit part. fvc:: is an explicit calculation. The Sp() term for me was always necessary for boundedness regarding the convective term (https://openfoam.org/release/2-2-0/n...s-boundedness/). However, I guess you are right that the Sp() term makes a source implicit and the SuSp does make it implicit and explicit. I never checked these things in detail.


However, my latest results:

[Tobi]

Hi,


I just want to refer to my post I made a few days ago: How to implement equation of mixture fraction and variance of mixture fraction?


Here, one can see the Sp() Term

PDF: Nikolas

[Mhd_Sal]

Hello Germilly,


Here i am replying for your 3 year old post. i am working on something similar to what you have done. i wanted to follow the tutorial you provided above. Sadly it says "Page not found", may be i'm too late and they have taken down the page. Is there any other way to get that tutorial back?


Thanks,
Sal

[Wowbagger]

Hi at all and thanks for all the contributions which helped me a lot with my current case of simulating a porous media.
Especially the catalyst tutorial from Tobi helped me saving hours!

Thanks for that.



I opened a new thread bcs I got a few thoughts about the setup with the geometrical data.
Feel free to have a look at: Porous media - heat transfer - set up of mean values