[hitzhwan]

Originally Posted by vinerm

Second order time-discretization scheme, either pure 2nd order or bounded one, are not available for Explicit VOF. This is because 2nd order requires a larger stencil of time, i.e., n-1, n, and n+1, which is not available while using explicit VOF (data at time n-1 is unavailable).

If mesh is good and time-step is small enough, first-order accuracy is as good as second-order.

Do you have email, I can send you a file.

[vinerm]

Is there any particular reason for using Explicit VOF? If you are modeling a falling film and the film is quite thin, better to use Eulerian Wall Film model. Modeling a thin film with VOF is always expensive. Other option is to use Implicit VOF along with LSM to improve accuracy.

[hitzhwan]

Originally Posted by vinerm

Is there any particular reason for using Explicit VOF? If you are modeling a falling film and the film is quite thin, better to use Eulerian Wall Film model. Modeling a thin film with VOF is always expensive. Other option is to use Implicit VOF along with LSM to improve accuracy.

(1)Because only use the explicit VOF can I use the Geo-reconstruct scheme for the volume fraction equation.
(2)I do not know how I can use a Eulerian Wall film model, can you tell me how I can set it? Or I can send my case to you. If you are available, I can send you by email. For the forum can not submit a file exceed 195.3 Kb.
(3)What is the LSM, how can choose the Implicit VOF along with LSM.
Thank you very much for your reply.

[vinerm]

Geo-Reconstruct is the best reconstruction algorithm available in Fluent, however, CICSAM is good as well and is available with implicit VOF.

EWF is a separate model under Models. But it is not available under older versions. I suppose this is available starting from R14.

LSM is level set method. It is not available as a stand-alone model in Fluent but in combination with VOF. This is useful for scenarios that have significant surface tension.

[hitzhwan]

Originally Posted by vinerm

Geo-Reconstruct is the best reconstruction algorithm available in Fluent, however, CICSAM is good as well and is available with implicit VOF.

EWF is a separate model under Models. But it is not available under older versions. I suppose this is available starting from R14.

LSM is level set method. It is not available as a stand-alone model in Fluent but in combination with VOF. This is useful for scenarios that have significant surface tension.

Does the "implicit VOF" means the transient formulation is implicit?

[vinerm]

Transient formulation in Fluent's pressure based solver is always implicit. Implicit means that VOF capturing is implicit.

[hitzhwan]

Originally Posted by vinerm

Transient formulation in Fluent's pressure based solver is always implicit. Implicit means that VOF capturing is implicit.

But we I choose the implicit VOF capturing, there is no CSCSAM option for the volume fraction, there are only "compressive" and "modified HRIC" optiions. Only I use explicit VOF capturing can I choose the CSCSAM option.

[vinerm]

In place of CICSAM, use compressive.

[hitzhwan]

Originally Posted by vinerm

In place of CICSAM, use compressive.

Do you think the compressive option can capture the film surface accuracy ? Have you use the option in simulation falling film?

[vinerm]

As stated earlier, Geo-reconstruct is the best available option. However, this is about reconstruction of the interface. Capturing depends a lot on the mesh. If the mesh is fine enough, results will be acceptable. If the film thickness is less than 1 mm, I'd recommend using EWF instead of VOF.

[hitzhwan]

Originally Posted by vinerm

As stated earlier, Geo-reconstruct is the best available option. However, this is about reconstruction of the interface. Capturing depends a lot on the mesh. If the mesh is fine enough, results will be acceptable. If the film thickness is less than 1 mm, I'd recommend using EWF instead of VOF.

Could you give me an example for using the EWF. I do not know how to use it?

[vinerm]

I hope the following is useful

http://dl.mr-cfd.com/tutorials/ansys...(Eulerian).pdf

[hitzhwan]

Originally Posted by vinerm

I hope the following is useful

http://dl.mr-cfd.com/tutorials/ansys...(Eulerian).pdf

I have a question, if some positions have film thickness higher than 1 mm, but some lower than 1mm, which option should I choose? In addition, why can not use the VOF at film thickness lower than 1mm? Thank you

[vinerm]

VOF is applicable as long as N-S is applicable, however, it becomes very expensive since a very fine mesh, spatial as well as temporal, is required. EWF can be used as long as film thickness is lesser than first cell height. So, you would need a coarser mesh. 1 mm is just a recommended value, you can use EWF for thicker films as well.

[hitzhwan]

Originally Posted by vinerm

VOF is applicable as long as N-S is applicable, however, it becomes very expensive since a very fine mesh, spatial as well as temporal, is required. EWF can be used as long as film thickness is lesser than first cell height. So, you would need a coarser mesh. 1 mm is just a recommended value, you can use EWF for thicker films as well.

Yes, I have found that when my first cell size is 0.06 mm, it will make the timestep at 1e-6 for me, which will take a long time to calculate. Do you know how to choose a mesh size for the first cell size? I do not know how to choose?

[vinerm]

For normal flows, mesh size is usually selected on the basis of . However, for thin films, you have to ensure that there are atleast 10-12 cells within the expected film thickness. If film is too thin, mesh would be fine and expensive to use.

[hitzhwan]

Originally Posted by vinerm

For normal flows, mesh size is usually selected on the basis of . However, for thin films, you have to ensure that there are atleast 10-12 cells within the expected film thickness. If film is too thin, mesh would be fine and expensive to use.

Do you know why the mesh is not uniform(see the attachment) around the tube surface?

[vinerm]

The mesh is already as uniform as it can get. It's the height of first cell and the growth rate that matters.

[hitzhwan]

Originally Posted by hitzhwan

Do you know why the mesh is not uniform(see the attachment) around the tube surface?

You mean it is reasonable for this phenomenon, I can not understand, but I find that the mesh in the article have not the phenomenon, why?

[hitzhwan]

Originally Posted by hitzhwan

I have a question, if some positions have film thickness higher than 1 mm, but some lower than 1mm, which option should I choose? In addition, why can not use the VOF at film thickness lower than 1mm? Thank you

Does anyone have this file?

Collection efficiency calculation and running wet and runback
analysis on wing



Copy the files, (NACA_0012.msh, film_accretion.c) to the working folder.




Thank you for sending me[/QUOTE]