CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)

 ztdep April 27, 2006 20:45

about the numerical method for the free surface

Hi:

!) How many methods do we have to treant the free surface numerically!

2) Does sombody know the "spine method" used in the commercial software "FIDAP"

regards

 Renato. April 29, 2006 17:49

Re: about the numerical method for the free surfac

Well, I'm novel in this area and I could say that most of the works related to free surface can be classified in two areas.

Those working with interface tracking - an Eulerian approach where the mesh follow the free surface movement and those working with interface capturing - when the mesh is captured with some kind of function. In the second approach, most of the works I've already seem are due to VOF and Level Set methods. VOF is older, hard to solve since it work with a step marking function, has good mass preserving and is not able to include some effects, such as surface tension. Level Set methods have been growing up in the last decade, mainly due the guys working with Cartesian grids (Sethian, Osher and co-workers) and structured solvers. There are, relatively, few works in LS area employing unstructured solvers like FVM or FEM.

Taking a fast look in the FIDAP users guide I realized that FIDAP has a capturing method and includes surface tension effects. It seems that FIDAP employs I kind of Level Set method, even without saying it explicitly in its manual.

Hope I've cleared up something...

Regards

Renato.

 ztdep April 30, 2006 10:31

Re: about the numerical method for the free surfac

Thak you very much Renato:

We now investigating the glass forming process with FIDAP, I have saw the method used in the FIdAP is called "spine" method", it regulate that the the interface moved in the direction of the SPINE, do you have any idear about this method!

by the way , how to set the touch angle ?

regards DP

 Jim_Park April 30, 2006 15:06

Re: about the numerical method for the free surfac

"VOF is older, hard to solve since it work with a step marking function, has good mass preserving and is not able to include some effects, such as surface tension."

The first papers I have describing VOF refer specifically to including surface tension.

Nichols, B. D., C. W. Hirt, and R. S. Hotchkiss, "SOLA-VOF: A Solution Algorithm for Transient Fluid Flow with Multiple Free Boundaries," Los Alamos (New Mexico) Scientific Laboratory, LA-8356, August, 1980, also

Hirt, C. W. and Nichols, B. D., 1981, J. Comp. Phys. v. 39, 201.

Torrey, Martin D., Lawrence D. Cloutman, Raymond C. Mjolsness, and C. W. Hirt, "NASA-VOF2D: A Computer Program for Incompressible Flows with Free Surfaces," Los Alamos (New Mexico) National Laboratory , LA-10612-MS, December, 1985.

Torrey, Martin D., Raymond C. Mjolsness, and Leland R. Stein, "NASA-VOF3D: A Three-Dimensional Computer Program for Incompressible Flows with Free Surfaces," Los Alamos (New Mexico) National Laboratory , LA-1009-MS, December, 1987.

Finally, the commercial CFD code, flow-3d, published by Flow Science, Inc (Santa Fe, NM, www.flow-3d.com) was originally developed from the SOLA series of codes. VOF surface tension effects are a part of their market.

 Renato. April 30, 2006 16:05

Re: about the numerical method for the free surfac

Ok Jim_Park, let me change the word to correct some injustice: "... VOF is not *suitable* to include some effects (*not able* I recognize is too strong), such as surface tension..." - it's widely written in LS articles.

well, I said I'm still novel in this area... ;-)

By the way, I have never gone deepest in VOF references to see what are the VOF extensions and evolutions...

Furthermore, we could also relate some other problems that have been overcame in the last decades by the free surface methods available, like mass preserving, surface reconstruction, hybrid methods mixing VOF+LS, etc...

In essence, VOF and LS are very similar. Both advect some sort of marking function. VOF advects a step function ranging from 0 to 1 and LS advects a signed distance function where the constant unitary gradient of this distance function is employed to model surface tension.

I'm not an expert in VOF but I'm almost sure that surface tension is not straightforward to implement in VOF like is in LS.

I'll read these articles to learn regarding surface tension in VOF ;-)

Regards

Renato.

 Jim_Park May 1, 2006 16:09

Re: about the numerical method for the free surfac

Renato,

I agree that VOF looks like a can of worms to an outsider - and I'm certainly one of those. There's a little research report on the Flow Science web site of a recent enhancement to VOF that you might want to read if you decide to wade through the references I gave above.

The NASA-VOF/3D documentation leaves parts of the surface tension application as "future development." I was told that funding for the work was diverted to analysis of the first space shuttle explosion. So the development was not completely finished for that project.

The Los Alamos group's later work with surface tension went in another direction [look for the "Ripple" code], described in

J. U. Brackbill, D. B. Kothe, and C. Zemach, 'A Continuum Method for Modeling Surface Tension,' submitted to J. Comp. Phys., 1990.

I hope your research is fruitful!

jp

 All times are GMT -4. The time now is 03:41.