The partial cell treatment is used to dealing with static curved boundaries. It's used in the NASA-VOF2D/3D code. The only reference given is:

Rivard,W.; Torrey,M.D., "K-FIX: A Computer Program for Transient, Two-Dimensional, Two-Fluid Flow", Los Alamos Scientific National Laboratory report LA-NUREG-6623, 1997.

But the LANL library is now closed for general public so that I can not obtain this technical report.

Is anyone kindly enough to point me to other references, because a search on "partial cell treatment" on Internet yield only a few useless results.

Thanks.

Look into the PARSOL functionality on the CHAM www site. Maybe Mike Malin would like to comment.

Fred.

It seems that PARSOL is what I am search for.

I can not find any detailed material about PARSOL on Internet ("PARSOL", "partial-solid"). Could anybody please point me other references? Thanks.

As a matter of interest Zi-Wei, what is your application?

Cut cell is great for multi part cluttered geometries if you can live with potential boundary layer diffusion effects on the surfaces of the objects cutting the grid. It certainly bypasses the grid creation issues associated with such cluttered geometries.

Fred.

The application is in computer graphics to simulate the motion of some liquid in an irregular container or in a 3D scene with many geometry objects randomly distributed. The diffiusion effect may be okay for computer graphics.

There are a number of commericial codes that employ the Cartesian cut-cell method. In PHOENICS the method is called PARSOL, and in FLOW3D it is called FAVOR (http://www.flow3d.com/CFD-101/fvsbfc.htm). I also believe that the method is used in the VECTIS code of Ricardo Consulting Engineers. Unfortunately, the CHAM website does not as yet provide a description of the PARSOL technique.

There are a number of scientific papers that document Cartesian cut-cell methods. For example, I recall, a paper from Nasa, and also one by Paul Tucker of Dundee University that appeared in Applied Mathematical Modelling. If you are interested I can try and dig out the details of these and other references from a few years ago.

I think that FLOTHERM from FLOMERICS also has this functionality.

That is very likely.

Sicilian and Hirt, J. Comput. Phys. 56, 428-447 (1984).