HI,

I'm mehdi who asked a bout Schemes of discretization in PHOENICS code III in this FORUM, My problem is not in the use of differencing schemes in PHOENICS but I try to make a comparaison between what I have do in the discretization and what is done in PHOENICS code. I have some difficulties on the stability conditions of the schemes. I want to find a mathematical reference which can help me. In all books what I have, the stability requirements are : the discretised maximum requirement. the requirement of positivity. (like in the book of Malalasekera & Versteeg :An introduction of computational fluid dynamics: the finite volume method) but always they present their work in case of incompressible flow and steady one.

I work on Navier Stokes Equations solved by PHOENICS code specialy in case of free surface. I have discretised the Navier Sotkes Equations with finite volume method and differnt finite difference schemes for convection and diffusion terms and in case of unsteady flow, it's a general case so the manner what I should take didn't exist in books.

Please, if you can help me by a best mathematical reference of PHOENICS EQUATIONS or you make with some persone who know the mathematical style of PHOENICS. If some one are agree, I can send him a part of my work in a document.ps

Thank you of your answer in CFD FORUM.

(1). It is somewhat difficult for me to follow your message. (2). I also had received similar request to review a piece of code being developed by one of the reader. (3). There are two issues here. One is the methods used in a commercial code, the other is the code which a reader is trying to develop. (4). I don't have any solution to this kind of problems. Basically because it is very hard to know exactly what were used in a commercial code over the years, through various researchers and programmers. (5). So, all one can learn is that the method used belong to certain school of algorithm. Definitely, there will be variations of the method and different ways of implementation. (6). So, my suggestion would be: Slow down, and follow an existing approach, which you really understand. Once you have the code in running condition, you can gradually expand it to include other features.

Yes, but can you give a reference

(1). No, I don't have any information about the code. (2). But 20 years ago, I had used the 2-D and 3-D codes developed by the same school of researchers. (3). I am sure that the current version of the algorithm used in the code is definitely very different from the formulation used 20 years ago.