grid independence test
 

Dear friends, wish you a happy new year. can anybody tell me how to do grid independence test in phoenics. Thanks

Re: grid independence test
 

Ramesh,

What type of test are you looking for.

I typically define my mesh - run the model. Then refine the mesh by 10 - 25% and repeat until the variation in results in as small as I am prepared to accept.

once the results are stable, the results are independent on the grid. This is what most people call 'grid independence'. I do not know of a feature that will do this automatically. If someone else does please advise.

Hope this helps,

Leon

Re: grid independence test
 

Leon , thank you very much for your suggestion. for one dimensional problem there is no problem of doing this method. but when you deal with 2 dimensional problem having non uniform grid size,i think this method is not that much useful. if you have uniform grid size then there is no problem of doing this method and also i am using this.so do you have any idea for non uniform grid size.i mean in which basis you will increase the no of grids in different regions. should we make it uniform grid size first and then we increase in equal proportion ?

Re: grid independence test
 

Ramesh,

I normally deal with problems in 3D, with non uniform grids, so I know your pain of implementing this approach.

I normally try to divide the area / volume of the cell by a similar percentage over the whole domain.

By using a percentage it tends to keep the cells from stretching and it does not matter if the grid is uniform or non-uniform.

This is a long and tedious task but it is the only one I know.

Hope this helps,

Leon

Re: grid independence test
 

Thank you very much for your valuable suggestion .I am doing in this way ,but I thought there may be a new method in phoenics and that is why I was asking. so thank you for your clarification. Ramesh

Re: grid independence test
 

There is a very good book by Roache that could be of use to you. The following site also contains a brief tutorial on his well documented method. http://www.grc.nasa.gov/WWW/wind/val.../spatconv.html