[Gerry Kan]

Howdy folks:

I am wondering if there is a way to create individual variables at the beginning of solving, and let the solver change it through out.

In my solver, there are a number of solution domain independent variables (think g, but changes in value through time) and I would like to be able to update their values at the start of each iteration while allowing the solver variables (U, p, T, etc) to access them in the inner iteration. Since the value will be identifical to all cells, so I don't want to create an extra scalar to save space and computational effort. How could this be done?

Thanks in advance, GK.

[Gerry Kan]

So it was more straightforward than I thought. As I am still at the prototyping stage, I don't want to be too fancy with my code, but it looks like I can declare variables (scalars, bool, what not) at the solver level and have them accessible later. So for now my problem is solved.

But again, once my solver works, I should think about packaging these variables nicely (in a class, for instance).

Gerry.

[Blumenkind]

Declare them as Fields (for example VolScalarField)
OpenFoam recognizes that they are uniform for your domain and converts them to uniformField, which is saved as one Value and your BC.

[akidess]

Originally Posted by Blumenkind

Declare them as Fields (for example VolScalarField)
OpenFoam recognizes that they are uniform for your domain and converts them to uniformField, which is saved as one Value and your BC.

Why on earth would you run all the machinery to create a uniform field with boundaries, when all you need is a scalar?

[Blumenkind]

You are right. I somehow thought, he wanted to save them and then he would have needed IOObjects.

[Gerry Kan]

Originally Posted by Blumenkind

Declare them as Fields (for example VolScalarField)
OpenFoam recognizes that they are uniform for your domain and converts them to uniformField, which is saved as one Value and your BC.

Hallo Blumenkind: That's the situation I want to avoid. But for now I am declaring them as "globals" and for now I am happy with it. The solver should have access to them in the inner / outer loops, should I need them.

VG, Gerry.