[schetti1704]

Is there any way to read out all flow data from a CFX calculation file, manipulate it an use to initialize a new calculation?

This is the background to the question:
I am working on the flow through the gap between a large and a small cylinder (much like the flow in a journal bearing). When both cylinders share the same centreline, the geometry is rotationally symmetric and the problem can be solved in a quasi 2D calculation. When the inner cylinder is excentric to the outer cylinder, it becomes three dimensional.
I would like to run a 2D calculation and use the results (flow field) to create a three dimensional .res file with a centric rotor.

I am familiar with the capabilities of user fortran, but haven't found any way to manipulate flow data in a results file.

Any ideas?

[singer1812]

Just use the 2D as is. It will extrapolate the results.

[schetti1704]

Thanks for the quick response.
It does extrapolate, but unfortunately the results aren't useful. It has the velocities in the wrong direction in the 3D case, and pressure also seams to be wrong.

[singer1812]

Are you not using the 2D to initialize a 3D run?
That seems to indicate a problem with your model. The starting values should not be your issue.

If you just want to have the 2D model look 3D in post processing, take a look into instance transforms in CFD POST.

[schetti1704]

The Case Starts with a rotationally symmetric geometry and flow, so a 2D calculation would suffice. Then the mesh deforms in a transient run and the problems becomes non-symmetric and therefore 3D.
If I could get the initial values for the transient 3D run from a 2D calculation, that would save me a lot of time.

[singer1812]

Run a SS 3D case?

[schetti1704]

That is plan B. But much more costly. I have to run 30 of those calculations, all slightly different, 3 million nodes each.
Is there a way to get data from one result file to another?