[esma]

Hi,
I'm using OpenFOAM to analyze the flow in the suction of a submersible pump.

I've been searching the forums to see if there is a way the velocity components in cartesian coordinates can be transferred to a polar (cylindrical) system.
most of the posts in CFD-online are from older days and I found in one people talk about compiling a code. I'm not familiar with compiling process and I thought these days there may be another way to get this component either in ParaView or as a function in system/controlDict. I couldn't find any clue.
Does anyone know a more straight forward solution than compiling code exercises?

As you can see attached, the velocity components extracted from the slice I need has Ux,Uy and Uz components and I'm looking for Ux, Uy to in radial and tangential Ur & Ut.

Appreciate any help.

[AtoHM]

If you need it only on this cross-section which seems to be perpendicular to a mayor coordinate axis, you can readily calculate that in PV itself. Grab a piece of paper, draw some triangles and derive the transformation needed.

[LuckyTran]

I recommend, using the calculator in paraview and putting in the formula for the transformation. Or saving it to a .csv and doing the same in excel.

There is indeed a fieldFunctionObjects library which contains a fieldCoordinateSystemTransform tool among many others that would let you output in a cylindrical system. Although it is a simple modification of the controlDict... because it is a functionObject, there is a compiling involved. All forms of compiling have apparently been outlawed even though the compiling is done automatically at runtime.

[esma]

Originally Posted by AtoHM

If you need it only on this cross-section which seems to be perpendicular to a mayor coordinate axis, you can readily calculate that in PV itself. Grab a piece of paper, draw some triangles and derive the transformation needed.

Thanks, but I thought about it before. please see the sketch attached. The problem is the location of the velocity vector at the cell center against the X and Y axis is not given. if I knew theta angle that was easy.

[esma]

Originally Posted by LuckyTran

I recommend, using the calculator in paraview and putting in the formula for the transformation. Or saving it to a .csv and doing the same in excel.

There is indeed a fieldFunctionObjects library which contains a fieldCoordinateSystemTransform tool among many others that would let you output in a cylindrical system. Although it is a simple modification of the controlDict... because it is a functionObject, there is a compiling involved. All forms of compiling have apparently been outlawed even though the compiling is done automatically at runtime.

great, thanks for the keyword. I'll try to see if I can implement "fieldCoordinateSystemTransform" in controlDict.

[AtoHM]

Originally Posted by esma

Thanks, but I thought about it before. please see the sketch attached. The problem is the location of the velocity vector at the cell center against the X and Y axis is not given. if I knew theta angle that was easy.

Ok, it is probably not readily available. Two options: I think OF has a function object to export the cell center coordinates. Once you have this, the calculation is possible directly.
Without that, you can use the PointToCellData or CellToPointData filters of PV to either interpolate the coords or the velocity to the points or cell centers and then you can calculate theta.