Axis Symmetric post processing

[gforti]

This an issue with Paraview.
When I calculate the potential integral in the coil volume, the results are equal to those from FEMM.

FEMM:
Integral of A over selection:
2.42274e-006 Henry Amp meter^2

Elmer:
2.424273847989E-006

Paraview:
1.28535e-05

[raback]

Good that things got sorted out. Probably I never understood the issue. So was there a missing weight in the integration?

\int f dA

is different than

\int 2pi r f dA

?

-Peter

[gforti]

We are still trying to figure it out what to do with Paraview. If it is needed, as Elmer is great.

[vencels]

How about creating a new field in the calculator
coordsX*A
where instead of coordsX put the proper variable that represents the distance from axis?

[thalesmaoa]

Sorry, but I'm not completely clear of what result I will achieve doing

Code: Select all

 coordsX*A

There shouldn't be x,y -> r,phi conversion?

From my understanding, Elmer is calculating everything as if (Z-axis) would be 2*pi*r. That is the reason why "SaveScalars" work.

However, when it comes to postprocessing, the results are wrong.

It seems to be considered as Cartesian. Elmer is not giving the full info to paraview.

It seems to lack of one post processing step or some kind of Elmer conversion.

I've asked the same quesiton in Paraview forum

https://discourse.paraview.org/t/axis-s ... ng/13889/3

and it seems that Paraview don't have the ability to understand that the result is axis symmetric.

[raback]

Hi,

In Elmer we assume

1) Cartesian 2D
x = Coordinate 1
y = Coordinate 2

2) Axially symmetric
r = Coordinate 1
z = Coordinate 2

There is nothing that could go to paraview that would indicate how it should intepret them. This is the metadata for coordinates included in VTU file:

Code: Select all

       <Points>
        <DataArray type="Float64" NumberOfComponents="3" format="appended" offset="1338012"/>
      </Points>

-Peter