[solved ]strange magnetic flux density and flux lines

Numerical methods and mathematical models of Elmer
Post Reply
swilske
Posts: 4
Joined: 08 Mar 2018, 15:48
Antispam: Yes

[solved ]strange magnetic flux density and flux lines

Post by swilske »

Hi I'm Stefan,

I'm new to elmer and want to use it for calculations related to my PhD thesis.

The goal is to get the torque, flux linkage, the flux density in the air gap and in an arbitrary arc around the center for a (switched) reluctance motor.

I followed the instruction given here:
https://www.researchgate.net/profile/Pa ... -Elmer.pdf

..and tried some snippet from:
viewtopic.php?f=3&t=4901&sid=96acf489df ... 0f51aedc40

But I ran into some problems. The magnetic flux density and the flux lines are looking wired.

You can see the distortion of the potential in the following picture (red spot in the air gap at 12 o'clock and the curly flux lines in the rotor at 8 o'clock ).
step2_small.png
second time step with distortion
(145.54 KiB) Not downloaded yet
I wonder if the reason for the distortion may be related to the warning I got : "WARNING:: InterpolateMeshToMesh: Points not found: 4 (found 812)".

Code: Select all

ComputeChange: NS (ITER=17) (NRM,RELC): (  4384.3816     0.15521086E-05 ) :: mgdyn2d
ComputeChange: NS (ITER=18) (NRM,RELC): (  4384.3880     0.14565775E-05 ) :: mgdyn2d
ComputeChange: NS (ITER=19) (NRM,RELC): (  4384.3881     0.25179950E-07 ) :: mgdyn2d
ComputeChange: NS (ITER=1) (NRM,RELC): ( 0.63325128      2.0000000     ) :: mgdyn2dpost
ComputeChange: NS (ITER=2) (NRM,RELC): ( 0.65572497     0.34870602E-01 ) :: mgdyn2dpost
MAIN: 
MAIN: -------------------------------------
MAIN:  Time: 2/4   3.3333333333400001E-003
MAIN: Estimated time left: 14.7 seconds
MAIN: -------------------------------------
MAIN: 
WARNING:: InterpolateMeshToMesh: Points not found: 4 (found 812)
ComputeChange: NS (ITER=1) (NRM,RELC): (  59932.260      1.7273248     ) :: mgdyn2d
ComputeChange: NS (ITER=2) (NRM,RELC): (  61190.413     0.20774846E-01 ) :: mgdyn2d
Here are all relevant project files
reluctance_motor_small.zip
switched reluctance motor with 2 coils, one active with DC current, transient simulation, result *.vtu only for two time steps
(1.68 MiB) Downloaded 284 times
Do you have any advise to get rid of the problems?
For other FEA-software (FLUX) it is recommended to mesh the air gap accordingly to the desired movement of the rotor in the simulation. Does something like this holds for elmer as well?

Stefan
Last edited by swilske on 15 Mar 2018, 11:32, edited 1 time in total.
swilske
Posts: 4
Joined: 08 Mar 2018, 15:48
Antispam: Yes

Re: strange magnetic flux density and flux lines

Post by swilske »

After trying different meshing in the air gap and mesh sizes as well, I was convinced that the problem is related to the warning:

Code: Select all

WARNING:: InterpolateMeshToMesh: Points not found: 4 (found 812)
. Because every time step without warning looks fine.

I found a really good hint here:
viewtopic.php?f=3&t=4652&p=17045&hilit= ... c45#p17021
raback wrote:Hi

Didn't try your case. However, the choices for mortar projector cannnot work. You could try to use these at the air gap:

Code: Select all

  Rotational Projector = Logical True  ! map rotating interface to plane 
  Galerkin Projector = Logical True     ! at the plane perform optimal weak projector
  Projector Skip Nodes = Logical True  ! if interface is nonconducting you can skip the nodes
-Peter
The only thing was, that there is a segmentation fault using the last statement. So without it, everything works fine now.

I search in the solver and model manual but can't find something about it. Are there any other resources for those keywords ?

Stefan
Post Reply