Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
infinite element domain and air conductivity
Posted 11 juil. 2016, 09:46 UTC−4 Low-Frequency Electromagnetics, Fluid & Heat, Heat Transfer & Phase Change, Chemical Reaction Engineering, Geometry, Materials, Mesh Version 4.3a, Version 5.2 2 Replies
Please login with a confirmed email address before reporting spam
Hi everyone,
I’ve got two questions concerning the AC/DC module. I’ve worked with it for a while to simulate an AC multiturn coil in a frequency-domain study. It's located inside a water halfspace and over a halfspace with varying magnetic permeability and electric conductivity. This was always done in the 2.5D axisymmetric mode where I only used sufficiently high model dimensions to avoid any kind of boundary effects/model artifacts. Since the solving time isn't a problem in 2.5D this worked quite well.
However now I need to move to 3D to implement non-symmetric features. After reading an article about infinite elements I decided to use these to keep the model dimensions and therefore the calculation time as short as possible. I didn’t find any advice on how to choose the dimensions of the infinite element domain though. I’m using a sphere (meshing = free tetrahedral) with r=30m which includes a 10m wide outer infinite element layer (spherical with a swept mesh). Is this sufficient for a coil with about 2000 Am²? The properties of the IE domain are chosen accordingly so that it’s centered correctly. Does the size of the IE domain matter anyway?
The second question concerns the modelling of air filled objects. I’ve noticed that in COMSOL the solution doesn’t converge for conductivities below ~0.0001 S/m. Since my lower halfspace has conductivities down to 0.001 S/m it would be nice to use a significantly lower value for air. Is there something like a workaround for this problem?
Kind regards,
Konstantin
I’ve got two questions concerning the AC/DC module. I’ve worked with it for a while to simulate an AC multiturn coil in a frequency-domain study. It's located inside a water halfspace and over a halfspace with varying magnetic permeability and electric conductivity. This was always done in the 2.5D axisymmetric mode where I only used sufficiently high model dimensions to avoid any kind of boundary effects/model artifacts. Since the solving time isn't a problem in 2.5D this worked quite well.
However now I need to move to 3D to implement non-symmetric features. After reading an article about infinite elements I decided to use these to keep the model dimensions and therefore the calculation time as short as possible. I didn’t find any advice on how to choose the dimensions of the infinite element domain though. I’m using a sphere (meshing = free tetrahedral) with r=30m which includes a 10m wide outer infinite element layer (spherical with a swept mesh). Is this sufficient for a coil with about 2000 Am²? The properties of the IE domain are chosen accordingly so that it’s centered correctly. Does the size of the IE domain matter anyway?
The second question concerns the modelling of air filled objects. I’ve noticed that in COMSOL the solution doesn’t converge for conductivities below ~0.0001 S/m. Since my lower halfspace has conductivities down to 0.001 S/m it would be nice to use a significantly lower value for air. Is there something like a workaround for this problem?
Kind regards,
Konstantin
2 Replies Last Post 13 déc. 2016, 00:11 UTC−5
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
Bump
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
There is a valuable information for me about electric conductivity of Air. Thank you.