Trouble with Magnetic Field Simulation Assembly
Posted Nov 14, 2024, 5:19 p.m. EST Electromagnetics, Geometry, Physics Interfaces 0 Replies
Please login with a confirmed email address before reporting spam
Hi
I'm trying to run a magnetic fields simulation on a model that i created in solidworks, yet im having some trouble getting it to run:
I started by importing my Solidworks model into COMSOL 6.2.0.415 using the import feature in the Geometry tab. I made an assembly in Solidworks and exported the assembly as a solidpart file (because COMSOL gave me source file errors when trying to import the assembly). I have the length unit selected as "From CAD document", and "solids" checked under the section "Objects to import".
After importing my file, I rotated it 45 degrees, placed a repair node in (from right clicking on Geometry -> Defeaturing and Repair -> Repair), and placed a block around the model (to simulate air around the model). I started with Form Union, but when I rendered my mesh, it gave me multiple errors because the Solidworks model is too complicated, so I proceeded to use Form Assembly. In the form assembly tab, I have "create pairs" selected, and the pair type as "identity pair".
Moving down to the magnetic fields interface, I noticed that when placing the node in, free space, magnetic insulation, and initial values are all defaults. Moreover, the magnetic insulation component selects every face inside the box (which is meant to be a magnetically permeable material) and the box itself (which is air). I add in Ampere's Law in Solid's to select the parts inside of the box that are supposed to conduct magnetic fields with the selection "Magnetization" highlighted with a given material magnetization, but when I run my simulation, I run into an issue where the flux is near zero (on the order of E-16 teslas) which makes sense because the tangential component of the fields on the parts of the model that are supposed to conduct magnetic fields are forced to zero; given Gauss' Law for magnetic fields the magnetic field should be zero. Therefore, the fields don't permeate outside of the model.
Given that analysis, what I want to do is have COMSOL assume only the box (air) to have the boundary condition of magnetic insulation, and all the parts of the model inside the box to show up as either (overridden) or (not applicable) in the Domain Selection tab, to where I would naturally select Ampere's Law for Solid's (this would normally work assuming that I would select Form Union in the Geometry tab, because COMSOL merges the boundaries between components in an assembly).
I looked online and used a few resources and I found that creating a continuity boundary between parts of the assembly sets a constraint where the boundary conditions on the outside of the box define everything inside of it; in other words creating a continuity boundary manually forces the boundaries on the outside of the box to be magnetic insulation and everything else inside is free to be assumed with Ampere's Law for Solid's. I read online that this is automatically done with Form Union, and Form Assembly needs this to be done manually.
Given that, I went and checked under "Definitions" to see the Identity Boundary Pair, and it made pairs between each of the components in the assembly. After running it again (where the Magnetic Insulation is still selecting everything inside the box), I run into the same issue. I then try to implement a Continuity Boundary in the Magnetic Field's interface, and I see that all of the components inside aren't selectable (it says "not applicable"). So i then try to select all components inside of the box and add a Magnetic Field's boundary condition, assuming that COMSOL will recognize that if I keep the initial parameter of the boundary condition of the tangential magnetic field to be zero, it will allow the magnetization to conduct in the material.
I either get an error message that says: Divergence cleaning failed for Auxiliary-Space Maxwell (AMS), or I run into the same issue where the magnetic flux doesn't permeate out of the material inside the box of air, and the magnetic field is still around the value of E-16 Teslas (essentially zero).
I even tried creating my own identity boundary between the model inside the box and the box itself (where I selected everything inside the box as the source, and the destination as the box, as well as vice versa, and it still doesn't work)
What am I doing wrong here? Do I have the correct approach when trying to solve this problem?
I attached a picture of my Model Builder Hierarchies, however, I'm not able to link my COMSOL model due to privacy concerns. Could anyone help me?
Edit: I have one section of my model to be a magnetization-based solution, so i set the magnetization myself, and another section is to be influenced by such magnetization, which i set the magnetic properties to be the material's relative permeability. When i run it as is, I get the AMS error mentioned above. When i set both components to be magnetization, it runs fine but it's not what I'm looking for. I have a hunch its an issue between the boundary conditions. Could anyone help me solve this issue?
Attachments:
Reply
Please read the discussion forum rules before posting.
Please log in to post a reply.