Skin effect not appearing with union of two cylinders in COMSOL

Please login with a confirmed email address before reporting spam

I am modeling a conductive disk in COMSOL under AC conditions (14 GHz and 80 GHz) with the following setup.

  1. Geometry and Physical Setup

Geometry: The conductor consists of two thin cylinders stacked on top of each other and merged using the Form Union geometry operation. Radius (per cylinder): 0.015 m Height (per cylinder): 0.0005 m Total Geometry: A disk with R=0.015 m and total height 1×10^−3 m.

Current direction: The current flows radially across the disk (from one side of the diameter to the opposite side, i.e., perpendicular to the stacking direction).

Physics: Coupled interfaces: Magnetic Fields (mf) and Electrical Circuit (cir). Coupling: The mf interface uses Lumped Ports for input and output, coupled to the cir interface via the External I vs. U node. The components in the cir circuit are connected in series.

  1. Problem / Observation

When analyzing the current density ? in a cross section: - With the Union of the two cylinders: The expected skin effect (concentration of current at the outer edges of the disk, parallel to the current direction) is not visible. The current density is uniform across the cross section.

  • As a control, using only one cylinder (R=0.015 m and total height 0.001 m): The skin effect is observed correctly.
  1. Question I don’t understand why, even though I apply a union to the two cylinders, I don’t get the expected skin effect, just like with a single cylinder?

4 Replies Last Post 2025/12/19 15:53 GMT-5
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 week ago 2025/12/11 19:25 GMT-5
  1. I suggest you post your .mph file to the forum.
  2. The Form Union operation does not physically join the cylinders together. If there is a space between them, there is not a connection.
  3. Are you also modeling computational space around these parts? You should be.

I expect that the specific issues will become clear to readers here if you post your .mph file to the forum. There are many ways to make mistakes in preparing Comsol Multiphysics models so it is not obvious what went wrong here.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
1. I suggest you post your .mph file to the forum. 2. The Form Union operation does not *physically join* the cylinders together. If there is a space between them, there is not a connection. 3. Are you also modeling computational space around these parts? You should be. I expect that the specific issues will become clear to readers here if you post your .mph file to the forum. There are many ways to make mistakes in preparing Comsol Multiphysics models so it is not obvious what went wrong here.

Please login with a confirmed email address before reporting spam

Posted: 3 days ago 2025/12/16 6:40 GMT-5

Hi Robert,

Thanks for your response. I’ve added both the single-cylinder and double-cylinder .mph files as requested.

Hi Robert, Thanks for your response. I’ve added both the single-cylinder and double-cylinder .mph files as requested.

Please login with a confirmed email address before reporting spam

Posted: 3 days ago 2025/12/16 6:42 GMT-5

the files

the files


Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 2 hours ago 2025/12/19 15:53 GMT-5

I am having trouble understanding how/why you have set up this problem (either file you posted) in this manner. I haven't done a lot of skin current calculations, but those in my experience are normally modeled including a computational domain (such as vacuum or air) around the metal parts, since the EM fields in that surrounding region are essential to the calculation. You do not appear to have included any such domains around the metal parts. In addition, lumped ports are normally set up to connect between two electric conductors, which no matter how peculiar their topology may appear, can at least in some sense be thought of as a transmission line. But your port doesn't look like that at all, to me. Where are your two conductors between which the circuit potential is applied? Your specified field direction is perpendicular to the port, which also doesn't make any sense to me. Next, you indicated that your single cylinder model executed to your satisfaction and gave the result you expected. But I executed it (in v6.4) and it failed to converge. So I'm not sure what you saw when you did it. In addition, I'm not sure you are connecting your .ec calculation to your .mf calculation properly either.

Now perhaps some others who read posts here (and are more expert on .ec & .mf coupled models than I am) can offer more helpful suggestions! But meanwhile, I encourage you to take a look at examples of coupled .mf and .ec computations that are provided to you in the Application Library. I often find the Application Library to be very useful.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
I am having trouble understanding how/why you have set up this problem (either file you posted) in this manner. I haven't done a lot of skin current calculations, but those in my experience are normally modeled including a computational domain (such as vacuum or air) around the metal parts, since the EM fields in that surrounding region are essential to the calculation. You do not appear to have included any such domains around the metal parts. In addition, lumped ports are normally set up to connect between two electric conductors, which no matter how peculiar their topology may appear, can at least in some sense be thought of as a transmission line. But your port doesn't look like that at all, to me. Where are your two conductors between which the circuit potential is applied? Your specified field direction is perpendicular to the port, which also doesn't make any sense to me. Next, you indicated that your single cylinder model executed to your satisfaction and gave the result you expected. But I executed it (in v6.4) and it failed to converge. So I'm not sure what you saw when you did it. In addition, I'm not sure you are connecting your .ec calculation to your .mf calculation properly either. Now perhaps some others who read posts here (and are more expert on .ec & .mf coupled models than I am) can offer more helpful suggestions! But meanwhile, I encourage you to take a look at examples of coupled .mf and .ec computations that are provided to you in the Application Library. I often find the Application Library to be very useful.

Reply

Please read the discussion forum rules before posting.

Please log in to post a reply.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.