Isotropic source in Electromagnetic Waves, Time Explicit

Please login with a confirmed email address before reporting spam

Hello Can COMSOL create an isotropic source in 3D or 2D (omni directinoal) in Electromagnetic Waves, Time Explicit ?

If yes, how?

Thank you


5 Replies Last Post 2024年6月23日 GMT+2 16:17
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 2 months ago 2024年6月19日 GMT+2 19:28
Updated: 2 months ago 2024年6月19日 GMT+2 19:29

There are no 3D isotropic sources of RF. The closest thing to that would be a sphere at a uniform temperature, radiating an incoherent unpolarized distribution of blackbody radiation. See https://en.wikipedia.org/wiki/Black-body_radiation . Needless to say, it would be very strange to seek to simulate blackbody radiation with Comsol's RF module. Now, in 2D, you will be (in effect) still modeling a 3D world, but one with no dependencies (of any kind) upon the spatial coordinate perpendicular to the 2D plane. I.e., your 2D model is a slice of a 3D world with "infinite translational symmetry" in the direction perpendicular to that plane. In that case, a "real-world" wire (of infinite length) oriented perpendicular to the subject 2D plane can be represented, in that 2D plane, as a point (or perhaps a small circle) source, which (if assigned an oscillating or otherwise time-dependent current perpendicular to the 2D plane) will radiate isotropically within the 2D plane, and (of course) with no radiation perpendicular to the plane. All of the above holds regardless of whether you are modeling in time- or frequency- domains.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
There are no 3D isotropic sources of RF. The closest thing to that would be a sphere at a uniform temperature, radiating an incoherent unpolarized distribution of blackbody radiation. See https://en.wikipedia.org/wiki/Black-body_radiation . Needless to say, it would be very strange to seek to simulate blackbody radiation with Comsol's RF module. Now, in 2D, you will be (in effect) still modeling a 3D world, but one with no dependencies (of any kind) upon the spatial coordinate perpendicular to the 2D plane. I.e., your 2D model is a slice of a 3D world with "infinite translational symmetry" in the direction perpendicular to that plane. In that case, a "real-world" wire (of infinite length) oriented perpendicular to the subject 2D plane can be represented, in that 2D plane, as a point (or perhaps a small circle) source, which (if assigned an oscillating or otherwise time-dependent current perpendicular to the 2D plane) will radiate isotropically within the 2D plane, and (of course) with no radiation perpendicular to the plane. All of the above holds regardless of whether you are modeling in time- or frequency- domains.

Please login with a confirmed email address before reporting spam

Posted: 2 months ago 2024年6月20日 GMT+2 17:42

Thank you

Thank you

Please login with a confirmed email address before reporting spam

Posted: 2 months ago 2024年6月22日 GMT+2 21:22

How to do it in comsol ?

Thank you

How to do it in comsol ? Thank you

Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 2 months ago 2024年6月23日 GMT+2 00:09
Updated: 2 months ago 2024年6月23日 GMT+2 00:12

Start Comsol Multiphysics. Choose dimension (2D, in your case) from "Select Space Dimension." Select the physics you prefer. In your case, you seem to be interested in "Electromagnetic Waves, Time Explicit (ewte)," although that wouldnotbe my preference. Click on "--> Study." Choose "Time Dependent." Click "Done." Under Geometry, draw a circle big enough to be your radiation space. Draw another much smaller circle, to represent your wire perpendicular to it. Subtract the smaller circle from the larger one using the boolean Difference tool. Under Materials, add a material (air, presumably). In the Model Builder, under "Component 1" expand Definitions. Right click Definitions, choose Function, and pick or define the function you want to represent your current vs. time. Give it an appropriate name. Plot it and make sure you like how it looks. In the Model Builder, look down to Electromagnetic Waves, Time Explicit (ewte) and right click that. Choose "Group by Space Dimension." Right click on Boundaries, choose More, and then Surface Current Density. Select the edges of the little circle in your model. In the Settings section, for Js0, put your function name in the edit box for the z component of the surface current density. E.g., if your function name is mycur, you would put "mycur(t)" in that box. Right click on Boundaries again and choose "Scattering Boundary Condition." Assign the outer boundaries (the circumference of the big circle) to that. Next, right click on Mesh and choose "Free Triangular." You can go ahead and mesh it already, but I wouldn't. I would instead right click Free Triangular 1 and choose Size. Then I would specify a custom "Maximum element size" that was small enough to allow, in my judgement, acceptable resolution in space of the fields of interest to me. Then mesh it. Next, go to Study. In the Settings panel, set the values of interest to you. Especially think about and choose the "Output times" in a reasonable way. Under the Model Builder, right click on Study 1 and choose "Show Default Solver." You can try the default solver as is if you want, although I typically set some features differently. Anyway, save your file, if you haven't already. Once you have everything set the way you want, then in the Settings panel for Study (or elsewhere), click "Compute." It will probably have some issue. If so, fix it and re-run it. But if all seems ok, then wait for it to finish. Next, go to Results. Right click on "Results" and start adding various plot groups to your liking. There is already a default one there for the Electric Field, and you can modify that one. To learn more, see the many, many tutorial (and other) examples provided by Comsol.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
Start Comsol Multiphysics. Choose dimension (2D, in your case) from "Select Space Dimension." Select the physics you prefer. In your case, you seem to be interested in "Electromagnetic Waves, Time Explicit (ewte)," although that would *not* be my preference. Click on "--> Study." Choose "Time Dependent." Click "Done." Under Geometry, draw a circle big enough to be your radiation space. Draw another much smaller circle, to represent your wire perpendicular to it. Subtract the smaller circle from the larger one using the boolean Difference tool. Under Materials, add a material (air, presumably). In the Model Builder, under "Component 1" expand Definitions. Right click Definitions, choose Function, and pick or define the function you want to represent your current vs. time. Give it an appropriate name. Plot it and make sure you like how it looks. In the Model Builder, look down to Electromagnetic Waves, Time Explicit (ewte) and right click that. Choose "Group by Space Dimension." Right click on Boundaries, choose More, and then Surface Current Density. Select the edges of the little circle in your model. In the Settings section, for Js0, put your function name in the edit box for the z component of the surface current density. E.g., if your function name is mycur, you would put "mycur(t)" in that box. Right click on Boundaries again and choose "Scattering Boundary Condition." Assign the outer boundaries (the circumference of the big circle) to that. Next, right click on Mesh and choose "Free Triangular." You can go ahead and mesh it already, but I wouldn't. I would instead right click Free Triangular 1 and choose Size. Then I would specify a custom "Maximum element size" that was small enough to allow, in my judgement, acceptable resolution in space of the fields of interest to me. Then mesh it. Next, go to Study. In the Settings panel, set the values of interest to you. Especially think about and choose the "Output times" in a reasonable way. Under the Model Builder, right click on Study 1 and choose "Show Default Solver." You can try the default solver as is if you want, although I typically set some features differently. Anyway, save your file, if you haven't already. Once you have everything set the way you want, then in the Settings panel for Study (or elsewhere), click "Compute." It will probably have some issue. If so, fix it and re-run it. But if all seems ok, then wait for it to finish. Next, go to Results. Right click on "Results" and start adding various plot groups to your liking. There is already a default one there for the Electric Field, and you can modify that one. To learn more, see the many, many tutorial (and other) examples provided by Comsol.

Please login with a confirmed email address before reporting spam

Posted: 2 months ago 2024年6月23日 GMT+2 16:17

Thank you so much.

Thank you so much.

Reply

Please read thediscussion forum rulesbefore posting.

Pleaselog into 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 theSupport Centerfor a more comprehensive response from the Technical Support team.

Baidu
map