Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
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.
Electromagnetic field components as integral expressions
Posted 17 nov. 2013, 16:03 UTC−5 RF & Microwave Engineering Version 4.3 1 Reply
Please login with a confirmed email address before reporting spam
I am trying to evaluate the electromagnetic field near the focus of high NA objective. The most common way to do it is to represent the electromagnetic fields in integral form , for instance -using Richards-Wolf diffraction theory.
Is it possible to use integrals in Ex, Ey, Ez input quantities of Port or Scattering Boundary Condition?
Is it possible to import 3D numerical data matrix for each of the Ex, Ey, Ez input quantities and to do some interpolation instead of integral ?
Is it possible to use integrals in Ex, Ey, Ez input quantities of Port or Scattering Boundary Condition?
Is it possible to import 3D numerical data matrix for each of the Ex, Ey, Ez input quantities and to do some interpolation instead of integral ?
1 Reply Last Post 18 nov. 2013, 22:53 UTC−5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I don't know much about Richards-Wolf diffraction theory, but I found this reference
[optics.nuigalway.ie/people/orodriguez/research.html ] from which it seems to be similar to the Stratton-Chu aperture-integral method for computing the radiating near-fields of an aperture antenna. So I have two responses to your question:
1. You can certainly implement a way to compute radiating near-fields based on near-field aperture integration methods, using Comsol's FE-computed aperture fields. (I've done this myself, for some local-illumination or focused antenna problems.) You'll have to set up the complex-vector aperture-field integrals yourself. (Note: Comsol provides a far-field aperture integration routine, but not a near-field one.) Hint: possibly the most important tool you'll need in setting up your near-field expressions is the "dest" operator, which allows you to create certain very important expressions like (for example): sqrt((dest(x)-x)^2+(dest(y)-y)^2+(dest(z)-z)^2), which is useful as the distance from an aperture point to an evaluation point.
2. Comsol Multiphysics is finite-element based and solves Maxwell's equations in differential form. If including both the aperture and the focal region (and all the relevant space between them) within your meshed computation space doesn't consume more memory than you have, you'll be better off (i.e., more accurate) doing the whole problem brute-force within the FE-computation volume, without resorting to creating your own near-field aperture integration expressions.
[optics.nuigalway.ie/people/orodriguez/research.html ] from which it seems to be similar to the Stratton-Chu aperture-integral method for computing the radiating near-fields of an aperture antenna. So I have two responses to your question:
1. You can certainly implement a way to compute radiating near-fields based on near-field aperture integration methods, using Comsol's FE-computed aperture fields. (I've done this myself, for some local-illumination or focused antenna problems.) You'll have to set up the complex-vector aperture-field integrals yourself. (Note: Comsol provides a far-field aperture integration routine, but not a near-field one.) Hint: possibly the most important tool you'll need in setting up your near-field expressions is the "dest" operator, which allows you to create certain very important expressions like (for example): sqrt((dest(x)-x)^2+(dest(y)-y)^2+(dest(z)-z)^2), which is useful as the distance from an aperture point to an evaluation point.
2. Comsol Multiphysics is finite-element based and solves Maxwell's equations in differential form. If including both the aperture and the focal region (and all the relevant space between them) within your meshed computation space doesn't consume more memory than you have, you'll be better off (i.e., more accurate) doing the whole problem brute-force within the FE-computation volume, without resorting to creating your own near-field aperture integration expressions.