Volumetric data for Near Field Magnetic Field Strength (H)

How to use openEMS. Discussion on examples, tutorials etc

Moderator: thorsten

Post Reply
Posts: 1
Joined: Tue 08 Jan 2019, 00:07

Volumetric data for Near Field Magnetic Field Strength (H)

Post by roblut74 » Tue 08 Jan 2019, 00:18


I am exploring the use of OpenEMS for 13.56MHz RFID applications. I have two questions based upon my experience so far:

- I see (and can build) examples of how to do 3D plots of the Far Field, but don't see how to do the same with the Near Field. Is there an example or pointers for doing so.

- Is there a way to get H field volumetric data out of OpenCMS (or generate it using ParaView). Bonus points if there is a way that I can build 3D prints of the field (a bit like a Matryoshka Doll) by converting it into a series of STL files?

Very sure more questions will come up as this project moves along, but that you in advance for any help that you are able to give.

- Shane

Posts: 173
Joined: Fri 13 May 2016, 02:54

Re: Volumetric data for Near Field Magnetic Field Strength (H)

Post by Hale_812 » Wed 09 Jan 2019, 00:54

When you define a dump box, you define the data to be dumped, its file name, and storing format.
Examine help AddDump() for details.
It is quite a complete tool, as you can find, capable of dumping not only rough data, but also subsampling and even extracting a frequency domain components.
Define separate identical dump boxes for E and H. When plotting, remember that H-data is shifted by a half of mesh-cell in space and time relatively to E-data. But for considerably dense mesh, you can combine and math-process these directly in ParaView, with small angular error.

Just be sure that you have some free space, like 500Gb-4Tb, when dumping all the fields for a large problem.

And do not try reading large dumped files to OpenEMS for later processing, like extracting a frequency-domain component from transient dump. OpenEMS can not process it through buffer by parts, and tries reading all the 500 Gb to your tiny 32Gb RAM... Instead, dump directly what you need when simulating, or process in ParaVies, which is also very slow and unstable when dealing with large meshes and time evolutions.

Post Reply