Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Meep – Free simulation software to model electromagnetic systems (ab-initio.mit.edu)
17 points by un_publishable on Aug 3, 2014 | hide | past | favorite | 8 comments


I've got the same question as johnm1019 - I've used FEM for mechanical stress analysis before, but never for electromagnetic stuff - what would the applications of this software be?


The primary focus of Meep is simulating the interaction of light with Photonic Crystals (periodic materials that control light based on both their shape and composition).

http://en.wikipedia.org/wiki/Photonic_crystal

http://ab-initio.mit.edu/wiki/index.php/Meep_Tutorial/Band_d...

You may have heard of optical metamaterials (aka "invisibility cloaks"). They are sometimes made with photonic crystals, but photonic crystals aren't always considered metamaterials. Butterfly wing coloring is a good example.


This is a cool example of solving a problem by developing a general purpose tool, then programming the tool to solve the specialized problem. It's cool for a couple of reasons. First, many aspects of the tool's behavior can be confirmed at the general purpose level, i.e., does it behave according to the expected laws of electromagnetism. Second, it provides the entire community with something useful.


So could this (would this?) be used to simulate antenna design?


MEEP is typically used in the nanophotonics community for simulating things metamaterials, photonic crystals, and waveguides. All of which can be thought of as optical antennas.


You could use it to do that, though my non-expert understanding is that FEM will beat out FDTD for antenna design on many occasions.


It depends on the application. FEM software like COMSOL and CST Studio is typically used for simulating individual antennas. FDTD software like Lumerical is more popular for simulating metamaterials, which are basically large periodic arrays of sub-wavelength antennas.


FEM is typically frequency domain, hence it's good for structures with high Q such as cavity filters, where you know the resonance a-prior. The time domain methods are nicer for transmission lines and antennas since the whole response can be computed at once.

CST has many solvers built it, not just FEM; the time domain is the most popular.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: