There are several things that need to be corrected in your script:
1) You need to put the (init-fields) line after (set! sources ...) to
properly initialize the source. If (init-fields) comes before the source
specification, as in your case, then the source will be effectively
ignored which is why your output is all zeros (there is also an error
with your amplitude function, see below). This is because the run
function will only call (init-fields), which also sets up the sources,
if the fields have not been initialized. You can find out exactly what
(init-fields) and the run functions are doing by looking at the file
libctl/meep.scm.in in the source code.
2) Note that your computational volume does not include a waveguide and
is just air: the block used to define your waveguide is the same size as
the geometry-lattice and has an epsilon of 1.
3) If you are interested in just the mode propagation within your
waveguide, use perfectly-matched layer (PML) absorbing boundaries to
absorb outgoing waves. Metallic/magnetic boundaries will produce
reflections which interfere with the guided mode(s).
4) Instead of specifying your own continuous wave source which abruptly
turns on and therefore excites spurious high-frequency components, use
the built-in continuous-src function which has a smooth hyperbolic
tangent turn on.
5) Your source amplitude function improperly combines vectors and
scalars which produces an error. The source amplitude must be a complex
number.
6) If you are trying to excite a single mode of the waveguide, use the
eigenmode-source function which is much more efficient at coupling into
a particular mode than a point/line source. For more details, see
Section 4.3 of our book (arxiv.org/abs/1301.5366) as well as the
following tutorial example
(http://ab-initio.mit.edu/wiki/index.php/Meep_Tutorial/Optical_forces)
7) Consider exploiting any symmetries intrinsic to your source and
structure to reduce the size of the computation. This is especially
important in 3d.
On 10/10/2016 09:01 AM, Kapidani Bernard wrote:
Hello I'm new to MEEP. I'm trying to simulate a rectangular waveguide
5 x 2.5 x 10 in size with a PEC termination at the longitudinal end
(the z direction) and PEC boundary conditions on the walls (I use
millimeters as units of space). Since I don't want to deal with what
the field does outside the waveguide, I made the simulation cell the
size of the actual waveguide, and I use the low level functions like
meep-fields-set-boundary to achieve perfect electric conductors on the
boundary of the domain.
Furthermore, I want to set a source at the entrance of the waveguide
so that it has the condition of a wave with frequency 5 GHz travelling
through the waveguide (which should be mono-modal from 3 to 6 GHz)
modulated with the TE_{10} mode spatial basis function. I try to
achieve this by setting the plane at z=inf{z} to PMC and setting an
y-directed electric current with a custom-src source. I'm interested
in probing the tranvserse electring field at the center of the
waveguide, but for some reason obscure to me, when i run meep, no
field propagates and all the outputs are zero, even if I check the
solution on the plane on which I've forced the sources... I attach my
input scheme code and apologize for my english (not my first language,
I try my best to be intelligible)
(reset-meep)
(define-param az 100)
(define-param ax 50)
(define-param ay 25)
(define-param epsr 1)
(define-param mur 1)
(define-param sigma 0)
(set! resolution (/ 5 6))
(set! geometry-lattice (make lattice (size ax ay az)))
(set! geometry (list
(make block (center 0 0 0) (size ax ay az)
(material (make dielectric (epsilon epsr))))))
(init-fields)
(meep-fields-set-boundary fields Low X Metallic)
(meep-fields-set-boundary fields High X Metallic)
(meep-fields-set-boundary fields Low Y Metallic)
(meep-fields-set-boundary fields High Y Metallic)
(meep-fields-set-boundary fields High Z Metallic)
(meep-fields-set-boundary fields Low Z Magnetic)
(define ( sinwt t)
(sin (* 2 pi (/ 1 60) t)))
(define (sinx x)
(sin (* pi (/ (vector3+ x (vector3 (/ ax 2) 0 0)) ax))))
(set! sources (list
(make source
(src (make custom-src (src-func sinwt)))
(component Ey)
(center 0 0 0)
(size (vector3 ax ay 0))
(amp-func sinx))))
(run-until 300
(to-appended "ey-probe"
(at-every 0
(in-point (vector3 0 0 0 ) output-efield-y))))
Thanks in advance,
Bernard
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
