Hello,
I believe the order of putting pml maybe an issue here, you should put pml
after defining your geometry and source not before. your defination of geometry
is probably overlaying the pml .
good luck.
gursoy b. akguc
________________________________
From: Adam Rogacki <roga...@niowaveinc.com>
To: "meep-discuss@ab-initio.mit.edu" <meep-discuss@ab-initio.mit.edu>
Sent: Friday, June 7, 2013 5:09 PM
Subject: [Meep-discuss] Regarding PML on multiple boundaries
Hello,
I have been running into some confusing issues regarding setting up a PML in an
environment in Meep. In order to simplify posting and understanding, I have
removed most of my code, leaving the problem areas (my PML section) intact.
; Simulation Domain Sizes
(define-param simX 1.84375)
(define-param simY 1.581)
(define-param simZ 1.71129)
(define-param pmlThickness 0.5)
(define-param freq 2.33)
(define-param df 0.5)
(define-param sourceX 0.221875)
; Set-up Simulation Domain
(set! geometry-lattice (make lattice (size simX simY simZ)))
(set! resolution 150)
; Set-up PML
(set! pml-layers
(list
(make pml
(thickness pmlThickness)
(direction High)
(side X)
)
(make pml
(thickness pmlThickness)
(direction High)
(side Z)
)
)
)
(set! geometry
(list
; Waveguide Outside
(make block
(center 0 0 0)
(size 1.84375 0.431 0.14525)
(material (make dielectric (epsilon -1E15)))
)
; Waveguide Inside
(make block
(center 0 0 0)
(size 1.79375 0.381 0.09525)
(material air)
)
; Open Waveguide End
(make block
(center 0.896875 0 0)
(size 0.1 0.381 0.09525)
(material air)
)
)
)
; Set up Sources
(set! sources
(list
(make source
(src
(make continuous-src
(frequency freq)
(fwidth df)
)
)
(component Ez)
(center sourceX 0 0)
(size 0 0.381 0.09525)
)
)
)
(run-until 10
)
The issue I am running into is that I want a PML layer of definable thickness
on the +X, and +Z side of the domain, as seen above. However, I am not sure
that the above PML specification is correct, since I would also imagine that I
could place PML on the -X and +Z sides as well. When I change my PML
specification for the X boundary to be (direction Low) (side X) and try running
the simulation I get the error: "meep: invalid boundary absorbers for this
grid_volume". Since meep finds an error in this configuration, I am wondering
if the above meep ctl file is really simulating what I expect it to be
simulating.
I have simulated my geometry (a rectangular waveguide through the center of the
domain) with a PML on the +X only side and observed the e-field in the
waveguide. I then added a PML on the +Z boundary (ensuring that the PML layer
didn't extend into my waveguide). When I ran this simulation (PML on +Z/+X
boundaries), the efield output was different. This doesn't make sense since
the Z PML didn't extend into the waveguide region.
Any insights would be greatly appreciated.
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
