[Getdp] 2D EM scattering problem

[Amit.Itagi at seagate.com]

Hi,

I am trying to solve a 2D scattering EM problem. I have a circular metal
particle embedded in a dielectric. I have PML (PEC backed) on all sides of
the dielectric. The polarization is TM (magnetic field out of the plane,
electric field in the plane). The unknown in the problem is the scattered
field. I am using a dual HE hp-adaptive approach.  Some more background
info : I am working in the frequency domain. I am using a scaled set of
units in which permittivity and permeability of free space are 1,
wavelength is 2*pi, and angular fequency w=1. I am also working at optical
frequency. Hence all material permeabilities are 1.

Here are my questions:

1) Should the PML region also get adaptively refined ? If not, how do I
turn off  the refinement in the PML.  I find that progressively cells in
the PML closer to the PEC get refined the most. However, the field close to
the PEC is negligible. I guess, the error that I define has the mean fields
(from the H and E formulations). These mean fields are close to zero near
the PEC. Hence, probably, the error is getting amplified. Is there a way to
overcome this ?

2) Also, in the hierarchal basis for the edge elements, I have BF_edge,
BF_edge_2E, and BF_edge_3E. I have to comment out the BF_edge_3E and set
the Adapt P1 range  - 1:2. Otherwise, I get an error "You should not have
reached here in the first place". What am I doing wrong here ?

3) I want to see how the global error is getting reduced as a function of
number of adaptive steps. How do I calculate the global error ? In my
dual.pro file, the error over an element, is the ratio of two integrals
over the element. How do I add up this ratio over all elements ? In short,
I want to calculate two integral quantities over every element, take the
ratio, and add up the ratio.


I am attaching the relevent files. I am also attaching a perl script that I
use to do the hp-adaptivity.

(See attached file: dual.pro)(See attached file: dual.geo)(See attached
file: myScript.pl)(See attached file: solver.par)


Thanks

Rgds,
Amit
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dual.pro
Type: application/octet-stream
Size: 6859 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/getdp/attachments/20070417/5b9a2588/attachment.pro>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dual.geo
Type: application/octet-stream
Size: 2235 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/getdp/attachments/20070417/5b9a2588/attachment.geo>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: myScript.pl
Type: application/octet-stream
Size: 421 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/getdp/attachments/20070417/5b9a2588/attachment.pl>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: solver.par
Type: application/octet-stream
Size: 4894 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/getdp/attachments/20070417/5b9a2588/attachment.par>