-
Filter by Topic
Release
Popular Topics
- List all discussions
LinkedIn
Facebook
Twitter
Delicious
Digg
Stumbleupon
More services...Identity pair _PML configuration
|
Thread index | Previous thread | Next thread | Start a new discussion |
RSS feed |
Turn on email notifications |
10 Replies
Last post: May 2, 2012 10:20pm UTC
September 21, 2010 12:57pm UTC
Identity pair _PML configuration
Hi,
I would like to perform a 3D Transmission-Reflection simulation in comsol 4.0 in a similar way as described in the pdf file i have attached.I have a problem of scattering and i would like to configure two PML domains in order to absorb the backward and the forward wave as in the example attached but after I created identity pair and used assembly an error occurres 'failed to evaluate expression' while executing.I had no problem to do it in version 3.5a.
Can anyone help me in order to build the configuration of such 3D simulation in version 4.0?
Many thanks,
Giuseppe.
Reply | Reply with Quote | Send private message | Report Abuse
May 23, 2011 9:38am UTC in response to Giuseppe Parisi
Re: Identity pair _PML configuration
Hi,
I have the same problem, Have you succeeded in modelling it in Comsol V 4.1 ? could you share your experience if you made it?
Regards
Honghui
Reply | Reply with Quote | Send private message | Report Abuse
May 23, 2011 11:50am UTC in response to Honghui Shen
Re: Identity pair _PML configuration
Hi,
you can't implement the identity pair condition analog to version 3.5 for version >3.5.
The only way to overcome the problem of absorbing boundary condition is to implement the scattering formalism with a background field as imput.The attached file is an exemple (which was sent to me from comsol support) of implementation for nano grating periodic structure on a substrate.
I hope this can help you.
Regards,
Parisi.
Reply | Reply with Quote | Send private message | Report Abuse
May 23, 2011 12:34pm UTC in response to Giuseppe Parisi
Re: Identity pair _PML configuration
Hi,
Thank you very much.
I looked into the the model you sent me, I found that it cannot sweep the frequency, how do you sweep the frequency? did you use script?
Regards
Honghui
Reply | Reply with Quote | Send private message | Report Abuse
May 23, 2011 1:21pm UTC in response to Honghui Shen
Re: Identity pair _PML configuration
Hi,
you can add a aprametric sweep like in the modified attached file or you can add a parametric under job configuration.
Consider that if you have not a substrate or a complicated initial field, you can operate with an analitical background field as imput field (for exemple a plane wave) in the 'Electric field components solved for ' under electromagnetic waves panel and use it for the scattered field.
regards
p.
Reply | Reply with Quote | Send private message | Report Abuse
May 23, 2011 3:00pm UTC in response to Giuseppe Parisi
Re: Identity pair _PML configuration
Hi
Thanks a lot
It is strange, I followed your note, added a parameter sweep sub-node to the study node, and added parameter list under the global definitions. but It did not work for me , the results is always the one for 5e14Hz. Could you please take a look my modified mph file in attachment?
Regards
Honghui
Reply | Reply with Quote | Send private message | Report Abuse
May 23, 2011 3:41pm UTC in response to Honghui Shen
Re: Identity pair _PML configuration
hi,
You have not need to add a parametric to job configuration if you have just added a parametric sweep (you can implement only one of them not both).The model i sent you was just modified and complete, you don't have to add anything else.
regards,
g.
Reply | Reply with Quote | Send private message | Report Abuse
May 1, 2012 11:08pm UTC in response to Giuseppe Parisi
Re: Identity pair _PML configuration
Hi, Giuseppe,
I also faced the same problem.
I am just wondering why does the mode need two EMWs (full field and scattering filed)?
Do it solve the absorb BC?
And how can we get the transmission and reflection?
Expecting to get your help!
Million thanks!
Reply | Reply with Quote | Send private message | Report Abuse
May 2, 2012 12:54am UTC in response to Marvin Farrell
Re: Identity pair _PML configuration
I am just wondering why does the mode need two EMWs (full field and scattering filed)?
Do it solve the absorb BC?
And how can we get the transmission and reflection?
"Full field" is used to calculate background field which is required for "scattered field" calculations.
Yes.
You compute the scattered field. Reflectivity is square of the field normalized by incident field. Check the guide on how S-parameters are calculated.
Reply | Reply with Quote | Send private message | Report Abuse
May 2, 2012 7:34pm UTC in response to Alexander Kuznetsov
Re: Identity pair _PML configuration
Hi Alexander
Thanks!
Based on the discussion, I rebuild the model (plasmonic grating) utilizing two EMWs i.e. full- and scattering-field,
However, I found a serious problem.
Actually the model of grating_on_substrate.mph also suffers the same problem.
As we know, there are two solutions (full-field and scattering-field)
The field distributions of the full field- and scattering-field are fine, when the refractive index of the substrate (i.e. n_sub=2) is different from that of the air.
But when I change the substrate as the air, then, the field distribution of the full-field is so disorderly. The field distribution of the scattering-field keeps well.
I really do not know the reason.
Regards!
Reply | Reply with Quote | Send private message | Report Abuse
May 2, 2012 10:20pm UTC in response to Marvin Farrell
Re: Identity pair _PML configuration
Anyone give me some help, thanks
Reply | Reply with Quote | Send private message | Report Abuse
Rules and guidelines
