NF2FF Phase

How to use openEMS. Discussion on examples, tutorials etc

Moderator: thorsten

Post Reply
LowDepth
Posts: 20
Joined: Mon 10 Nov 2014, 16:59

NF2FF Phase

Post by LowDepth » Tue 18 Dec 2018, 14:33

Dear openEMS-users,
if I do a RCS calculation in the fashion of the sphere-RCS example, how can I determine the correct complex Phase of the RCS?
Is it done by specifying the radius R? If I don't specify anything the default should be R=1 m. Do I get the complex RCS at distance R from the phase-center, if I multiply the outcome of calcNF2FF by Image... I guess the convention Image is used in openEMS, right?

Hale_812
Posts: 172
Joined: Fri 13 May 2016, 02:54

Re: NF2FF Phase

Post by Hale_812 » Thu 20 Dec 2018, 02:12

I didn't do it in traditional fashion, as in the example. but I did phase extraction with a plane wave box. this instrument extracts the plane wave from itself in the box exterior, so it is zeroed, but leaves the excitation inside the box. If there is something disturbing the field inside it, the the field difference appears outside and on the boundaries, which is dumped and processed by "nf2ff", and if I am not mistaken, the data is complex, or contains phase. I do not have a reference code now to check. You need to explore the resulting structure.
So the trick is that it is very stable when scanning frequency without random phase bias, and you can plot phase walk vs frequency and direction the same time.

thorsten
Posts: 1411
Joined: Mon 27 Jun 2011, 12:26

Re: NF2FF Phase

Post by thorsten » Thu 20 Dec 2018, 09:14

It's of course "-jwt" ;)
And yes, by default everything is for 1m distance, but you can tell nf2ff to calculate for every other radius instead. But I must confess I have never used it since I was never really interested in the absolute phase and I guess there are more things to consider too, like the phase of the Gaussian pulse and so on?

regards
Thorsten

Hale_812
Posts: 172
Joined: Fri 13 May 2016, 02:54

Re: NF2FF Phase

Post by Hale_812 » Fri 21 Dec 2018, 00:47

isn't phase of the pulse removed when normalizing the frequency response at the output? Then again, if it is a true synthetic Gaussian pulse, why should there be deviations? AFAIK it should be flat as a board.

Post Reply