Re: NEC-LIST: comparison of the results of the nec2 and feko

Hello Xiaolin. Your wire segmentation lengths are too small. I
normally use about 1/10 of a wavelength.
In NEC2 which uses the Method of Moments, the usual recommended
segmentation length is typically between 1/8 and 1/20 of a wavelength
(using the double precision PC version of NEC2). For smaller
lengths, convergence tests should be performed on the problem to
ensure that numerical round-off errors are negligible. This is then
a case by case study, and is highly dependent on the geometrical structure.
~Kin

At 18:03 7/30/2009, xlzhang0606 wrote:
>hi,friends,
>
> Recently ,I was modeling a wire-duoconical antenna with the 4nec2, and I
>found that the input impedance from NEC2 differs much with that from FEKO.
>
> NEC2 uses the point match and FEKO uses the Garlekin's Method,
> so I think
>the numerical results of the FEKO should be of more accuracy. What
>should I do
>to reduce the difference between the results from two
>softwares? The nec file is
>listed below:
>
>CE
>GW 1 19 0.0000 0.0000 0.1000 2.3000 0.0000
>4.9000 0.0030
>GW 2 12 2.3000 0.0000 4.9000 0.0000 0.0000
>7.3000 0.0030
>GR 2 6
>GW
>13 1 0.0000 0.0000 0.0000 0.0000 0.0000 0.1000 0.0030
>GE 1
>GN 1
>EX 0 13 1 0 1.0000 0.0000
>FR 0 81 0 0 7.0000 0.2500
>EN
>
>The modeling frequency is between 7-27MHz. When using nec2 (the 4nec2 ), I
>set the segment length as 1/40 of the minmum wavelength initially, and then I
>reduce the segment length and find no good convergence. What mistakes I have
>made in the modeling? or is the nec2 not recommended to model such wire
>antennas with several benches connecting to a single node? I hope someone
>can help me and give me some useful suggestion, THANKs.
>
> sincerely
> Xiaolin Zhang
> 28/07/09
>
>
>
>
>MxRWYEAH.NETCb7QSJOd#:Dz5DVUImCb7QSJOd
>
>--
>The NEC-List mailing list
>NEC-List_at_robomod.net
>http://www.robomod.net/mailman/listinfo/nec-list

-- 
The NEC-List mailing list
NEC-List_at_robomod.net
http://www.robomod.net/mailman/listinfo/nec-list